Ship Sensors 2024

Ship Sensors
When combined with artificial intelligence, advanced computing architec-

tures and enhanced communications, sensor technologies can monitor vessel
performance and the adjacent environment to detect conditions that may
hinder voyage completion. This book provides insight into the present and
future of sensor architectures and configurations that can enhance vessel per-
formance and further improve the safety of navigation. It covers topics such
as traditional and expanded sensor functions in engineering and navigation,
as well as new sensor capabilities that can provide greater insight into vessel
behavior and performance and enhance awareness of passenger, crew and
other human activities.
Chapters offer background information on typical legacy vessel sensor
configurations and current International Maritime Organization (IMO)
requirements for onboard sensors and future regulatory trends before dis-
cussing modern sensors and current-generation “smart” sensors that provide
enhanced situational awareness to watchstanders. A vision of next-generation
sensors currently being investigated for shipboard use is provided along with
long-term trends in quantum sensing and computing that promise radical
change across a wide variety of vessel functions. Insight is also given into
cybersecurity factors so essential to all sensor systems.
Ship Sensors: Conventional, Unmanned and Autonomous is ideal for pro-
fessional seafarers, maritime academics and university students, and develop-
ers of maritime sensors and systems.
R. Glenn Wright holds a BS in Electrical Engineering, MS in Computer

Science and PhD in Maritime Affairs with 45 years of industry experience
leading research in sensor-based systems for surface vessels and unmanned
underwater vehicles, meteorological and oceanographic data systems, and
the automated testing of electronic equipment. He is also a master mariner
and research vessel captain investigating meteorological, oceanographic, and
electromagnetic phenomena and sensor data fusion, with direct knowledge
of ship operations and the application and usefulness of the foundational
technologies involved. This book complements his book Unmanned and
Autonomous Ships, also published by Routledge.
Ship Sensors
Conventional, Unmanned and

Autonomous
R. Glenn Wright
Cover image: © Johnny Haglund/Getty Images
First published 2024
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2024 R. Glenn Wright
The right of R. Glenn Wright to be identified as author of this work has been
asserted in accordance with sections 77 and 78 of the Copyright, Designs and
Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-P ublication Data
Names: Wright, R. Glenn, author.
Title: Ship sensors : conventional, unmanned and
autonomous / R. Glenn Wright.
Description: New York, NY : Routledge, 2024. |
Includes bibliographical references and index.
Identifiers: LCCN 2023042461 | ISBN 9781032456218 (hbk) |
ISBN 9781032456225 (pbk) | ISBN 9781003377900 (ebk)
Subjects: LCSH: Ships–Electronic equipment. |
Intelligent sensors. | Sensor networks.
Classification: LCC VM480 .W75 2024 |
DDC 623.8/504–dc23/eng/20231108
LC record available at https://lccn.loc.gov/2023042461
ISBN: 978-1 -0 32-4 5621-8 (hbk)
ISBN: 978-1 -0 32-4 5622-5 (pbk)
ISBN: 978-1 -0 03-3 7790-0 (ebk)
DOI: 10.1201/9 781003377900
Typeset in Sabon
by Newgen Publishing UK
Contents
List of Figures xii

List of Tables xiv
Preface xv
Acknowledgments xvii
List of Acronyms and Abbreviations xviii
1 Introduction to Ship Sensors 1

1.1 Ship Sensors 2
1.2 What Exactly Is a Sensor? 2
1.3 The Sensor Revolution 3
1.4 Non-technical Factors Driving Maritime Sensor
System Innovation 4
1.5 Electromagnetic Spectrum Allocation to Maritime
Sensors 4
1.6 Types of Ships and Maritime Vehicles 5
1.7 Sensors Discussed, and Not Discussed 8
1.8 Flow of the Chapters in This Book 8
2 Engineering Sensors 11
2.1 Ship Function, Health and Performance 11
2.2 Types, Functions and Applications 12
2.3 Performance Parameters 14
2.4 Engineering Sensors 15
2.4.1 Fluid and Bulk Level 15
2.4.2 Fluid and Gas Flow 16
2.4.3 Pressure 17
2.4.4 Temperature 18
2.4.5 Humidity 18
2.4.6 Position 19
vi Contents
2.4.7 Vibration 20
2.4.8 Gases 21
2.4.9 Fire 21
2.4.10 Microphones 22
2.4.11 Video Cameras 22
2.5 Sensor Interfaces, Controls and Panels 23
2.5.1 Cables and Connectivity 23
2.5.2 Power Supplies 24
2.5.3 Signal Conditioning 24
2.5.4 Displays and Computers 26
2.6 Sensor Relationships and Engineering System
Automation 27
3 Navigation Sensors 28
3.1 Sensor Use in Navigation 29
3.2 Overview of Navigation Sensor Functionality 30
3.3 Types of Navigation 33
3.3.1 Satellite Navigation 33
3.3.2 Radar Navigation 36
3.3.3 Piloting 41
3.3.4 Radio Navigation 43
3.3.5 Dead Reckoning 45
3.3.6 Celestial Navigation 47
3.3.7 Weather Instrument Use in Navigation 47
3.3.8 Other Sensor Data 48
3.4 The IMO e-Navigation Initiative 49
4 Cargo Sensors 52
4.1 Ships and Cargos 53
4.2 Containerized Cargo 54
4.2.1 Container Hold Sensors 54
4.2.2 Shipping Container Sensors 55
4.2.3 Parametric Rolling 56
4.3 Bulk Cargo 57
4.3.1 Cargo Liquefaction 57
4.3.2 Explosion 58
4.3.3 Unpreparedness to Fight Fires 59
4.4 Liquid Tank Cargo 59
4.4.1 Cargo Tank Overheating 59
Contents vii
4.4.2 Explosion and Fire 60

4.4.3 Another Explosion and Fire 61
4.4.4 Combustion Caused by Welding 62
4.5 Ro-Ro Ships 62
4.5.1 Fires Caused by Lithium-Ion Batteries 63
4.5.2 Refrigerator Unit and Other Fire Sources 64
4.5.3 Fuel Spill 64
4.5.4 Improper Stability 65
4.6 Other Ship Types 66
5 Crewmember and Passenger Sensors 71

5.1 A Burst of Non-traditional Ship Sensors 71
5.2 Identification Cards 72
5.3 Closed Circuit Television (CCTV) 73
5.4 Facial Recognition 74
5.5 Radio Frequency Identification (RFID) 75
5.6 Eye Scanning 76
5.7 Fingerprints 77
5.8 Body Temperature Sensors 78
5.9 Space Occupancy Detection 78
5.10 Accident (Slip and Fall) Detection 79
5.11 Man Overboard Detection 80
6 Artificial Intelligence in Sensor Systems 83

6.1 Smart Sensors 84
6.2 Sensor Data Comprising Multiple Perspectives of the
Same Subject 86
6.3 Machine Learning and Deep-Learning AI 88
6.4 Training Dataset Development 90
6.5 Smart Sensor System Verification, Validation and
Test 91
6.6 Multi-Sensor Data Fusion 93
6.7 AI and Maritime Training 94
6.8 Effects of AI on the Maritime Workforce 95
7 Shore and Intership Communications 101

7.1 Communications Infrastructure 101
7.2 Satellite Communication 102
viii Contents
7.2.1 Very-Small-Aperture Terminal (VSAT)

Satellite Systems 103
7.2.2 Satellite-Based Automatic Identification
System (S-AIS) 103
7.2.3 Remote Vessel Control via Satellite 104
7.2.4 Limitations of Satellite Communications 104
7.3 Very High Frequency (VHF) Radio 104
7.3.1 Digital Select Calling (DSC) 105
7.3.2 Weather Broadcasts, Notices to
Mariners 108
7.3.3 Emergency Services 108
7.3.4 VHF Signal Propagation Effects 109
7.4 Medium Frequency (MF) and High Frequency (HF)
Radio 109
7.4.1 Digital Select Calling (DSC) 110
7.4.2 Weather Broadcasts and Marine Safety
Information 110
7.4.3 Precision Time Broadcasts 113
7.4.4 HF Signal Propagation Effects 114
7.5 Global Maritime Distress and Safety System
(GMDSS) 114
7.6 Visual Communications 115
7.6.1 Lights and Dayshapes 115
7.6.2 Coded Light Signals from Aids to
Navigation 117
7.6.3 International Code of Signals (ISC) 118
7.6.4 Visual and Audible Emergency and Distress
Signals 120
7.7 Sound and Light Signals 122
7.7.1 Maneuvering and Warning Signals 122
7.7.2 Sound Signals in Restricted Visibility 124
7.7.3 Signals to Attract Attention 125
7.8 Summary 125
8 Intraship and Internal Communications 129

8.1 Integrated Bridge Systems 130
8.2 Essential Factors in Sensor Data
Communications 130
8.2.1 Data Quality 131
8.2.2 Data Rate 132
Contents ix
8.2.3 Bandwidth 133

8.2.4 Latency 133
8.2.5 Security 133
8.2.6 Flexibility 134
8.3 LAN (Local Area Network) 134
8.4 Overview of Sensor Data Bus Standards and
Protocols 134
8.5 Ethernet 137
8.6 Controller Area Network (CAN) Bus 138
8.7 NMEA-0183 139
8.8 NMEA-2000 140
8.9 RS-485 140
8.10 Universal Serial Bus (USB) Ports and Devices 141
8.11 Bluetooth 141
8.12 Wireless Fidelity (Wi-Fi) 142
9 Cybersecurity and Ship Sensors 144

9.1 Aspects of Cybersecurity Unique to Sensor
Systems 146
9.2 Cyberattacks and Sensor Systems 148
9.2.1 Engineering Sensors 148
9.2.2 Navigation Sensors 149
9.2.3 Crew, Passenger and Cargo Sensors 152
9.3 Where the World Ends and the Ship Begins 153
9.4 The Attack Surface, or the Protect Surface? 155
9.5 Ship and Port Cyber Infrastructure 156
9.5.1 Ship Infrastructure 156
9.5.2 Port Infrastructure 159
9.6 Present-Day Precautions Against Cyberattacks 161
9.7 Future Directions to Enhance Cyber Security 163
10 Vessel Sensor Degradation 168

10.1 Background and Investigative Approach 170
10.2 Sensor Systems Examined 172
10.3 Sources of Sensor Degradation 172
10.3.1 Biofouling 173
10.3.2 Interference 173
10.3.3 Damage 174
10.3.4 Degradation Modes 174
10.3.5 Symptoms and Characteristics 175
x Contents
10.4 Sensor Degradation Detection 175

10.4.1 Design and Implementation 176
10.4.2 Rapid Prototyping 176
10.4.3 Verification, Validation and Test 177
10.4.4 Physical Research Environment 178
10.5 Detection Results 179
10.5.1 Visible Light Camera 180
10.5.2 Infrared Camera 181
10.5.3 Radar 182
10.5.4 Sonar 182
10.5.5 Lidar 185
10.5.6 Global Positioning System (GPS)/
GNSS 188
10.5.7 Inertial Navigation (INS) 188
10.5.8 Heading Sensor 189
10.5.9 Automatic Identification System (AIS) 190
10.5.10 Engineering Sensors 190
10.5.11 Weather Sensors 191
10.5.12 Satellite Sensors 192
10.5.13 Microphone 193
10.5.14 Electronic Navigation Chart/Electronic
Chart Display System (ENC/ECDIS) 193
10.6 Image Snippets for Remote Analysis 193
10.7 Remote User Interface 194
10.8 Sensor Degradation 195
10.8.1 Identification 196
10.8.2 Remediation 197
11 Next-Generation Sensing 201

11.1 Ship-Based Drone Vehicles 202
11.1.1 Logistics and Resupply at Sea 202
11.1.2 Man Overboard Search and Rescue 203
11.1.3 UAV Use for Sea Ice Navigation 204
11.1.4 Ship Inspection via UAV 205
11.2 Forward-Looking Navigation Sonar (FLS) 206
11.2.1 The Problem of Blind Navigation 206
11.2.2 Forward-Looking Sonar (FLS)
Technology 208
11.2.3 Approach for Estimating Advance FLS
Warnings 208
Contents xi
11.2.4 FLS Installations 211

11.2.5 Integrated Bridge Alerts 212
11.2.6 FLS Information and Warnings 214
11.2.7 FLS Potential to Avert Disaster 217
11.2.8 Acknowledgments 218
11.3 eCelestial Navigation System (eCNS) 218
11.3.1 eCNS Technical Specifications 219
11.3.2 eCNS Problem Identification 219
11.3.3 Present Development Status 221
11.4 Virtual Aids to Navigation (VAtoN) 221
11.4.1 VAtoN Need and Application 222
11.4.2 VAtoN Development 222
11.4.3 Sensor Utilization 223
11.4.4 VAtoN Experiments 224
11.5 Quantum Sensing 229
11.5.1 Quantum Ship Sensor Applications 229
11.5.2 How Quantum Sensing Works 231
11.5.3 Future Directions of Quantum Sensing 232
Index 236
Figures
6.1 Radar Image in Pixels, with Associated Radar Chirp

Waveform Variation over Time 87
7.1 Masthead Lights and Dayshapes 116
7.2 International Code of Signals, Flags and Pennants 119
7.3 Visual Distress Signals 121
9.1 Port and Ship Infrastructure Segments of the Entire Shipping
Process 154
10.1 Ten-Meter Vessel Autonomy Testbed 179
10.2 Degraded Visible Spectrum Camera Images 181
10.3 Nominal and Degraded Infrared Camera Images 182
10.4 Nominal and Degraded Radar Images 182
10.5 Nominal and Degraded Side-Scan Sonar Sensor Images 183
10.6 Nominal and Degraded Fish Finder (top) and Chirp Sonar
(bottom) Images 184
10.7 Nominal and Degraded ForVϋ, LiveVϋ and EchoPilot
Navigation Sonar Sensor Images 185
10.8 Nominal and Degraded FarSounder Navigation Sonar Sensor
Images 186
10.9 Nominal and Degraded Lidar Images 186
10.10 Lidar Degradation due to Laser Interference (Top),
Heavy Rain (Bottom, Left) and Insect Landing on Sensor
(Bottom, Right) 187
10.11 Inertial System Degradation over Time without GPS/GNSS
Inputs 189
10.12 Degraded Engineering Sensor Detection via NMEA-2000
Data Bus 191
10.13 Nominal and Degraded Weather Sensor Images 192
10.14 Nominal and Degraded Satellite Sensor Images 192
10.15 Microphone Placement and Audio Waveforms 193
10.16 Image Snippets to Return to Remote Control Center for
Analysis 194
List of Figures xiii
10.17 Transmission of Snippets via Email Using High Frequency

(HF) Radio 194
10.18 User Interface for Remote Display of Sensor Status 195
10.19 Example Use of Sensor Fusion to Remediate Sensor
Degradation 198
11.1 FLS Transducer Installations 212
11.2 Forward-Looking Sonar Alarm Volume Setting 213
11.3 FLS Electronic Navigation Chart (ENC) Overlay 213
11.4 Ground Track of Costa Concordia, Annotated with Coverage
for FLS Navigation 215
11.5 Aid to Navigation (AtoN) Detection and Identification
Experiment Route 225
11.6 AtoN Detection and Identification Combined with Depth
Contour Following 228
Tables
1.1 Electromagnetic Spectrum Allocation to Maritime Sensors 6

2.1 Example Engineering Sensor Functions 13
3.1 Example Navigation Sensor Functions 31
7.1 VHF International Transmitting Frequencies 106
7.2 MF Single Sideband Maritime Radiotelephone Channels 110
7.3 HF Single Sideband Maritime Radiotelephone Channels 111
8.1 Comparison between Shipboard Sensor Data Bus Standards 136
10.1 Sensor Suite Matrix 178
10.2 Identification of Classes and Types of Degradation Shown
in Figure 10.2 196
11.1 Comparison between Forward-Looking Sonar (FLS) Systems 209
Preface
Ship sensors are being viewed from a new perspective that considers the
fusion of their data, capabilities and performance through a comprehensive
approach that sees every aspect as a cohesive entity rather than a collec-
tion of isolated sensors and systems. Driven by revolutionary technological
advances by vessel autonomy initiatives, new sensor capabilities can enhance
seafarer awareness of vessel status, identify trends in vessel performance
and the adjacent environment that improve potential for voyage success and
detect conditions that hinder and may jeopardize successful voyage comple-
tion. Combined with artificial intelligence, advanced computing architectures
and enhanced communications, the management and control of shipboard
systems can achieve broad new capabilities to improve safety for tradition-
ally staffed vessels and implement new strategies to achieve full independence
for autonomous vehicles.
This book on ship sensors logically follows the author’s previous book,
Unmanned and Autonomous Ships, providing a detailed look at the various
organs and appendages comprising a vessel’s architecture that makes possible
partial and even full automation of various onboard functions. However,
sensors are increasingly prevalent throughout all levels and functions of con-
ventionally staffed ships, making this topic vitally relevant for all vessels. The
ultimate goal is to provide a window into the present and future of sensor
architectures and configurations that can enhance vessel performance, safety
and further improve the safety of navigation. In an attempt to eliminate
potential confusion, unmanned and autonomous watercraft of all types and
sizes are referred to as “vehicles” to distinguish them from the traditional use
of the word “ships” throughout the chapters of this book. The word “ves-
sels” can apply to both ships and vehicles.
Topics covered include traditional and expanded sensor functions in engin-
eering and navigation as well as new sensor capabilities integrated into occu-
pied ship’s spaces and cargos that can provide greater insight into vessel
behavior and performance and enhance awareness of passenger, crew and
other human activities. Background is provided regarding typical legacy
vessel sensor configurations and current IMO requirements for onboard
xvi Preface
sensors and future regulatory trends. Discussed also are modern sensors
and current-generation “smart” sensors that provide enhanced situational
awareness to watchstanders. A vision of next-generation sensors currently
being investigated for shipboard use is provided along with long-term trends
in quantum sensing and computing that promise radical, revolutionary and
game-changing alterations across a wide variety of vessel functions. A topic
unique to this discussion is that of sensor degradation in terms of what it is,
how it may be detected and the potential for overcoming the many limita-
tions that result from its occurrence.
Additional topics include the means by which sensor data communica-
tion is accomplished and how data fusion can achieve far greater promise
to enhance situational awareness and safety of navigation than the sum of
the individual sensors themselves. Insight is provided concerning cybersecu-
rity essential to all sensor systems, the means by which sensor data can be
protected and how confidence can be assured in the information they pro-
vide across an operational environment. Also discussed is how sensor system
design can be verified with respect to the requirements they are intended to
fulfill, validated to ensure they actually accomplish these requirements and
tested to determine their resilience and robustness under a wide variety of
conditions.
Two definitions are offered to help provide context for the discussions in
this book:
Automaton1 (noun) au· tom· a· ton
1 A mechanism that is relatively self-operating, especially a robot.

2 A machine or control mechanism designed to follow automatically a pre-
determined sequence of operations or respond to encoded instructions.
Smart Sensor2
1 A sensor that conditions measured signals before transmitting them to the

control network.
Notes
1 www.merriam-webster.com/dictionary/automaton
2 www.pcmag.com/encyclopedia/term/smart-sensor
Acknowledgments
My eternal love, thanks and appreciation are extended to Magdalena, whose

ability to achieve software miracles and results that make this and all of our
work possible is exceeded only by her inspiration and support throughout
these endeavors. I am forever most grateful.
My thanks to Adam Zgol who greatly aided in performing the experiments
used in detecting sensor degradation described in Chapter 10, and who was
also instrumental in establishing the underwater remotely operated vehicle
(ROV) capabilities used in our research.
Passing mention is made in the way of thanks to my two freelance editors,
Rubi (English Bulldog) and Semper (German Shepherd), who drooled on the
pages of this manuscript they liked and ripped up those they did not like. Far
too often they were correct.
Acronyms and Abbreviations
ABS American Bureau of Shipping

AC Alternating Current
A/D Analog to Digital (Conversion)
AI Artificial Intelligence
AIS Automatic Identification System
AM Amplitude Modulation
APT Automatic Picture Transmission
ARPA Automatic Radar Plotting Aid
AtoN Aid to Navigation
BeiDou Navigation Satellite System (China)
BIMCO Baltic and International Maritime Council
CAN Controller Area Network
CCS Cargo Control System
CNN Convolutional Neural Network
COLREGS Convention on the International Regulations for Preventing
Collisions at Sea
COSPAS Space System for Search of Distress Vessels (Russia)
COTS Commercial Off The Shelf
D/A Digital to Analog (Conversion)
DARPA Defense Advanced Research Projects Agency (United States)
DC Direct Current
DCSA Digital Container Shipping Association
DGPS Differential Global Positioning System
DoD Department of Defense (United States)
DoS Denial of Service
DR Dead Reconning
DSC Digital Select Calling
ECDIS Electronic Chart Display Information System
ECU Electronic Control Unit
eLORAN Enhanced Long-Range Radio Navigation
EHF Extremely High Frequency
List of Acronyms and Abbreviations xix
ENC Electronic Navigation Chart

EPIRB Emergency Position Indicating Radiobeacon
FM Frequency Modulation
FLIR Forward-looking Infrared
FR Fouling Rating
Galileo Global Navigation Satellite System (European Union)
GAN Generative Adversarial Network
GHz Gigahertz
GLA General Lighthouse Authorities of the United Kingdom and
Ireland
GMDS Global Maritime Distress and Safety System
GNSS Global Navigation Satellite System
GPS Global Positioning System (United States)
GT Gross Tons
HEO High Earth Orbit
HF High Frequency
HVAC Heating, Ventilation and Air Conditioning
Hz Hertz (frequency)
IACS International Association of Classification Societies
IALA International Association of Marine Aids to Navigation and
Lighthouse Authorities
IAPH International Association of Ports and Harbors
IBS Integrated Bridge System
IC Integrated Circuit
IMO International Maritime Organization
INS Inertial Navigation System, Integrated Navigation System
IoT Internet of Things
IR Infrared
ISC International Code of Signals
ISO International Organization for Standardization
IT Information Technology
kHz Kilohertz
LAN Local Area Network
LDUUV Large Displacement Unmanned Undersea Vehicle
LEO Low Earth Orbit
LF Low Frequency
LIDAR Light Imaging Detection and Ranging
LORAN Long-Range Radio Navigation
LUSV Large Unmanned Surface Vehicle
mA Milliamperes
MAIB Marine Accident Investigation Branch (UK)
MASS Maritime Autonomous Surface Ships
xx List of Acronyms and Abbreviations
MAtoN Mobile Aid to Navigation

MEO Medium Earth Orbit
MF Medium Frequency
MHz Megahertz
mmRADAR millimeter RADAR
MMSI Maritime Mobile Service Identity
MSC Maritime Safety Committee
MSI Maritime Safety Information
MTS Maritime Transportation System
MUSV Medium Unmanned Surface Vehicle
NavIC Navigation Indian Constellation (India)
NAVTEX Navigational Telex
NIST National Institute of Standards and Technology
(United States)
NMHS National Meteorological & Hydrological Services
NTSB National Transportation Safety Board (United States)
NTSC National Science and Technology Council (United States)
ODU Offshore Drilling Units
OSI Open Systems Interconnection
OT Operational Technology
PC Personal Computer
PCS Port Control System
PLA Programmable Logic Array
PLC Programmable Logic Controller
PM Pulse Modulation
PNT Positioning, Navigation and Timing
PRM Parametric Rolling Movement
QZSS Quasi-Zenith Satellite System (Japan)
RACON Radar Beacon
RADAR Radio Detection and Ranging
RDF Radio Direction Finding
RDP Remote Desktop Protocol
RF Radio Frequency
RFID Radio Frequency Identification
ROC Remote Operations Center
Ro-Ro Roll on-Roll off
RTD Resistance Temperature Detector
SAIS Satellite-AIS
SAR Synthetic Aperture Radar, Search and Rescue
SARSAT Search and Rescue Satellite-Aided Tracking
SART Search and Rescue Radio Transmitter
SAS Synthetic Aperture Sonar
newgenprepdf
List of Acronyms and Abbreviations xxi
SBAS Satellite-Based Augmentation System

SOLAS Safety of Life at Sea convention
SSB Single Side Band
TCP/IP Transmission Control Protocol/Internet Protocol
TEU Twenty-Foot Equivalent Unit
THz Terahertz
TTE Time-Triggered Ethernet
µA microamperes
UMAA Unmanned Maritime Autonomy Architecture
U.S. United States of America
USV Unmanned Surface Vehicle
UUV Unmanned Undersea Vehicle
VATON Virtual Aid to Navigation
VHF Very High Frequency
VHF-ATON VHF Aid to Navigation
VLCC Very Large Crude (oil) Carrier
VLF Very Low Frequency
VPN Virtual Private Network
VSAT Very Small Aperture Terminal
VTMIS Vessel Traffic Management Services
VTS Vessel Traffic Services
WAN Wide Area Network
Wi-Fi Registered trademark of the Wi-Fi Alliance
WIG Wing in Ground craft
WMO World Maritime Organization, World Meteorological
Organization
XLUUV Extra Large Unmanned Undersea Vehicle
Chapter 1
Introduction to Ship Sensors
This book describes a ship’s senses of vision, hearing, smell, touch and taste,
which are vital to how a vessel and her crew, human and cyber, interpret and
react to things that are going on within and to the external world outside the
hull, superstructure and on deck. Events taking place may aid or hinder the
goal of planning, undertaking and completing a voyage. Watchstanders on
the bridge ensure the ship can make way and doesn’t bump into things en
route. Crewmembers in the engineering department are entrusted with pro-
pelling the ship through the water as well as operating and maintaining the
various onboard systems that make the ship come to life. Deckhands ensure
the safe and efficient loading, overseeing, and unloading of cargos and any
passengers that may be along for the ride; except when the number of pas-
sengers begins to exceed crew sizes, where hospitality staff are entrusted with
the duties of passenger safekeeping and morale.
To illustrate the roles played by human beings and ship sensors, two short
stories depicting different points of view of the same sequence of events are
told. The first describes an observer who assigns their own thoughts and feel-
ings to events as they unfold. The second describes the underlying processes
to interpret these events and make appropriate decisions and take action.
One is very human, the other a matter of fact.
Ship Without A Soul

Standing at the rail in the starlit night and listening intently to the silence,
the sea was glazed with skim ice that appeared in the early hours of the
morning. The moon was quartered and shrouded in mystery. To starboard
the fog shimmered into the air, to port the reflected moonlight reached to
the horizon. No wind, no movement, adrift, awaiting our turn to proceed
further.
Then came the sound. Clinking and tinkling like a thousand champaign
glasses. There was a growing sense of anticipation as the echoes grew
louder, yet a pleasantness filled the air. You could feel her presence. A soft
Doi: 10.1201/9781003377900-1
2 Introduction to Ship Sensors
glow appeared in the fog and gradually her sleek lines were revealed.
White, then red shown through the silver apparition.
I knew her name, where she was going and where she was from.
She moved with purpose intent on making her way, knowing I was
there but she would not stay.
Cold and indifferent, no crew is aboard. Emotionlessness without feel-
ings, but nothing untoward.
Neither a ghost from the past, nor anything old. Just an automaton of
today. A ship without a soul.
This, of course, is a story of two ships passing in the night except with a
modern twist where one vessel is unmanned and autonomous, operating
and navigating using her own resources. The ship’s presence and identity
are announced to others using Automated Identification System (AIS), with
her every move choreographed by an automated captain viewing the world
through electronic eyes, ears and other senses based on the knowledge of
what has been learned from terabytes of training data. However, none of
this is possible without conventional and unmanned vessels being laden
with sensors to provide full-time situational awareness across all engineering
functions throughout the ship and navigation functions below, on and above
the sea during the voyage. Both vessels in this story have benefitted from
advances in ship automation. It is hoped that the contents of this book help
to bring these concepts into view in a logical manner.
1.1 Ship Sensors

The use of ship sensors has long been considered in the context of the engin-
eering and deck departments. However, new technologies now also consider
sensor applications that have expanded into the sensing of crew and pas-
senger activity and cargo characteristics, new vessel systems and operational
parameters, and many other factors previously incapable of being observed.
Smart sensor technologies that provide far greater insight into system per-
formance and trends have also blurred these traditional areas of responsi-
bility as interrelationships between factors having to do with engineering,
navigation, deck, passenger and cargo operations become more complex.
Related concepts are also introduced in subsequent chapters to stimulate dis-
cussion on sensor data communications, cybersecurity of sensor applications
and communications, sensor degradation, and future trends in sensor design
and operation.
1.2 What Exactly Is a Sensor?

Simply stated, a sensor may be characterized as a device that detects and
responds (usually with an electrical signal) to an input from the physical
Introduction to Ship Sensors 3
environment.1 In practice, the terms “sensor” and “transducer” are often

confused, but for the sake of discussion, they are interchangeable with each
other. In the context of ships and unmanned vehicles, these inputs can come
from on or within the vessel itself, as in the case for engineering, passenger,
and cargo sensors; or external to the vessel from the surrounding sea, air
and space as may heighten situational awareness of engineering and bridge
watchstanders as well as autonomous agents tasked with advising seafarers
and in decision-making roles to increase safety of navigation. These sensors
are typically connected to electronic displays in engineering spaces and on
the bridge, to internal networks that distribute sensor data and information
derived from sensor data to locations within the vessel for fusion, additional
processing and sensor analytics. Sensor data and information can also be
distributed external to the vessel through broadband communication chan-
nels and via “the Cloud” to remote centers worldwide for vessel monitoring
and control. The analysis of diverse data from multiple sensors using dif-
ferent sensing methods can provide greater insight and dimensionality into
behavior and processes to formulate situational awareness and problem solu-
tions than the individual sensors themselves, with the whole picture being
greater than the sum of the individual inputs. An introduction is provided to
several of these topics in the following paragraphs.
1.3 The Sensor Revolution

New sensors and innovations in their application and use are expanding
our insight into not only the world and the universe that surrounds us, but
also our perception and knowledge of how things operate and the processes
they use. Accompanying these sensors has been the evolution of computer
hardware technology that facilitates the rapid processing of great volumes
of sensor data. Statistical analysis and artificial intelligence-based software
make possible the detection of trends, attributes and objects based on the
characteristics of sensor data contents and enable the subsequent taking of
appropriate action.
One example of how sensor technology has evolved is the breathtaking
imagery obtained in 1995 by the U.S. National Atmospheric and Space
Administration (NASA) Hubble Space Telescope of the “Pillars of Creation”
located some 6,500–7,000 light-years from Earth revealing never before seen
details of three giant dark columns of cold gas. The Hubble telescope repre-
sented the most advanced technology of its era. Yet in less than 30 years the
James Webb Space Telescope in 2022, using newly developed near-infrared
imaging technology, can now see inside these columns to reveal internal struc-
tures filled with many stars and interactions in the process of star formation.2
During this same period enormous advances were made in medical imaging
technologies that include ultrasound, magnetic resonance imaging (MRI)
and computed tomography (CT) that, through the use of modern digital
computing, today makes possible non-invasive visualization of the interior

human body with unprecedented detail and resolution.3 Simultaneously, at
the small end of the size spectrum scientists with the aid of atomic force
microscopy (AFM) are now able to image, measure and manipulate particles
as small as 10 nano-meters (nm) or 0.000000001 meters (m) in size for the
development of new medicines.4
Nestled snugly between galactic and atomic dimensionalities lie ship sen-
sors and sensor technology that have also continued to evolve and change the
face of shipping and ships as we have traditionally known them. Significant
advances have been made across all aspects of ship sensors in terms of moni-
toring the functions and processes of the various onboard engineering, navi-
gation, cargo, and passenger systems which are the subjects of this book.
1.4 Non-t echnical Factors Driving Maritime Sensor

System Innovation
Much of the innovation in shipboard sensor systems stems from industry
and International Maritime Organization (IMO) digital ship and related
initiatives designed to accommodate advances in artificial intelligence and
smart sensors with greater data content, volume and flow, and to integrate
these new and advancing technologies into the regulatory framework.5 These
include developments associated with Maritime Autonomous Surface Ships
(MASS) and generally smaller Unmanned Surface Vehicles (USV); and issues
pertaining to enhanced Safety of Life at Sea (SOLAS), modernization of the
Global Maritime Distress and Safety System (GMDSS), e-Navigation and
resulting enhancements and improved maritime services, and the standard-
ization of shipborne satellite navigation system receiver equipment.
1.5 Electromagnetic Spectrum Allocation to Maritime

Sensors
Any discussion on ship sensors is incomplete without considering the char-
acteristics of the physical environment in which they operate. Using human
senses and our many sensory limitations as a baseline, our visibility into the
world must be considered deficient at best. Many birds and cats see clearer
and at greater distances, especially at night. Dogs have a much better sense
of smell, and wolves can hear over great distances. Whales and elephants
can hear the lowest frequencies; the Great Wax moth has the highest ultra-
sound hearing, while the rabbit has the best overall hearing range amongst
land animals. Insects such as beetles have no hearing at all but make up
for this by being sensitive to vibrations through a sense of touch. However,
human beings have an immense ability to observe other creatures and adopt
their unique skills for themselves. Through observation and experimenta-
tion, we have created sensors that provide us with awareness where we can
communicate great volumes of related information across large portions of

the electromagnetic spectrum, thereby appropriating many common capabil-
ities from other inhabitants of this planet. Table 1.1 illustrates the scope of
many sensors available today, along with their operating and data communi-
cation frequencies in the electromagnetic spectrum.
Note that there exist many references throughout the literature to different
band and sub-band allocations within the Extremely High Frequency (EHF)
millimeter-wave frequencies, including Q, U, E, F, D and G-band designa-
tions not identified in Table 1.6,7 Considering ongoing levels of research and
experimentation with these frequencies and for clarity, we identify mmRadar
as operating in the W-band.
Immediately evident is the fact that seafarers can now have super-human
capabilities to sense objects and events. Indeed, distance is no longer a limita-
tion as we also have the ability to communicate sensor data and related infor-
mation in real time to any location on Earth. However, gaining new powers
to see, hear and communicate provides little advantage without adequate
training in their use. Indeed, a lack of knowledge can often lead to overcon-
fidence in our decision-making capabilities with disastrous results, especially
when we are unaware of the limitations of these new senses.
The evolution of all types of ship engineering, navigation, cargo, and
human activity sensors since 1995 has been remarkable, with increases in
speed, clarity, resolution, accuracy and utility occurring seemingly almost
on a daily basis. The following paragraphs elaborate on these developments
as they presently exist in the maritime world and sheds some light on many
of the near-future trends that continue to revolutionize the industry and dra-
matically affect the lives of seafarers worldwide.
1.6 Types of Ships and Maritime Vehicles

Before describing sensors used onboard ships, it is necessary to provide con-
text for their use in the many different types of ships on which they reside and
the specialized roles they fulfill in world commerce. A partial list of the types
of ships considered in this book includes:
• Passenger Ships • Dry Cargo

• Container Ships • Refrigerated Cargo
• Tanker Ships • Ro-Ro Ships
• Bulk Carriers • Livestock Carriers
• Gas Carriers • Offshore Ships
• General Cargo • Fishing Vessels
newgenrtpdf
Table 1.1 Electromagnetic Spectrum Allocation to Maritime Sensors
6
Frequency Range Frequency Units Sub-B and Maritime Navigation Sensors
RADIO 30–3 00 kHz Low Frequency (LF) Human hearing: Sonar, Acoustic underwater
20 Hz–2 0 kHz communication,
Engineering, eLoran
300 kHz–3 MHz Medium Frequency (MF) MF SSB, GMDSS, Sonar, RDF,
NAVTEX, Engineering
3–3 0 MHz High Frequency (HF) HF SSB communications,
GMDSS, RDF
30–3 00 MHz Very-H igh Frequency (VHF) VHF communications, EPIRB,
VTS, GMDSS, AIS, RDF,
APT
200–2 50 MHz G-b and Few maritime services,
possibly Radar
250–5 00 MHz P-b and EPIRB Satellite Signal, SAR
300 MHz–3 GHz Ultra-H igh Frequency (UHF) 500–1 ,500 MHz L-b and GNSS, Mobile Phone,
Inmarsat, Weather, WAN,
Cargo, EPIRB
2–4 GHz S-b and GNSS, Satellites, Weather,
Microwave, Radars, LAN,
Bluetooth
3–3 0 GHz Super-H igh Frequency (SHF) 4–8 GHz C-b and Communication and Weather
satellites, Wireless, Cargo
8–1 2 GHz X-b and Meteorological, Earth
Observation/C omm. Sats,
Radar, SART
12–1 8 GHz Ku-b and Communication and Weather
satellites
18–2 7 GHz K-b and Communication satellites,
mmRadar
30–3 00 GHz Extreme-H igh Frequency (EHF) 27–4 0 GHz Ka-b and Communication and Weather
satellites
40–7 5 GHz V-b and Radars, Communication
satellites
75–1 11 GHz W-b and mmRadar
OPTICAL 300 GHz–4 00 THz Infrared Light (IR) Infrared Sensors and
Cameras, Lidar
400–7 50 THz Human Visible Light (VIS) Cameras
Source:
ITU 2015. Nomenclature of the frequency and wavelength bands used in telecommunications. Recommendation ITU-R V.431-8 (08/2 015).
International Telecommunications Union, Geneva.

7
These exemplify the vast majority of sensor applications found on modern

and future vessels most likely to be encountered by seafarers and outfitters.
General discussions about common sensor applications across these and
other types of vessels are included in the following paragraphs throughout
this chapter.
Naval ships feature many commercially available engineering and bridge
navigation sensors discussed in this book, which are common among all
ships. However, no attempt is made to address tactical information nor
naval-specific sensors such as acoustic, Sonar, Radar, and shipboard air and
torpedo defense systems unique to operating within the maritime battlespace
for which technical specifications are not readily available.
The text also does not address special-purpose vessels with specific and
unique engineering requirements and sensors beyond those common to most
ships. There are many in this classification, including nuclear and wind-
powered ships, submarines, surface effect ships, hovercraft, hydrofoils, wing-
in-ground (WIG) craft, liftboats, offshore drilling units (ODU), ice breakers,
cable and pipe layers, research vessels, tugboats, barge and timber carri-
ers, salvage vessels, diver support ships, survey vessels, buoy tenders, crane
barges, dredgers, floating cranes and other such vessels.
1.7 Sensors Discussed, and Not Discussed

The heart of this book discusses sensors associated with monitoring engin-
eering, navigation, cargo, and human functions onboard ships. However, not
covered and acknowledgment is made regarding numerous sensors found
onboard ships but also commonly found across many maritime and land-
side applications. These include sensors associated with the operation and
monitoring of lifts and elevators, housekeeping and food services, cooking
facilities, swimming pool and spa machinery, entertainment systems and
other functions that exist on ships, from the smallest of vessels to large-scale
and complex operations that rival the world’s largest hotels, restaurants and
catering establishments.
1.8 Flow of the Chapters in This Book

The chapters of this book are organized to introduce many of the essential
elements of shipboard sensor systems, provide specifics on several of their
key features and then touch on some of the peripheral subjects that directly
affect sensor operations. This is followed by discussion of the author’s per-
ception of future near and long-term trends in sensor design, technology and
applications.
Chapters 2–10 focus on the present state of sensor technology, describe

many types of sensors likely to be found in and on various modern ships and
unmanned vehicles and highlight some important sensor-related issues being
considered throughout the industry.
Chapters 2, 3, 4 and 5 discuss sensors from perspectives specific to func-
tions associated with engineering, navigation, cargo, crewmembers and pas-
sengers, respectively.
Chapter 6 provides background into essential technologies associated with
artificial intelligence (AI) and how these are revolutionizing sensor designs
by creating smart sensors and sensor analytics that provide unprecedented
insight into sensed phenomena and processes. Also discussed is the use of
communications as a means to extend human senses worldwide.
Chapter 7 identifies many of the methods by which sensor data is commu-
nicated between ships and from ship to shore.
Chapter 8 describes how sensor data is communicated within the ship
itself.
Chapter 9 addresses the critical subject of sensor system cybersecurity and
how this fits in with overall vessel cybersecurity.
Chapter 10 addresses the subject of sensor degradation that stymies sensor
operation and performance and how this phenomenon can adversely affect
vessels of all types in voyage completion, as well as introduce factors that
decrease the safety of navigation and
Chapter 11 bridges the gap between the previous chapters, future enabling
technologies, and the directions sensors will likely to take regarding their
application and use.
References
1 Robert Sheldon. What Is a Sensor? WhatIs.com, TechTarget. August 2022.
www.techtarget.com/whatis/ definition/sensor
2 Tereza Pultarova. Why the James Webb Space Telescope’s Amazing ‘Pillars of
Creation’ Photo Has Astronomers Buzzing. Space.com. 24 October 2022. www.
space.com/james-webb-space-telescope-pillars-creation-excites-astronomers
3 Dennis Hung. The Evolution of Medical Imaging and Its Impact on the
Healthcare System. www.healthworkscollective.com/evolution-medical-imag
ing-impact-healthcare-system/
4 Yuki Takechi-Haraya, Yukihiro Goda, and Kumiko Sakai-Kato. Imaging and
size measurement of nanoparticles in aqueous medium by use of atomic force
microscopy. Anal Bioanal Chem. 2018 Feb;410(5):1525–1531. doi: 10.1007/
s00216-017-0799-3. Epub 18 December 2017.
5 International Maritime Organization. Resolution A.1149(32) Adopted on 15
December 2021 (Agenda item 9(a)) Revised Strategic Plan for the Organization
for the Six-Year Period 2018 to 2023. A 32/Res.1149 28 January 2022.
6 David Sumi. V and E Bands Key to the 5G Revolution. July 2019.

BroadbandCommunities Magazine, www.bbcmag.com/rural-broadband/v-
and-e-bands-key-to-the-5g-revolution
7 mm wave frequency bands. RF Wireless World. www.rfwireless-world.com/
Terminology/mm-wave-frequency-bands.html
Chapter 2
Engineering Sensors
Engineering sensors have evolved to support multiple roles and functions

across the entire ship, with many active and passive sensors designed to
measure vessel systems and critical infrastructure parameters and detect
abnormal conditions. This is distinguished from sensor systems used on the
bridge to assist with tasks such as navigation, safe passage and communi-
cation with other vessels and those designed primarily for monitoring pas-
senger and crew activity, and cargo conditions. However, there is an overlap
in rudder position, propeller shaft speed, thrust indication and other sensor
applications where both engineering and deck seafarers have shared interests.
This chapter considers the various tasks performed by sensor systems hav-
ing to do primarily with traditional engineering functions onboard ships.
Different types of sensors typically found on most ships are described
along with performance characteristics that help to determine whether the
machinery, other apparatus, or processes they monitor are operating within
or outside of nominal limits, or detecting events as they occur. The topic of
“smart sensors” is discussed highlighting their use and advantages towards
increasing safety and improving performance. Finally, shipboard process
automation is discussed, emphasizing sensor fusion and how these systems
can extend and enhance sensor performance.
2.1 Ship Function, Health and Performance

Engineering systems cover many functions with unique modalities and char-
acteristics. These are generally associated with engine, propulsion, steering,
electrical, heating, ventilation and air conditioning (HVAC); distillers, con-
densers, evaporators and associated pumps, valves, and piping. Control and
health monitoring of engineering systems distributed throughout the vessel
is accomplished through the use of a network of actuators and sensors to
monitor nominal machinery performance and generate alarms resulting from
out-of-tolerance and abnormal conditions. Vast improvements have been
made in monitoring engines and drive trains, bearings, and other compo-
nents that consider temperature, wear and vibration, water infiltration into
Doi: 10.1201/9781003377900-2
12 Engineering Sensors
oil as well as monitoring exhaust gas and cylinder liners. Significant advances
have also been made using new technologies for level sensing, water ingress,
pressure monitoring, winch torque and load monitoring, and many other
engineering sensing applications.
In addition to the above, other inputs from the engineering environment
include motion, moisture, torque, velocity, acceleration, pressure, audio, and
video. These inputs are often represented as a function of range, intensity,
volume, resistance, conductance, and other characteristics of mechanical and/
or electrical (analog and/or digital) signals created by sensors and sent dir-
ectly to displays or indirectly using one or more data communication buses.
Such functions form the underpinnings of engineering sensors for onboard
systems that indicate system performance characteristics such as vacuum,
quantity, liquid and gas flow, voltage and electrical current flow. The wide
range of measured parameters and characteristics of data obtained from sen-
sors used in engineering applications prohibit detailed discussion short of a
full text dedicated exclusively to this task. However, generalizations can be
made based on sensor function.
2.2 Types, Functions and Applications

Sensors are recognized primarily according to the things they measure. A par-
tial list includes:
• Pressure • Angle • Capacitance • Fluid • Open/

properties Closed
• Temperature • Direction • Impedance • Heave/ • On/Off
Pitch/Roll
• Humidity • Force • Power • Air flow • Error
• Position • Strain • Frequency • Audio • Other
• Rotation • Torque • Count • Video
• Speed • Vibration • Fluid level • Motion
• Velocity • Voltage • Fluid flow • Noxious
gas
• Distance • Current • Fluid • Time
capacity
• Acceleration • Resistance • Fluid • Date
remaining
Some sensors have the capability to measure multiple characteris-

tics. Table 2.1 provides an overview and summary of various engineering
Engineering Sensors 13
Table 2.1 Example Engineering Sensor Functions a
Function Signal Descriptions
Alternating Voltage, Current, Power, Frequency, Average Frequency, Average L-

Current N Voltage, Average L-L Voltage, Phase A Frequency, Phase A L-N
Voltage, Phase AB L-L Voltage, Phase B Frequency, Phase B L-N
Voltage, Phase BC L-L Voltage, Phase C Frequency, Phase C L-N
Voltage, Phase CA L-L Voltage.
Direct Current Current, Voltage, Battery Temperature, Battery State of
Charge, Battery Time Remaining, Ripple Voltage, Power.
Electrical Resistance, Conductance, Impedance.
Electric Switch/Breaker, Breaker Current, Hardware OFF Counter,
Distribution Hardware ON Counter, Hardware ERROR Counter, Hardware
ON Timer, Hardware OFF Timer, Hardware ERROR Timer.
Engine Monitor Order, Response, Boost Pressure, Rotations per Minute
(RPM), Hours, Coolant Pressure, Coolant Temperature,
Exhaust Gas Temperature, Fuel Pressure, Fuel Rate, Oil
Pressure, Oil Temperature, Tachometer, Tilt/Trim, Voltage,
Fuel Economy, Fuel Consumption, Percent Load, Percent
Torque, Trip Fuel Used, Vibration, Warning Indicator.
Fluid Flow Flow Rate, Trip Volume.
Fuel Management Distance to Empty, Time to Empty, Total Fuel Capacity,
Total Fuel Level, Total Fuel Consumption, Total Fuel Used,
Total Fuel Economy, Total Fuel Rate, Total Fuel Remaining.
Hull Opening Status, Acceleration, Stress, Heave, Roll, Pitch.
Humidity Inside, Outside.
Indicator Status, Hardware OFF Counter, Hardware ON Counter,
Hardware ERROR Counter, Hardware ON Timer,
Hardware OFF Timer, Hardware ERROR Timer.
Mechanical Force, Strain, Torque.
Motion Acceleration, Angle, Angular Velocity, Angular Acceleration,
Distance, Rotational Rate, Velocity.
Pressure/Vacuum Barometric Altimeter, Compressed Air, Engine Boost,
Engine Coolant, Engine Fuel, Engine Oil, Hydraulic Oil,
Steam, Transmission Oil, User Defined, Water.
Propeller Rotations Per Minute (RPM), Pitch, Order, Response.
Rudder Angle, Order. Response.
Sound Level, Decibels, Direction.
Tanks Capacity, Level, Remaining, Properties.
Temperature Battery, Engine Coolant, Engine Oil, Engine Room, Exhaust
Gas, Freezer, Heat Index, Heating System, Inside, Outside,
Refrigeration, Sea, Transmission Oil, User Defined, Wind Chill.
Thrusters Status, Direction, Thrust %/R PM.
Time Local Date, UTC Date, Moon Phase, Sunrise, Sunset, Local
Time, UTC Time, Twilight AM, Twilight PM.
Transmission Gear, Oil Pressure, Oil Temperature, Warning Indicators.
Vessel Heave, Pitch, Roll, Trim Tabs, Stabilizers, Keel.
Video Visible Imagery, Infrared Imagery.
Note:
a
Derived in part from National Marine Electronics Association (NMEA) 2000/I EC 61162-3 .
Maritime navigation and radiocommunication equipment and systems – Digital inter-
faces – Part 3: Serial data instrument network.
applications and the characteristics measured in monitoring specific func-

tions. Some overlap exists between functions for the sake of clarity and
comprehensiveness.
Although there is great commonality among vessels, this table should be
viewed as representative and general in nature due to the wide variety of
ships and unmanned vessels and vehicles, their diversity in form and function
and the extensive range of tasks and missions they perform. Engineering-
related applications in which sensors are used can be classified into three
general categories: ship hull and superstructure; engines, propulsion, and
generators; and electrical, HVAC, and other auxiliary systems. Some of the
locations on the ship where engineering sensor use is common include the
hull (strain, fatigue, damage, motion, acceleration, opening status), service
and cargo tanks (level, pressure, capacity), cargo holds (level, gases), engine
room (temperature, pressure, vacuum, speed, flow, position, sound, video,
consumption, load, torque, time, gases) and enclosed spaces (carbon mon-
oxide, oxygen, explosive gases).
2.3 Performance Parameters

While a sensor detects and responds to physical input from the environment
to produce an output (generally electrical), measuring a physical input and
converting it into data can bring loss or distortion that can change the value
or meaning of what is measured. These parameters can be viewed as static or
dynamic; static is generally associated with a steady state input and dynamic
with a changing state input.1,2 Common parameters include:
• Transfer Function: A mathematical function that models sensor output

for each possible input.
• Linearity: The deviation of sensor output from a straight line based upon
a straight-line input.
• Range: Span of inputs the sensor can measure with a specified accuracy.
• Sensitivity: Ratio between change in sensor output to change in sen-
sor input.
• Hysteresis: The degree to which a sensor output lags behind changes in
the sensor input.
• Repeatability: The ability to consistently obtain the same output when
applying the same input.
• Drift (also stability): Ability to maintain the same output when the input
does not change over time.
• Accuracy: The degree to which the result of a measurement conforms to
the correct value.
• Resolution: Smallest possible input that is detectable by a sensor.
• Threshold: Amount of change required at an input before a sensor changes
the output.
Static and dynamic parameters are often a reflection of the quality of the
sensor. Sensors can also be classified as being either analog or digital. Analog
sensors produce an output that is continuous over time in proportion to the
input, whereas digital sensors produce pulses or digital words corresponding
to the input. Analog and digital sensors, where the output alternates between
two fixed levels, are binary sensors that effectively function as on-off switches.
Analog signals are usually converted to digital for transmission and analysis.
2.4 Engineering Sensors

Ships come in all sizes, exist for different purposes and perform a wide variety of
functions. It is only natural that ship sensors follow the same pattern. There are
many different sensors; some perform the same functions yet look completely
different, while others look alike but perform different functions. The following
paragraphs discuss the most common sensors used onboard ships of all types.
2.4.1 Fluid and Bulk Level

Sensors measure the levels of different fluids in tanks, bulk cargo in holds
and elsewhere onboard ships and are designed according to the character-
istics of what they measure. These include potable, ballast, bilge and black
water; hydraulic fluid, fuel, lubricant oil, waste oil and liquified gases such
as ammonia and natural gas. Some of the different types of fluid level sensors
are listed below:
• Resistive: One of the more basic sensors is simply a resistive element

attached to a float, lever, or other device where the value of electrical
resistance changes as the device moves up or down. Some types of fluid
level sensors may be stacked together with several in line for deep tanks
beyond the effective sensing range of a single sensor.
• Hydrostatic: These sensors measure liquid column static pressure that
varies in proportion to weight to calculate fluid level.
• Capacitance: Radiofrequency (RF) signals vary due to dielectric changes
with the rise and fall of fluid level using a transmitter mounted externally
to the tank.
• Radar: Microwaves are directed downward from the top of a tank or
vessel, reflected back from the surface and then measured form the basic
principles involved in Radar-based level sensors.
• Ultrasound: Sound waves are emitted, and the reflected waves from the
surface are measured, with the transit time being proportional to the dis-
tance traveled.
In addition to cost, the key considerations for determining the proper type
of level sensor to be used in any application include the characteristics of
the material to be measured in terms of solid, liquid, slurry, or powder; the

pressures and temperatures at which the material is kept in the container or
vessel, flammability and explosion potential, and the nature and characteris-
tics of the container or vessel itself within which the material is kept onboard
the ship.
2.4.2 Fluid and Gas Flow

Flow can be measured as the mass or volume of a fluid or gas that passes
through a particular point in a pipe, duct, or other conduit or structure. This
is generally accomplished by taking a velocity measurement combined with
a factor that considers the cross-section at the area where the measurement
takes place. Additional factors that must be considered are the density and vis-
cosity of the fluid or gas, which can vary widely and contribute to variations
in flow measurements. Density is defined simply as mass per unit volume. It
is thought of as the “weight” being measured based on the number of mole-
cules occupying a given space (volume), whereas viscosity is the internal fric-
tion or resistance that exists within the internal layers of a fluid or gas. Both
density and viscosity decrease as the temperature increases. However, density
follows a fairly linear relationship to temperature, while viscosity mostly has
an exponential relationship with temperature. Flow rates are expressed in
units of volume or mass.
Fluid level and flow measurements are complementary and help deter-
mine how much of something exists and is available to go over a distance
or for a period of time. Flow measurements are used in determining fuel
oil consumption, monitoring engine coolants, controlling draught, trim and
list through ballast water movements; balancing water systems, monitoring
seawater inlets and monitoring the onload and offload rate of gas and liquid
cargo transfer.
Different types of gas and fluid flow sensors most commonly found on
ships include:
• Thermal Anemometers: This sensor operates based on the relationship

between the amount of heat removed from a heated temperature sensor
and the velocity of the gas or fluid passing over the sensor.
• Differential Pressure: Used for both gases and liquids, flow rate measure-
ments are based upon the pressure drop across measured points, which is
proportional to the square of the flow rate.
• Ultrasound: Sound waves are used to determine flow using either of two
methods. One method is based on frequency shifts (Doppler effect) caused
by liquid flow, whereas the other is based on the time it takes for sound
to travel from one point to another based upon the direction and rate of
the fluid being measured.
• Laser Doppler: This sensor uses a continuous wave laser combined with
a laser detector to determine the change in wavelength caused by the fluid
or gas being measured as a velocity function.
• Mass: The flow rate of gas or fluid mass is often measured to determine
fuel-to-air ratios for proper combustion.
• Vortex: Flow measurements are based upon the creation rate of eddies
and vortexes in proportion to the velocity of gas or fluid flow passing
around a fixed object.
• Turbine: The gas or liquid flow rate through a propeller or turbine will
cause it to turn or spin at a speed directly proportional to the flow.
Key considerations for determining the proper type of flow sensor include
the characteristics of the gas or liquid to be measured in terms of purity, vis-
cosity, density, liquidity or slurry, the pressures and temperatures at which
they operate, and the properties and characteristics of the path in which the
sensor is placed and operates.
2.4.3 Pressure
Pressure is measured by converting physical energy exerted against a surface
into an electrical signal that is proportional to the force. Various types of
pressure sensors exist using different technologies to make measurements.
Pressure measurements are taken at various locations on a ship in support
of functions such as engine fuel and lubricating oil, manifold vacuum and
coolant; potable water, firefighting equipment, hull strain and fatigue,3,4 shaft
torque, and winch loads. Sensors commonly found on ships that are used to
measure pressure include:
• Piezoresistive: Solid-state devices embedded within a sensor change their

electrical resistance values in accordance with the pressure exerted on the
sensor.
• Piezoelectric: Crystalline and ceramic devices embedded within a sensor
that change their electrical polarity in accordance with pressure exerted
on the sensor.
• Variable Capacitance: Materials embedded within a sensor vary in their
ability to store an electrical charge according to pressure exerted on the
sensor.
• Variable Reluctance: Materials embedded within a sensor that vary in
magnetic reluctance (resistance to magnetic flux) in accordance with pres-
sure exerted on the sensor.
• Fiber Optic: Fibers embedded within a sensor that change their optical
characteristics to enable measurement of the phase differences between
light signals that have taken different paths.
The four kinds of pressures include gauge, differential, absolute, and

vacuum. All are measured in units of pounds per square inch (psi) or bars.
Gauge pressure is measured in relation to atmospheric pressure. It is positive
(above zero) when the pressure is greater than the ambient pressure (14.7 psi/
1.01 bar at sea level) and negative when below ambient pressure. Differential
pressure is a relative measurement of the difference in pressure between two
measured points. Absolute pressure is a measurement against an absolute
vacuum. Vacuum pressure is a measurement of pressure that is below normal
atmospheric (ambient) pressure.
2.4.4 Temperature
Temperature is an expression of how hot or cold something is compared
to a scale. Temperature measurements are made throughout a ship include
engine coolant, cargo, exhaust, passenger and crew quarters, and food freez-
ers. Both contact and noncontact measurements may be taken. Examples of
different types of temperature sensors include:
• Electro-mechanical: Two dissimilar metals are physically bonded together,

forming one device. As temperature changes, the two metals different
rates of expansion and contraction cause torsion and physical movement
that can be calibrated to temperature.
• Thermocouple: Temperature measurements are made using two different
metal alloys that produce a small electrical signal when the temperature
changes at one junction with respect to the other.
• Resistance: Resistive sensors are composed of two or more bonded met-
als or conducive polymers that change in electrical resistance as the tem-
perature changes. Resistance may increase or decrease with temperature,
depending on the materials used.
• Infrared: Noncontact measurements of liquid or material temperatures
may be performed using infrared sensors that sense heat (or energy) in
direct proportion to temperature.
• Capillary: Liquid expansion or contraction properties within a reservoir
are often used to complete an electrical circuit between two contacts.
Temperature is generally expressed in degrees Fahrenheit or Celsius. Water

freezes at 32°F/0°C and boils at 212°F/100°C. The Rankine and Kelvin scales
are based on the Fahrenheit and Celsius scales, referencing absolute zero
(−459.67°F/−273.15°C).
2.4.5 Humidity
Humidity represents the concentration of water vapor that exists within a
gas. Absolute humidity refers to the actual amount of water present in terms
of weight or volume. Relative humidity is the ratio between the absolute

humidity and the saturation point at the same temperature. Measurements of
humidity on ships are performed to monitor condensation levels in sensitive
cargo and spaces; heating, ventilation and air conditioning system perform-
ance; maintain environmental conditions for proper operation of sensitive
equipment, determining conditions conducive to fire and combustion, and
for weather reporting and forecasting. Common types of humidity sensors
include:
• Capacitance: Humidity is measured based upon the effect of absorbed

water vapor on the ability of a sensor to store an electrical charge. As
water vapor increases, the capacitance also increases based upon changes
in the dielectric constant (permittivity) of the sensor materials.
• Resistance: Similar to a capacitance sensor, humidity is measured based
upon the effect of absorbed water vapor on the resistance property of a
sensor. This is usually an inverse relationship to humidity, where resist-
ance decreases as humidity increases.
• Thermally Conductive: Humidity is measured using thermistors to deter-
mine the difference in thermal conductance between ambient air and a dry
air reference (e.g., nitrogen), resulting in a change in resistance directly
proportional to humidity.
Absolute humidity is measured in units such as grams per cubic meter

(g/m3), whereas relative humidity is measured as a percentage compared to
saturation.
2.4.6 Position
This term is defined within the context of engineering departments on ships
as being the location of something in a particular space or relation to another
object or reference. Position sensors are designed to detect movement and
convert this into signals for transmission, reporting, or control. Examples
include rudder position to port or starboard of the ship’s centerline; throt-
tle being open, closed, or somewhere in between; and the detection of limits
such as high vs. low, on vs. off, up vs. down, open vs. closed, etc. They can be
divided into linear, rotary, and angular categories. Several different types of
position sensors are likely to be installed throughout most ships and vessels.
Common types of position sensors include:
• Synchro/Resolver: A synchro converts an angular position into an elec-

trical signal transmitted via wire connection to a remotely located resolver,
where the angular and rotational position is physically reproduced.
• Limit: Limit sensors are generally electromechanical devices used to detect
the presence or absence of an object. Their use is associated with limiting
the travel of an object to within specified limits.
• Resistive: A resistive sensor can be either linear or angular, usually a

potentiometer forming an adjustable voltage divider using a three-terminal
resistor. The resistance value changes with the position of a brush on the
resistor, resulting in a variable electrical output dependent upon position.
• Proximity: Several technologies can be used to detect the nearby presence
of an object. A switch can be used for simple detection (near/not near),
whereas photoelectric, inductive, or ultrasonic techniques can provide
degrees of closeness or proximity information.
• Magnetic: Magnetic position sensors are noncontact sensors that detect
a magnetic field and adjust an existing electrical output based upon the
strength and direction of the field in proportion to the position.
• Inductive: An inductive position sensor interacts with a magnetic field
to create an electrical output whose strength is proportional to the mag-
netic field.
• Capacitive: A change in capacitance value is detected to determine the
position of a measured object.
• Optical: Optical position sensors transmit light from a source to a
receiver, or transmit and receive reflected light from the monitored object.
Information about object position is based upon changes in light charac-
teristics, including intensity, wavelength, polarization, or phase.
• Ultrasonic: Similar to optical sensors, an ultrasound source is used instead
of a light source.
• Hall Effect: This noncontact sensor detects variations in magnetic fields to
determine orientation, especially on a rotating shaft or device. This sensor
is commonly used for speed detection.
Position sensor units of measurement include angle (degrees), direction,

speed (rpm), velocity, position, range, and limits.
2.4.7 Vibration
Monitoring machinery vibration is an essential element of detecting failures
and predicting future component failures. Vibration often stems from mis-
alignment, imbalance, and wear on engines, motors, couplings, pumps, and
other machinery. Measurement is accomplished mostly through the use of
strain gauges and piezoelectric accelerometers that can sense back-and-forth
motion. Vibration can be considered as being of two types:
• Axial: Vibration that is in line with the thrust projected from the
machinery.
• Radial: Vibration that radiates perpendicular to the line of thrust.
Units to express vibration measurement can take several forms.5 When

considering the repetitive back-and-forth motion (frequency), this can be
expressed in cycles-per-second or Hertz (Hz). The amplitude between vibra-

tion peaks or displacement from a fixed position can generally be measured
in meter (m), centimeter (cm), millimeter (mm), or micrometer (µm) units.
The velocity achieved during the vibration cycle representing the changing
rate of displacement can be expressed as distance per unit of time; or cm/S,
mm/mS, or mm/µS. Acceleration may be defined as the changing rate of vel-
ocity per unit of time and can be expressed as m/S2, cm/S2, or mm/S2. Sound
levels resulting from vibration can generally be expressed in decibels (dB) or
watts per square meter (W/m2).
2.4.8 Gases
Exposure to hazardous gases and deficiencies in essential gases in confined
spaces onboard ships has gained significant attention in the press and by
regulatory authorities.6,7,8 Hazard sources include leaking gases and liquids,
cleaning fluids, combustion (e.g., welding), dust concentrations and decom-
posing organic matter. Among the most significant gases monitored are:
• Oxygen (O2): Depletion of oxygen, especially in cargo holds, has been the
cause of many deaths.
• Carbon Monoxide (CO): Caused by the accumulation of exhaust gases
from combustion.
• Hydrogen Sulfide (H2S): Usually caused by leakage of blackwater tanks.
• Lower Explosive Limit (LEL): Combustible gases such as methane and
natural gas.
Many permanent installation and portable gas detectors are available for
detecting these and other gas hazards onboard ships. Multiple gases are often
detectable using the same device.
2.4.9 Fire
Fires on board ships are one of the most dreaded occurrences, costing many
lives and billions of dollars in damage and losses annually. Many nations have
adopted IMO regulations under the Safety of Life at Sea (SOLAS) conven-
tion that specifically address shipboard fire detection capabilities.9 All ships
are subject to fire, but car-carrying vessels, in particular, have experienced
even greater exposure as a result of carrying vehicles that contain not only
combustible fuel but also lithium-ion batteries that are subject to seemingly
spontaneous combustion.10 Three different types of sensors are available that
are capable of detecting the characteristics of fire. These include:
• Smoke: Operation is based upon the detection of particulates caused by

smoldering materials.
• Flame: Detects ultraviolet and infrared radiation emitted by flames.

• Heat: Measures temperature above a threshold where the presence of fire
is likely to exist.
Smoke is the most likely killer of people as toxic fumes may be emitted
over a period of hours before the presence of flame or heat can be detected.
Smoke detectors use a photo-electric sensor to visually detect particulates
or a radioactive isotope as an ionization source to measure electrical con-
ductivity between particles. Flame detectors use optical sensors to rapidly
detect ultraviolet and/or infrared emissions associated with fire. Heat detec-
tors are temperature sensors calibrated to a threshold that corresponds to a
probability of fire. They are particularly useful in environments laden with
dust and where high moisture levels exist. Many fire detectors contain two
or more combinations of sensors. In addition to spaces where fire detection
equipment is mandatory, car-carrying vessels are also adding networks of fire
detectors throughout their cargo decks to detect fires with greater specificity
as to location to further speed response times and fire suppression efforts.
2.4.10 Microphones
Microphones installed in the engine room and in other locations that are
usually unoccupied can provide great insight into adverse situations as they
develop. Easily detected using a simple live microphone connected to a remote
speaker are vibrations, bearing squeals, clanking noises and other audible
clues to actual or pending equipment malfunctions with which engineering
staff are already familiar but may otherwise go unnoticed in the absence of
staff within these spaces.
Sound monitoring devices are also available that can continuously monitor
and record vibration, sound and noise levels, and air overpressure. These
devices may be equipped with extended dynamic range microphones that
can detect sounds outside the human range of hearing, are water-resistant
and ruggedized to withstand harsh conditions. Monitoring can be performed
continuously for days and weeks at a time, and thousands of events may
be recorded when equipped with sufficient memory. Some such devices are
standalone units equipped with internal batteries, while others are operated
under the ship’s power with battery backup that is network-connected to
supplement other alarms and system capabilities.
2.4.11 Video Cameras

The use of video cameras onboard ships is dramatically increasing due to
security requirements and to help make up for reduced crew sizes in main-
taining surveillance of deck areas and enclosed spaces. Much like micro-
phones, video cameras can offer visual clues as to events when they occur.
Some cameras have extended bandwidth capabilities to cover both the visible
and infrared spectrums, or two cameras with complementary capabilities can
be installed. This can be advantageous for identifying machinery that is run-
ning hot. Video monitoring can be performed continuously and recorded for
days or weeks at a time over many events, usually at a location remote from
the cameras connected via a network.
2.5 Sensor Interfaces, Controls and Panels

All sensors must be connected to interfaces and data buses that distribute
sensor data to their destinations to be processed and initiate actions appro-
priate to sensor measurements. These include controls, alarms, panels and
other devices such as computers and controllers associated with all sensed
phenomena such as point and continuous levels and flows, leak detection
and overfills, hazardous and nonhazardous atmospheres associated with ship
hull and superstructure, critical system infrastructure, passenger and crew
monitoring and cargo operations.
At these endpoints and in between there exist system components critical
to efficient and reliable operation of sensor networks. Beginning from the
sensors themselves, these components include:
• Cables and Connectivity: The channels and conduits through which sen-
sor data is communicated.
• Power Supplies: Delivery of electricity to provide power to sensors and
sensor networks.
• Signal Conditioning: Amplification, filtering, level changing and digitiza-
tion of sensor signals.
• Displays and Computers: Destination for sensor data processing, alarm
generation and user interfaces.
Each of these topics is discussed in the paragraphs that follow.
2.5.1 Cables and Connectivity

Electrical and fiber optic signal paths in the form of wires, cables and other
conduits form a critical part of ship infrastructure through which individual
sensors communicate the signals and data they acquire. Some sensors are indi-
vidually wired through interfaces or directly to receivers, signal conditioning
and data processing elements; while others are connected to common data
buses that share many sensors and exchange large volumes of many types of
data distributed to converters, other processors and displays. Wireless and
Internet of Things (IoT) sensors designed specifically for the widespread shar-
ing of data are also becoming more common.
Several onboard networks comprise digital communication buses for data

transfer between sensors and onboard computers. Internal ship communi-
cation is accomplished using one or more data buses that include local area
network (LAN), RS232/ RS485, NMEA- 0183/IEC 11623, NMEA- 2000/
IEC 61162-3, controller area network (CAN), wide area network (WAN)
and Bluetooth. These and some other standards coalesce under the IMO e-
navigation initiative, where maritime services transition from conventional
transmission methods to contemporary digital technologies by providing
and exchanging relevant information and data in a harmonized and unified
format.
A thorough conversation on this topic is warranted when discussing engin-
eering sensors. However, since the scope of this topic also covers navigation,
passenger and crew, and cargo sensors; this topic is covered separately in
Chapter 6 under Sensor Data Communications.
2.5.2 Power Supplies

Provisioning of electrical power to sensors is generally accomplished using
centralized communication centers located in one or more standard 19-inch
racks within which power supplies reside connected via interfaces that pro-
vide signal conditioning and power to the sensors. Power inputs are obtained
from the ship’s power at 115 V/230 Vac with power outputs provided to sen-
sors at 12–28 Vdc at current levels sufficient to meet requirements depending
upon the number of sensors connected to the system. Voltage and current
requirements vary according to individual sensor requirements. However,
many sensor interfaces are standardized, greatly simplifying this problem.
2.5.3 Signal Conditioning

Electrical signals present at the sensors themselves generally need to be
adapted to coincide with the signal levels and formats that are compatible
with the receivers at the processors, computers, and displays that use and act
upon sensor data. Sensor signals are often too weak to be transmitted across
long distances, can also be noisy and must be filtered to be useable. Sensor
data can originate from several different forms that must be reformatted into
digital word formats.
An example of signal conditioning is in the case where sensors that are
based upon resistance can operate across ranges from 100 ohms (Ω) to
500 kΩ or more, depending on the type of sensor used. These are very com-
mon sensors that measure temperature, strain, humidity, fluid flow, force,
pressure, and acceleration. Resistive measurements are often converted into
voltage levels by bridge circuits that use a pair of voltage dividers, with one
containing known resistance values and the other containing the resistive
sensor. The circuit output is taken as the difference in voltage potential
between the two dividers. When the balance between the dividers is the
same there is no voltage difference. However, when the resistance of the
sensor changes, the output voltage changes as a function of the measured
value. In most cases, the magnitude of the voltage is small and requires amp-
lification. This analog voltage needs to be converted into digital format to
be useful. Bridge circuits generally feature low power consumption in the
microamp (µA) to milliamp (mA) range and are often integrated directly
into the sensor.
Another example is capacitance sensors, where capacitance values are con-
verted directly into digital words at the sensor using an integrated circuit
(IC) with correction and compensation for offset, sensitivity and drift made
according to the specific characteristics of the sensor itself. This applies to
single capacitive sensors where both terminals are accessible and differential
capacitive sensors. Capacitance sensors also feature low power consumption
in the microamp (µA) to milliamp (mA) range.
In many cases, the signals generated by the sensor itself and available at its
output terminals are not directly compatible with circuit path characteristics
to transmit sensor data, nor are they at the proper levels or in the correct
format. It is, therefore, necessary to further process sensor signals for them to
be useful. Several more common types of sensor signal processing techniques
are listed in the following paragraphs:11
• Amplification: When sensor signal strength is below a threshold where

a receiver cannot directly acquire data, it is necessary to boost or amp-
lify the signal power to a level and range that is compatible with receiver
input requirements. This applies to analog signals where the sensor
output signal range should match the full range of analog to digital (A/D)
converter capabilities, and to digital signals to ensure appropriate logic
levels are attained. This is also important to overcome signal loss where
long transmission paths exist between the sensor and receiver, and to help
increase the signal-to-noise ratio such that environmental noise is mini-
mized. Impedance matching between in-line system components is neces-
sary to minimize signal loss, and sufficient output drive current must be
available to ensure the signal is usable. Thermocouples and strain gauges
typically require amplification to be effective.
• Attenuation: The opposite of amplification, attenuation entails reducing
or limiting sensor output signal strength for analog sensors that may
saturate or overload receiver input capabilities. This specifically applies
when the sensor output signal is stronger than the range required for A/
D conversion.
• Filtering: Electronic filtering eliminates environmental and other noise
from sensor measurement data. This can include 50/60/400 Hz signals
from local power sources and other frequencies radiated from equipment
nearby the sensor and along the signal transmission path. With respect to
the operating bandwidth of the receiver, low-pass filters reduce or elim-

inate signals at the low end of the frequency spectrum from local power
sources, while high-pass filters eliminate signals that exist above the high-
est frequency that is necessary to be coded at a given sampling rate to fully
reconstruct the signal. Low-pass and high-pass filters are used together to
eliminate signals within a specific band of frequencies.
• Isolation: Electrical isolation of sensor signals may be required in cases
where signal parameters are significantly outside the range of the receiver,
when voltage or current spikes can be introduced by equipment switch-
ing and other circuits, or where stray AC or DC signals may leak into
a sensor signal path that may cause damage to equipment. This can be
accomplished through the use of capacitors, transformers, optoisolators,
or optocouplers.
• Excitation: This involves using supplemental voltage or current sources
to facilitate the proper operation of active sensors that may include resist-
ance temperature detectors, bridge sensors and Hall elements.
• Linearization: Sensors that produce nonlinear signals with respect to the
physical measurement often require a process to transform their output
into a linear signal. This may be accomplished using either hardware or
software.
• Compensation: Sensor compensation involves calibrating a signal with
respect to offset and span correction over a specific measurement range.
This may be accomplished using either hardware or software.
Other types of signal processing can be performed to ensure optimal

performance and flexibility by manipulating a signal to prepare it for the
next stage of processing. The result must be in compliance with display and
computer interface requirements at the next level of integration for users of
these data.
2.5.4 Displays and Computers

Once sensor signals have been generated and processed, they are routed to
the location(s) where they are recorded, used to provide notifications and
alarms; or undergo additional computer processing as needed to operate and
maintain the vessel. Display of sensor data may be accomplished using a
simple analog gauge or digital numeric display, or integrated into complex,
multi-sensor displays to monitor a combination of sensors. Such displays
are often placed in one or more standard 19-inch racks that include one or
more power supply(s), level converters, and signal processors, central pro-
cessing units, a multi-function keyboard, and a display unit providing meas-
urement and alarm data. This may reside at a single point location, such as
a control room, or be distributed throughout the ship based upon areas of
responsibility.
2.6 Sensor Relationships and Engineering System

Automation
With increasing ship digitalization and automation, the role of sensors is
ever-growing as they form the heart of how onboard systems view their
environment. In the engineering department alone, automation has been
introduced to reduce costs, improve operations and reduce the occurrence of
human error and therefore accidents. Areas served represent a wide variety
of integrated control systems including monitoring alarms and implementing
auxiliary control systems, power management, propulsion, ballast systems,
cargo and heating, ventilation and air conditioning. Smart sensors form the
core of what is making automation possible.
References
1 AllICData. What Are the Static and Dynamic Characteristics of the Sensor?
allicdata.com. Updated 21 November 2021. www.allicdata.com/news/sensor/
what-are-the-static-and-dynamic-characteristics-of-the-sensor.html
2 Tom Kenny. Sensor Technology Handbook. Chapter 1, Sensor Fundamentals.
Elsevier, Oxford, UK. 2005. Edited by Wilson, Jon S., pp. 2–5.
3 Sudripto Khasnabis. Different Technologies to Measure Hull Stresses in
Ships. Marine Insight, Naval Architecture. 16 September 2019. www.marine
insight.com/naval-architecture/different-technologies-to-measure-hull-stres
ses-in-ships/
4 Y. Takaoka, K. Nihei, P. Vargas, P. Aalberts, and M.L. Kaminski. Application
of Fatigue Damage Sensors in the Monitas System. Offshore Technology
Conference (OTC), Houston, TX, USA. 30 April 2010. https://doi.org/10.4043/
20870-MS
5 ByJu’s.com. Unit of Vibration. https://byjus.com/physics/unit-of-vibration/
#vibration-units
6 Maritime Executive. Two Killed in Gas Leak Aboard Chinese Bulker. 23 April
2018. https://maritime-executive.com/article/two-killed-in-gas-leak-aboard-
chinese-bulker
7 Maritime Executive. Police Inspect Sydney Harbor Cruise Boats After Passenger
Fatality. 7 February 7, 2019. www.maritime-executive.com/article/police-insp
ect-sydney-harbor-cruise-boats-after-passenger-fatality
8 Cargo and Cargo Hold Ventilation. INTERCARGO, the Standard Club and
DNV- GL. January 2021. www.dnv.com/maritime/publications/Cargo-and-
Cargo-Hold-ventilation-guidance-download.html
9 IMO. Part IV. SOLAS Chapter II- 2. Construction –Fire Protection, Fire
Detection and Fire Extinction. International Maritime Organization. London.
10 Valdes-Dapena, Peter. Burned Ship Carrying Luxury Cars Has Now Sunk.
CNN Business. 2 March 2022. www.cnn.com/2022/03/02/business/felicity-ace-
car-ship-sunk/index.html
11 David Ashlock and Anjelica Warren. The Engineer’s Guide to Signal
Conditioning. National Instruments, 2015. https://download.ni.com/evaluat
ion/signal_conditioning/20712_Benefits_of_Integrated_SC_WP_HL.pdf
Chapter 3
Navigation Sensors
Many sensors are dedicated to different aspects of the safe and efficient move-
ment of ships through perilous waters and the avoidance of other vessels and
hazards to navigation while en route to their destinations. Biological sensors
provide human vision, hearing, touch, smell and other abilities that have
formed the core ship navigation capabilities for millennia and continue to
do so, with the human brain the ultimate destination for sensor fusion and
interpretation. However, the inventions of mechanical and electronic sensors
have vastly expanded the range, scope, resolution and accuracy of the human
senses to better assist mariners in performing essential pilotage, dead reckon-
ing, celestial and other navigation techniques to locate position, determine
time, steer a proper and safe course and continuously monitor the voyage
through completion. Examples of invention include the ancient astrolabe,
back staff, chip log, leaded line and sextant right up to modern electronic
sensors such as Radar, Sonar, Global Navigation Satellite Systems (GNSS),
autopilots, and weather instruments so vital to planning and executing a voy-
age. This has also led to secondary uses of time-delayed sensor data found in
nautical charts, Electronic Chart Display and Information Systems (ECDIS),
Notices to Mariners and Coast Pilots.1,2,3,4 The data these products contain
have been compiled in the past from many diverse sensor technologies and
multiple sensors of the same kind including single-beam and multi-beam
Sonars, Lidar, aerial and satellite imagery.
Marine navigation has been described as a blend of both science and art,
and that science can be taught but art must be learned from experience.5
Ship sensors illustrate how science can improve situational awareness and
greatly enhance navigation safety. The principles of what ship sensors can
do and how they function have been the basis for many courses of study and
practical demonstrations. It is also the focus of research in machine learn-
ing that attempts to replicate, improve and accelerate human methods for
viewing and interpreting sensor data. However, the art of marine naviga-
tion stems from prudent discernment of sensor limitations considering the
Doi: 10.1201/9781003377900-3
Navigation Sensors 29
types of vessels on which mariners serve and encounter along their voyages,
and the environment through which they must pass. Such knowledge takes
years to acquire and master. It is this art of navigation that the developers
of deep-learning Artificial Intelligence (AI) technologies are attempting to
master through the fusion of multiple sources of sensor data combined with
vast resources portraying vessel and human behavior.
Sensor technologies continue to adapt to new navigation applications, and
the evolution of advanced sensor capabilities promotes the invention of new
and enhanced methods to navigate. Historical examples include the replace-
ment of the leaded line with Sonar, bells with high-power air and electric
horns, and speed logs and Long-Range Radio Navigation (LORAN) with
GNSS. The use of electronic sensors in vessel navigation is expanding across
all maritime frontiers, whether they advise crewmembers on conventionally
staffed ships or guide and control all navigation functions on autonomous
vehicles.
This chapter proceeds from the perspective of how ship navigation is
achieved with the benefit of sensors rather than focusing solely on the sen-
sors themselves, although details on sensors are indispensable to their proper
understanding. A review of the different types of ship navigation establishes
a historical basis for how these inventions have been and continue to be
adopted. This includes a discussion of methods developed over the centuries
to perform navigation, emphasizing the contributions of sensors toward
extending the human senses to accomplish these tasks. Finally, an overview
is presented of the many sensors commonly found on ships, along with a
preview of future navigation sensors and sensor technologies looming on the
horizon.
3.1 Sensor Use in Navigation

An integral part of the navigation process is the ability to avoid collision and
allision by communicating the meaning behind actions taken in response to
sensor indications from Radar and Automatic Radar Plotting Aid (ARPA),
Sonar, Automatic Identification System (AIS), radio, audible and visual sig-
naling to and from nearby vessels and shore-based operation and control
centers. Underlying these principles is the need for precise knowledge of the
vessel’s relative position with respect to other vessels, landmarks and aids to
navigation (AtoN), land masses and sea bed topography along the route to
the intended destination.
Essential to any discussion about navigation is the requirement to main-
tain a proper lookout by bridge watchstanders based upon the International
Maritime Organization (IMO) Convention on the International Regulations
for Preventing Collisions at Sea (COLREGS) rule 5 that states:6
30 Navigation Sensors
Every vessel shall at all times maintain a proper look-out by sight and
hearing as well as by all available means appropriate in the prevailing cir-
cumstances and conditions so as to make a full appraisal of the situation
and of the risk of collision.
In practice, the lookout task has traditionally been performed by one or more
watchstanders posted at all times on the bridge and, weather permitting,
stationed forward in restricted visibility conditions. They are tasked with
reporting on lights, vessels and small craft, landmarks, large floating objects
and marine mammals, indications of shoal waters, hazards to navigation
and fog signals likely to herald the risk of collision. Efforts to further vessel
autonomy find increasing assistance being rendered to bridge watchstanders
by independent electronic bridge aids to monitor sensor data, detect potential
hazards and suggest possible courses of action to be taken. Such aids satisfy
the need for sight and hearing and all appropriate means to fully assess the
situation.
Information obtained from navigation sensors are interconnected using
multiple workstations, providing overlapping and redundant capabilities
through an Integrated Navigation System (INS). Such systems are a con-
venient means for seafarers to prioritize and display navigation informa-
tion in an organized manner and check the validity, consistency, latency and
integrity of the data. The products of these displays are used to assist in
route planning and monitoring, collision avoidance, alarm monitoring and
to enhance overall navigation situational awareness.
Since approximately 1995 navigation sensors have witnessed the devel-
opment and introduction of new instruments and technologies as diverse
as Digital Select Calling (DSC) in Very High Frequency (VHF) and High
Frequency (HF) radio, Navigational Text Messages (NAVTEX), enhanced
Long- Range Radio Navigation (eLORAN), Forward- looking Navigation
Sonar (FLS), the GNSS, Global Maritime Distress Safety System (GMDSS),
AIS and Satellite AIS (S-AIS), digital Radar/ARPA, multibeam Sonar, broad-
band satellite communications, ECDIS and the Internet.7 The rate of improve-
ment in existing sensors, the invention of new sensor technologies and their
augmentation to achieve new functionality through the use of AI continues
to increase at an accelerated pace.
3.2 Overview of Navigation Sensor Functionality

A comprehensive list of navigation-related sensors and sensor systems com-
monly found on commercial and special-purpose vessels, along with a sum-
mary of associated signal descriptions, is included in Table 3.1. However,
standards do not exist to enable seafarers trained in the equipment of one
manufacturer’s product to be proficient in the same type of equipment pro-
duced by other manufacturers. The bridges of many naval, coast guard
Table 3.1 Example Navigation Sensor Functions
Automatic Identification System (AIS) Maritime Mobile Service Identity

(MMSI), Navigation Status, Rate of
Turn, Position, Speed Over Ground,
Course Over Ground, True Heading,
True Bearing, UTC seconds, IMO
Identification, Radio Call Sign, Vessel
Name, Cargo Type, Ship Dimensions,
Type of Positioning System, Location
of Positioning System Antenna,
Ship’s Draught, Destination, ETA…
AutoPilot* Heading, Direction. Data
sources: GNSS. Rate Gyro/INS,
Rudder Indicator, Speed Log…
Audio (Sound) Decibels, Frequency, Direction.
Camera – Visible Video, Range, Direction.
Camera – Infrared Video, Range, Direction.
Compass – Gyro True Heading, Bearing.
Compass – Magnetic Magnetic Heading, Bearing.
Depth Sounder Depth, Range, Video…
Electronic Chart Display Information Coastline, Own Ship’s Safety Contour,
System (ECDIS)* Isolated Underwater Dangers,
Isolated Dangers, Scale, Range,
North Arrow, Units of Depth and
Height, Display Mode, Display Base.
Drying Line, Buoys, Beacons, other
Aids to Navigation, Fixed Structures,
Fairways, Channels, Visual and Radar
Conspicuous Features, Prohibited
and Restricted Areas, Chart Scale
Boundaries, Cautionary Notes, Ship
and Ferry routes, Archipelagic Sea
Lanes, Spot Soundings, Submarine
Cables and Pipelines, ENC Edition
Date, Chart Update Number,
Magnetic Variation, Graticule, Place
Names….
Data sources: Electronic Navigation
Chart (ENC) database, GNSS, AIS,
Radar, ARPA…
Emergency Position Indicating Radio Vessel ID, Vessel Type, Vessel size,
beacon (EPIRB) GNSS coordinates, Contact
Information …
Forward-L ooking Sonar Depth, Range, Video…
Global Maritime Distress Safety Call Sign, MMSI, Selective Call
System (GMDSS) Number, Mobile Earth Station
Identification Number (MESIN),
Ship Earth Station Identify Number
(SESIN), 406 MHZ EPIRB Unique
Identification Number.
(Continued)
Table 3.1 (Continued)
Global Navigation Satellite System Positioning, Navigation and Timing

(GNSS) (PNT) data…
Heading Indicator Heading.
HF SSB Radio Audio, Imagery, Digital data, MMSI,
GPS data…
Laser Radar (LIDAR) Video, Range, Direction…
Microwave Radar Video, Range, Direction…
NAVTEX Navigational Warnings, Meteorological
Warnings, Ice Reports, Search
& Rescue, Pirate Warnings,
Meteorological Forecasts, Pilot
Service Messages, AIS Messages,
SATNAV Messages, Other Electronic
Navaid Messages and Navigational
Warnings.
Automatic Radar Plotting Aid (ARPA) Video, Range, Direction, Closest Point
of Approach, Time to Closest Point
of Approach, Alarms, Target Data…
Rudder Angle, Order, Response.
Search and Rescue Transponder X-b and Radar Beacon
(SART)
Satellite Imagery, Audio, MMSI, GPS data…
Side-S can Sonar Depth, Range, Video…
Speed Log Speed Through Water, Over Ground,
Total Log, Trip Log.
VHF-D SC Radio Audio, Imagery, MMSI, Digital data,
GPS data…
Voyage Data Recorder (VDR) Data from: GNSS, Speed Log, Gyro
Compass, Radar, AIS, ECDIS,
Bridge Audio, VHF, Echo Sounder,
Main Alarms, Hull Opening Status,
Hull Accelerations and Stresses,
Rudder, Engine, Propeller, Thrusters,
Anemometer, Weather Vane…
Weather Temperature, Barometric Pressure,
Humidity, Dew Point, Wind
Direction, Wind Speed…
*Note: The instrument itself is not a sensor but uses sensor data to perform its navigation-
related function.
and other military vessels also generally have the same navigation equip-
ment as civilian vessels and often comply with the IMO Safety of Life at Sea
Convention (SOLAS) requirements. It should be noted that military vessels
are not necessarily required to operate in the same manner as civilian vessels
in terms of compliance with COLREGS and SOLAS and often do not regu-
larly utilize equipment such as GMDSS and AIS.
3.3 Types of Navigation

Different types of navigation are used depending upon the physical environ-
ment in which the ship is found: harbor, harbor approaches, inland, coastal
and ocean environments. Common to all modern vessels is the use of satel-
lite navigation, which in practice almost universal reliance is unwisely made
for positioning information. All large ships and many smaller vessels also
rely upon Radar navigation for relative positioning information and col-
lision avoidance. Radio navigation has been transformed from its origins
as a means of obtaining positioning information via Omega, LORAN, and
direction-finding systems to become a key provider of inter-ship communi-
cation via AIS, in addition to retaining its original functions of exchanging
intentions between ships’ bridges to coordinate close passage, ship to shore
communications and for emergency use.
Piloting for navigating in restricted waters based on relative positioning
to geographic and hydrographic features continues to be of great import-
ance. Dead reckoning based upon course, speed, current, leeway, steering
error and other factors remains a vital component of position estimation.
Celestial navigation using lines of position to the moon, planets and stars is
seeing a resurgence in significance, especially in view of the many limitations
and frailties encountered with satellite navigation technologies. With sensors
playing increasingly vital roles across all types of navigation, overreliance
upon one or even several of these technologies without proper correlation
between their results can lead to reduced safety of navigation and even disas-
trous outcomes. The following paragraphs describe these types of navigation
and highlight the roles sensors play in each.
3.3.1 Satellite Navigation

Most conversations on satellites for ship navigation focus exclusively on
GNSS services. However, often taken for granted and overlooked is the indis-
pensable role played by weather satellites in providing the information neces-
sary for the planning, continuous monitoring and execution of voyages. The
following paragraphs attempt to offer comprehensive insight into all uses of
satellites to permit and facilitate safe navigation.
3.3.1.1 Global Navigation Satellite System (GNSS)
The GNSS is comprised of several constellations in Medium Earth Orbit

(MEO) positioned between 3,000 and 12,500 miles and Geostationary Earth
Orbit (GEO) positioned geostationary 22,236 miles above the Earth with
each constellation containing between 20 and 32 or more satellites devel-
oped by different countries to provide positioning, navigation and timing
(PNT) services to government, commercial and private users worldwide.8

They enable precise determination of location in three dimensions (latitude,
longitude and altitude) at any given time, although maritime use is generally
at sea level except for some inland navigation locations.
Satellite systems that form the GNSS have been developed and are oper-
ated by various countries across the world and include:
• Global Positioning System (GPS) –United States.

• Global Navigation Satellite System (GLONASS) –Russia.
• Galileo –European Union.
• BeiDou Navigation Satellite System –China.
• Indian Regional Navigation Satellite System (IRNSS) –India.
• Quasi-Zenith Satellite System (QZSS) –Japan.
GNSS operates primarily in the L-band (1–2 GHz) and S-band (2–4 GHz)
frequencies. The method used in the operation of GNSS includes the trans-
mission of signals from various satellites in each constellation, with each
signal containing a message with precise satellite location and current time
information. Position is determined through triangulation by measuring the
time it takes for signals to travel from multiple satellites to the ship’s onboard
receiver. Modern GNSS receivers often receive signals independently from
several different GNSS systems to improve reliability in the event problems
exist with one system, and to enhance accuracy.
Continuous PNT information is provided directly to the ship’s navigation,
communications and other onboard systems through constant communica-
tion with satellites to enable tracking of vessel movements and determine
velocity and direction. Projection of future positions based on the vessel’s
current trajectory helps to determine adherence to planned routing and to
detect and avoid potential hazards to navigation that may include other ves-
sels, shoals and insufficient depths, and potential conflicts with charted and
sensed objects. GNSS also provides onboard VHF and High Frequency (HF)
radio transmitters and satellite terminals with position data to be communi-
cated in the event of an emergency to help speed rescue efforts, and many sys-
tems offer features to mark waypoints that assist in man-overboard recovery
situations. Capabilities to provide messaging for emergency warning services
are being considered for the Galileo, IRNSS, and QZSS constellations.9 This
includes features for danger notification and subsequent instruction to escape
from disaster situations. Several commercially available communication
devices also locate positions using GNSS and then connect to other satellites
to send and receive messages.10
Many limitations exist in the implementation of GNSS technology that
reduces its usefulness and allows it to be exploited for nefarious purposes by
nation states and individuals. One of its main flaws is that it operates using
relatively weak signals from space that can easily be overridden by spoof-
ing that displays false position information to users and can lead vessels
to stray into shoal waters or directly into the hands of pirates and adver-
saries. These weak signals can also be interfered with through jamming
resulting in denial of service (DOS) attacks that render it useless. GNSS
effectiveness can also be limited due to signal path obstruction between
the satellite and antenna by ship structures, and multipath interference due
to reflected signals from the many metal surfaces on ships that can lead
to errors in position, speed, or velocity calculations. Accuracy can also be
affected by sunspots and space weather events that influence atmospheric
conditions, especially with the ionosphere. It can also be adversely affected
by severe weather conditions such as heavy precipitation, dense cloud cover
and sometimes even fog. Satellite geometry, where satellite configurations
are altered as parts of the constellation go out of service, can also adversely
affect accuracy.
An additional potential limitation of GNSS services has to do with the
ability of national operators to selectively degrade or dither signals to reduce
positioning accuracy to nonmilitary users. This was called “selective avail-
ability” for the United States GPS, where autonomous horizontal position-
ing accuracy was advertised to be no worse than 100 meters 95 percent of
the time. This function was turned off in 2000 and officially discontinued in
2007 after the U.S. DoD demonstrated the ability to selectively deny GPS
signals on a regional basis as needed in a military area of operations when
U.S. national security is threatened.11 Such capabilities are likely retained for
other nation-state controllers of GNSS satellite constellations.
It should be noted that Differential GPS (DGPS) services can be used where
further signal augmentation is necessary to achieve centimeter-level or sub-
meter-level accuracy. With one receiver established at a precisely known loca-
tion as a reference, another roaming receiver calculates its position based
on satellite signals and compares this location to the known location.12 The
difference is applied to the roaming GPS receiver in real-time in the field using
radio signals. Differential techniques may be used with other GNSS systems
as well.
3.3.1.2 Weather Satellites
Increasing bandwidth for ship satellite communications allows many real-

time weather services and long-term forecasts to be provided directly to ves-
sels at sea. This data is increasingly important for efficient route planning,
weather avoidance and better-informed voyage decision-making. Commonly
reported conditions provided by the World Meteorological Organization
(WMO) and National Meteorological & Hydrological Services (NMHSs)

worldwide include:
• Cloud cover • Atmospheric pressure

• Precipitation • Currents
• Temperature • Buoy observations
• Wind speed and direction • Sea ice concentration
• Sea surface temperature • Earth observations
• Wave height
Satellite providers of Maritime Safety Information (MSI) in terms of

meteorological forecasts and warnings and a variety of other services include
Inmarsat/ Viasat SafetyNET and Iridium SafetyCast, as well as relatively
lower-cost solutions requiring small antennas provided by SkyMate, Sirius
XM Marine Weather, and Baron WxWorx. (Trademarks and Service Marks
are properties of their respective service providers.)
Weather satellites generally operate within the same frequency bands
as communication satellites, including the L, S, C, X, Ku and Ka bands.
Exceptions exist where some weather satellites use frequencies in the
Extremely High Frequency (EHF) and millimeter-bands up to 100 GHz, and
analog Automatic Picture Transmission (APT) signals transmitted at frequen-
cies between 137 and 138 MHz.
Concern has been voiced on whether 5G global networks can interfere
with weather satellites in terms of their ability to sense water vapor, tempera-
ture and other atmospheric phenomena vital to weather forecasting.13 There
is not yet a consensus on this issue, and the full effects of 5G will be deter-
mined in the future once the inevitable expansion of this technology becomes
more prominent in the satellite realm.
3.3.2 Radar Navigation

Radar as a navigation tool is a specialized form of piloting that enables
navigators to measure distances and bearings to chart objects, Radar AtoN,
and other vessels. Radar provided game-changing technology in terms of
new navigation capabilities for position identification and collision avoid-
ance, especially under low visibility and night time conditions, qualifying
this navigation technique as a unique and separate discipline complemen-
tary to other types of navigation. However, the complexities involved in
its operation required extensive training to master its use. The dangers of
inadequate training resulting in improper navigation were illustrated by
a Radar-assisted collision in 1956 when the third mate had misread his
screen and inadvertently steered Stockholm straight into Andrea Doria.
Over 1,600 lives were at risk, 51 lives were lost and Andrea Doria was
sunk.14
Radar system designs have changed over the years making them more
accurate with greater range and resolution, providing more information than
in previous generations, and easier to use and interpret. Also being intro-
duced are tools such as Doppler technology that instantly shows whether a
target is approaching or moving away, and ARPA that can create tracks and
calculate the target course, speed and closest point of approach. New deriva-
tives of Radar technology are also being introduced in the form of Synthetic
Aperture Radar (SAR), millimeter Radar (mmRadar), and Light Imaging
Detection and Ranging (Lidar) that promise new levels of situational aware-
ness for not only traditionally staffed ships but also unmanned and autono-
mous vehicles. The following paragraphs describe some of the key technical
aspects of Radar-based sensors and their use in modern vessel navigation.
3.3.2.1 First Analog, Then Digital Radar
Early Radar systems produced high-power continuous wave signals that fea-
ture combinations of amplitude modulation (AM), frequency modulation
(FM), and pulse-width modulation (PWM) characteristics to detect and locate
targets. Distance is measured by determining the time required for the signal
to travel from the ship’s transmitter to a target and then return to the ship in
terms of the amplitude of the reflected signal being proportional to distance
with variation in frequency to distinguish between targets. Pulse width that
varies with reflected signal strength is used to measure target speed.
The range and resolution of the signal are dependent in part on the trans-
mission frequency, with lower frequencies generally resulting in higher range
and lower resolution than higher frequency transmissions. Radar signals are
highly directional, thereby enabling bearing measurements based on the dir-
ection from which they were transmitted. Signal pulse width helps to deter-
mine resolution, and the higher the frequency at which the pulses repeat,
the better it can discern faster-moving targets. Phase shift and Doppler shift
characteristics detected in the returned signal also help to determine target
speed and direction.
Analog Radar systems were prevalent in early maritime applications and
are generally being phased out in favor of modern digital technology that uses
much less power at reduced radiation levels to achieve greater performance.
The heart of analog and early digital Radar systems is a klystron vacuum
tube used as a signal amplifier. The processing of returning analog signals
reflected off-targets is accomplished by converting them into electronic sig-
nals compatible with filters and mixers, whereby various signal characteris-
tics are examined to reveal target distance, bearing and speed. Results are
displayed on an electronic screen using symbols and characteristics appro-
priate for the target.
Digital Radar technology can generate complex signals and transmit mul-
tiple pulse widths where the returned signals are sampled, converted to digital
signals and analyzed using a wide array of digital signal processing techniques
that provide far greater control and manipulation of data than analog tech-
niques. This results in greater accuracy and resolution than analog systems
and new insight not available using analog signal processing to better detect
and classify moving targets and to distinguish them from sea clutter, atmos-
pheric effects and stationary backgrounds. Modern digital Radar systems are
fully solid state, eliminating the need for a klystron vacuum tube amplifier.
They are also less susceptible to interference due to better available filter-
ing techniques. Digital Radar technology can provide features not available
using analog technology, including SAR that can deliver higher spatial target
resolution than conventional digital Radar systems.15,16 Additional function-
ality can also be provided whereby, with each sweep of the Radar antenna,
dual progressive scan transmissions may be transmitted and processed to
simultaneously display two separate Radar ranges that can be manipulated
separately. Additional capability is also available to overlay information
from other systems, such as AIS, onto the Radar display, further enhancing
situational awareness.
3.3.2.2 Radar Operating Frequency Trade-o ffs
Many early marine Radar systems operated in the L-band (500 MHz–1.5
GHz) frequency range that provided a good mix of range and resolution con-
sidering the technological state of the art of the time. Longer wavelengths are
less susceptible to attenuation due to precipitation, and wider beam widths
cover a larger area than progressively shorter wavelengths found at higher
frequencies. As technology progressed, successively higher transmission fre-
quencies were adopted offering greater resolution at the cost of lesser range.
However, the advent of digital Radars helped to extend range as digital signal
processing techniques improved. L-band Radars provided the foundation for
the development of Radar-assisted navigation and collision avoidance tech-
niques currently in existence.
Shorter wavelengths and higher operating frequencies have established
S-band (2–4 GHz) Radars as the workhorse of the maritime industry for
decades. Increasingly lower costs combined with long range on the order of
up to 100 miles, good resolution and a robust ability to penetrate precipita-
tion and other weather events provide a reliable sensor system upon which
regulatory advancements were established to enhance the safety of naviga-
tion worldwide. These are the first systems where automated tools for target
tracking and positioning were introduced.
Today’s state-of-the-art X-band (8–12 GHz) Radar systems operating at
shorter wavelengths and higher frequencies than S-band and L-band systems
provide excellent short-range capability. Combined with very high resolution
and good weather penetration, their use is ideal for waterway and port close-
in maneuvering and for search and rescue. They feature many advanced
capabilities including target tracking, identification and future position pre-
diction made possible through enhanced signal processing and sensor fusion
techniques. The simultaneous use of the combination of S-band and X-band
Radars on the bridge to provide complementary perspectives of short and
long-distance vessel and landmark characteristics can achieve greater situ-
ational awareness than was previously possible. X-band Radar systems can
display Search and Rescue Radio Transponder (SART) positions transmitted
from vessels in distress.
The most advanced short-range Radar systems available today operate
in the L-band (24–27 GHz) and have even higher resolution than S-band
and X-band systems. They are ideal for use in densely populated marine
environments, target identification, port and lock maneuvering and collision
avoidance. Due to their short wavelengths inherent to L-band signals, small
targets, buoys and other AtoN are easily detected and identified.
3.3.2.3 Radar-L ike Devices
New inventions that include mmRadar and Lidar are being developed for use
by Unmanned Surface Vehicles (USVs) to enhance precise close-in maneuver-
ing capabilities. These technologies are likely to also be adopted for use in
navigation by conventionally crewed vessels in the future as their utility and
benefits are more fully developed.
Operating primarily in the W-band (75–111 GHz), mmRadar has been
used for a variety of military purposes to provide terminal guidance for
missile targeting as an integral part of aircraft and ground vehicle fire con-
trol systems and to assist in aiming automated guns in ship defenses against
incoming missiles. While recognizing limitations associated with their short
wavelength operating frequencies, there are many opportunities for mmRa-
dar use on civilian ships and vehicles. These include collision avoidance with
other vessels and allision avoidance with objects, especially in port navi-
gation, for real-time tracking of small boats in close proximity, for human
in-the-water search and rescue, and ice navigation. The ELVA-1 mmRadar
system for shipboard use is currently being marketed as capable of elimin-
ating close-in blind spots and having a useful range of between 0 and 600 m.17
Lidar utilizes lasers that operate in the higher range of infrared light fre-
quencies up to 400 THz. Its uses mirror those of mmRadar except with
higher resolution due to the shorter wavelengths involved. Experiments have
shown that ship position and heading angle were estimated using Lidar with
higher accuracy than GPS, QZSS and inertial navigation, with a position esti-
mation of approximately 1.2 m.18 The Ouster OS2 Lidar system operates at a
frequency of 345 THz (865 nm) and detects 10% reflective targets at 200 m
with a maximum range beyond 400 m.19
3.3.2.4 Piloting by Radar
Piloting by Radar is the act of navigating a vessel through restricted waters

using electronic eyes that can see relative positions only from the perspec-
tive provided by the interpretation of radio signals reflected from the local
environment. The same practices that apply to piloting by eyesight also
apply to Radar piloting. Still, their implementation requires learned skills
and thought processes demanding intimate knowledge of Radar capabilities
and limitations. Of particular interest are Radar AtoN, which should be
specifically noted in the voyage plan. Terrain features, landmarks and AtoN
that are easily verified visually in good weather can be distorted or omit-
ted entirely from a Radar screen. For example, at long distances, low-sand
beaches and shoals that are easily detected by the eyes of a mariner may not
be detected at all by Radar unless covered by vegetation.20 Structures on
low sand beaches may appear to be other ships. Shadows in Radar returns
caused by land features can significantly distort shoreline configurations and
hide landmarks prominent to the eyes. Indirect or false echos caused by
reflections from vessel structures may also appear on the screen and cause
confusion. Other phenomena that may be encountered include side-lobe
effects that cause circles or a series of echos around targets, multiple trace
echos received from targets beyond the Radar range setting, and interference
from other ship Radar systems and electronic emission sources that are in
close proximity.
3.3.2.5 Radar Collision Avoidance
Whereas a major premise of the previous paragraph is that the relative

positions of objects present on the Radar screen are fixed, the relative posi-
tions of most moving targets are constantly changing and require extreme
due diligence. The monitoring of safe passage and evaluation of the risk of
collision is covered under COLREGS rule 7, Risk of Collision, which states
in part:21
proper use shall be made of Radar equipment if fitted and operational,

including long-range scanning to obtain early warning of risk of colli-
sion and Radar plotting or equivalent systematic observation of detected
objects.
This is a requirement that Radar shall be used under all visibility conditions,
day and night. The rule continues with,
assumptions shall not be made on the basis of scanty information, espe-

cially scanty Radar information.
This part of the rule specifically acknowledges the limitations of Radar that
make it unable to provide a complete or even an adequate representation
of events. All decision-making should consider Radar as only one source of
information that must be verified through other means such as visual contact,
AIS, and bridge-to-bridge exchange of intentions.
3.3.2.6 Radar Beacons
The use of a Radar beacon (Racon) to enhance the presence of physical AtoN
is accomplished by adding a transmitter that, when triggered by a Radar
signal, will transmit a coded reply consisting of a series of dots and dashes to
the interrogating X-band Radar. This reply appears on the Radar display in a
line emanating radially from just beyond the echo of the AtoN. Racons may
be used on both laterally significant and non-laterally significant aids alike;
the Racon signal itself is for identification purposes only and therefore carries
no lateral significance. They may also be used as bridge marks to mark the
best point of passage.
3.3.3 Piloting
The term “piloting” refers to the navigation of a vessel through restricted
waters. This involves one or more properly credentialled crewmembers, and
sometimes a pilot knowledgeable of local conditions, present on the bridge
performing the required tasks to achieve a successful transit. Integral to this
process is proficient knowledge of sensor operation, interpreting and applying
sensor data, and the real-time ability to combine and compare clues obtained
through visual monitoring of the physical environment with multiple sensor
readings to achieve and maintain comprehensive situational awareness.
The vast majority of sensors available on the bridge can assist in pilot-
ing navigation. Radar as a significant contributor to effective piloting has
already been discussed in the previous paragraph. Other major contributing
sensor sources include time-delayed data in nautical charts and ECDIS, and
real-time data sources, including different types of Sonars, compass heading
and bearing indicators, two-way audio hailers and even weather instruments.
Of course, the human eyes, ears and other senses combined with seafarer
training, experience and judgment are indispensable to successful piloting
navigation. Unmanned and autonomous vehicles have additional sensors,
supplemented by powerful computers containing machine learning and deep-
learning AI-based firmware and software that attempt to replicate seafarer
presence on board. Despite industry assertions to the contrary, as of 2024,
the practical realization of such technologies is still years away into the next
decade for general use and is only suitable now and in the immediate future
for demonstration projects.
3.3.3.1 Historical Sensor Data
Piloting begins before getting underway with a thorough review of historical

sensor data found in nautical charts, ECDIS, Light Lists, Notices to Mariners,
Coast Pilots, Sailing Directions, tide charts, and other documents (both phys-
ical and electronic) of the correct versions, scales and resolutions appropriate
to planning the voyage. These data have been obtained from a wide variety
of sensors at some time in the past, and may or may not reflect actual condi-
tions that may exist during the voyage but are representative of what might
be expected at various points along the route. These data sources indicate
the types and positions of AtoN, landmarks and other notable terrain and
landscape features, in addition to hydrographic information including depth
soundings and contours. Source diagrams and zones of confidence consist
of graphic representations of the extent of hydrographic surveys included
within nautical charts and accompanying tables of related survey quality
categories. This information helps mariners assess hydrographic survey data
and the associated level of risk to navigate in a particular area based upon a
combination of the survey date, position accuracy, depth accuracy and sea-
floor coverage to detect objects on the seafloor.22
3.3.3.2 Visual Features and Landmarks
Once underway, the human senses supplemented with binoculars and other
physical and electronic aids such as infrared (IR) and low-light cameras
remain the standard for visual piloting navigation of conventionally staffed
ships. Of specific interest to help accomplish this task are lighthouses, water
towers, piers, terrain and any prominent objects or features displayed on
navigation charts and contained within ECDIS to be used as AtoN that sea-
farers may correlate with visual sightings by taking compass bearings while
en route. This includes critical buoys and lights whose characteristics may
be verified as corresponding to their Light List entries, especially at night.
The positioning of all visual AtoN should be cross-checked using Radar and
GNSS whenever possible.
In the case of unmanned and autonomous vehicles, visible, low-light and
infrared cameras are combined with microphones to help achieve situational
awareness by sight and hearing through 360 degrees. However, seeing and
hearing alone is insufficient without proper knowledge of what is being seen
and heard. Appropriately trained and credentialed mariners remotely con-
trolling unmanned vessels are constrained by and must rely entirely upon
the perspective provided by shipboard sensors regarding the field of view,
image and audio resolution and fidelity, signal path latency and other factors
for decision-making. For example, a mariner having a lifetime of experience
viewing the world through two eyes must adjust to a world where six eyes
are the norm, assuming six cameras with a 60-degree field of view evenly dis-
tributed through 360 degrees, where each camera approximates a mariner’s
normal field of view. Fewer cameras could be used, but not without some
loss of directionality and resolution, along with a distorted view to which the
mariner is unaccustomed.
3.3.4 Radio Navigation

The earliest forms of radio navigation included now obsolete systems
involving radio beacons, Decca, Omega, and Loran-C technologies. GNSS
eventually replaced these systems, providing global coverage with far greater
positioning accuracy. However, should GNSS services be denied or unavail-
able, few backup options exist to provide replacement PNT capabilities. The
radio components and infrastructure are no longer produced or in place for
long-range radio navigation except for time signal broadcasts over HF radio.
There still exist requirements for radio use, especially VHF but HF as
well, for navigation purposes. In addition, several new developments have
occurred regarding AIS capabilities to broadcast and display AtoN on Radar
and ECDIS displays. These and other subjects are covered in the paragraphs
that follow.
3.3.4.1 Radio Direction Finding (RDF)
Direction-finding has traditionally been accomplished using low frequency

(LF) radio signals between 200 and 415 kHz and middle frequency (MF)
signals between 550 and 1,500 kHz from radio beacons transmitting ampli-
tude modulation (AM) signals commonly used for broadcasting stations and
services. Dedicated maritime and aeronautical beacons transmitting Morse
code identification signals, as well as commercial broadcast stations situated
along coastal areas, would be received by the ship using a directional antenna
to identify the compass bearing to the transmitter, which was then be used to
guide the ship toward the station’s location. RDF bearings are subject to inac-
curacies due to quadrantal error caused by the influence of the local envir-
onment, coastal refraction due to differences in conductivity of the land and
water, polarization error due to ionospheric activity and reciprocal error that
may result in 180° ambiguity in the reading.23 Radio beacons were generally
phased out of use by the year 2000.
Direction finding is still used to locate the source of radio transmissions
on 121.5 MHz, marking the position of an activated Emergency Position-
Indicating Radio beacon (EPIRB). This signal is in addition to the satellite
signal. It is accomplished while close proximity to the beacon and, presum-
ably, lifeboats and survivors in the water near where an accident occurred
on the water.
3.3.4.2 Hyperbolic Navigation
The Omega, Decca, and Loran C radio navigation systems began in the
1940s and 1950s and were based on using hyperbolic navigation principles
that measured phase differences between different signals to generate lines
of position from which a navigator could determine the position on a chart.
Each system used different frequencies, with Omega operating on the Very
Low Frequency (VLF) portion of the radio spectrum at 10–14 kHz, Decca
operating on Low Frequency (LF) at 70 kHz and Loran-C operating on LF
and Medium Frequency (MF) between 1.9 and 10 MHz. Each band of fre-
quencies has its own unique signal propagation characteristics and limita-
tions, with lower frequencies generally better able to cover longer distances.
Omega accuracy was about 1,000 m at ranges of up to 10,000 nautical miles,
Decca accuracy was about 100 meters at ranges of up to 400 nautical miles,
and Loran C had an accuracy of about 1 nautical mile at ranges of up to
1,000 nautical miles (1,800 km).24
3.3.4.3 Enhanced Loran (eLoran)
eLoran has been considered as a possible backup system for GNSS. eLoran
operates like to DGPS in providing differential correction to GNSS in local
areas. It is currently not widely used except for research and testing in the
United States and the United Kingdom, and it is in active use in South Korea.
Testing performed by the General Lighthouse Authorities of the United
Kingdom and Ireland (GLA) with eLoran experienced accuracies of between
8 and 10 m (95%) to seven ports on the east coast of the United Kingdom.25
3.3.4.4 Automatic Identification System (AIS)
AIS is used to provide vessel identification and other information for use in
collision avoidance and general maritime domain awareness. Data is commu-
nicated in real time between ships using VHF radio channel 87B (Simplex),
also known as AIS channel 1 (161.975 MHz), and ship to shore for VTS
using VHF channel 88B (Duplex) as AIS channel 2 (162.025 MHz).26,27
Other channels may also be used depending on national and regional regu-
lations. Class A AIS generates output power of 1, 12.5, or 25 watts and is
primarily intended for use by larger vessels of greater than 300 gross tons
(GT) and all passenger vessels. Class B AIS, capable of generating an output
power of 2, 5, or 12 watts, is intended for smaller vessels and recreational
craft use. AIS messages comprised of digital data packets encoded with vessel
name, Maritime Mobile Service Identity (MMSI) number, position, speed,
course, type of vessel, ports of origin and destination, and other naviga-
tional information are continuously broadcast every 2–15 seconds between
messages depending on vessel speed. AIS messages can also be received by
satellite-based AIS receivers, allowing for global coverage and monitoring of

vessel movements.
3.3.4.5 AIS Aids to Navigation (AIS-A toN)
AIS can be used to augment the presence of a physical AtoN, such as a buoy,
by providing a corresponding electronic presence at the same location on
Radar and ECDIS broadcasting both laterally (e.g., Port Hand Mark) and
non-laterally significant marine safety information (e.g., environmental data,
tidal information and navigation warnings).28 This is accomplished by fitting
an AIS transmitter broadcasting using VHF radio on AIS channels 1 and 2
directly to the physical AtoN itself, installing one remotely within the line of
sight of the physical AtoN and referred to as a synthetic AtoN, or virtually
with installation in a location within line of sight to where a physical AtoN
should exist but cannot be placed due to inaccessibility, consistently rough
weather conditions or other environmental factors and referred to as a vir-
tual AtoN. AIS AtoN can broadcast autonomously and at fixed intervals,
providing the name, position, dimensions, type, characteristics and status
from or concerning an AtoN.
3.3.4.6 Mobile Aids to Navigation (MAtoN)
Another type of AIS VHF radio-based AtoN is the Mobile Aid to Navigation
(MAtoN), which is defined as a non-fixed or un-moored Aid to Navigation
(AtoN) and does not include a fixed or moored buoy that is adrift from sta-
tion, temporarily or otherwise.29 MAtoN may exist in physical or virtual
form and can be used for Ocean Data Acquisition System (ODAS) (e.g., to
gather data on currents and weather), wreckage (e.g., containers, debris),
water quality and pollution monitoring equipment, dynamic guard zones and
convoys, underwater operations, enhancing navigational safety during mili-
tary operations (e.g., no sail zones during minesweeping, target, exercises
areas), towed and deployed applications (e.g., cable laying), search & rescue
applications and special events (e.g., swimming competitions). It is specific-
ally noted that MAtoN should not be used for unmanned vessel applications.
3.3.5 Dead Reckoning

Dead reckoning is an ancient navigation technique for determining the pre-
sent approximate position based on projecting past courses steered and
speeds over the ground from a known past position.30 Not considered in
DR navigation are the effects of leeway in the form of lateral drift off course
caused by wind, currents and other environmental factors for which course
adjustments are made as appropriate. Error is cumulative in that the error
inherent in the present position estimate builds upon that which existed in all
previous position estimates since leaving a known accurate position. DR is

useful for predicting the times of sunrise and sunset, arrival at landfalls, and
the sighting of lights and prominent landmarks. This technique is still used in
modern times to provide redundancy in navigation methods and cross-check
the accuracy of celestial and electronic navigation sensors. However, the sen-
sors involved and methods used to perform them have changed dramatically.
3.3.5.1 Traditional DR Instruments
Various devices are necessary to perform DR, including a properly calibrated

magnetic compass, navigation chart, logline or chip log, accurate chronom-
eter or timepiece, parallel ruler, navigation divider and a log book to record
observations and events. The compass, corrected for magnetic deviation
caused by the vessel’s unique physical and magnetic characteristics, deter-
mines the ship’s course relative to magnetic north. The navigation chart, in
addition to containing information essential to navigation and magnetic vari-
ation in the angular difference between true north and magnetic north, is
used to plot the ship’s course, measure distances, and identify landmarks
or reference points encountered and anticipated along the way. The logline
and chip log are devices used to measure the speed of a vessel through the
water based upon the amount of line played out from a spool over a period
of time, or a device deployed in the water and timed as to how long it takes
to move between two known distances such as the bow and stern of the
vessel. A watch, chronometer, or other timepiece, or electronic time refer-
ence known to be accurate is essential to record the times of observations or
events that have occurred throughout the voyage from which elapsed time
and distance traveled may be calculated. A parallel ruler and navigation div-
ider are used for plotting and measuring distances on the navigation chart.
3.3.5.2 DR using Inertial Navigation
Modern-day Inertial Navigation Systems (INS) use the principles of dead

reckoning to estimate position and velocity using a combination of acceler-
ometers and gyroscopes to measure the acceleration and rotation of a vessel
along several axes, oftentimes nine or more. Calculating changes in vessel
position and velocity from an initial known position and velocity is per-
formed by integrating these measurements over time. It is updated at regular
intervals to provide a new estimated position.
Similarly, as traditional DR with the absence of external references for
known positioning and velocity in PNT data, INS suffers the same problem of
progressively increasing errors in velocity compounded into even larger errors
in position resulting from integration drift. These errors include accelerom-
eter bias, gyroscope drift, sensor noise, scale-factors, g-sensitive-factors, non-
linearity sensitivity, packaging, gravity modeling errors, and environmental
factors.31 Unlike traditional DR, INS employs advanced signal processing,

error modeling and filtering algorithms combined with sensor fusion with
external references, including GNSS. However, inertial navigation errors
accumulate over time due to sensor inaccuracies and external disturbances.
3.3.6 Celestial Navigation

The use of astronomy in terms of time, celestial coordinates and motion of
the stars, planets, moon and the sun for ship navigation dates from before
the ancient Egyptians and the ancient Greeks developed many of the celestial
navigation techniques still in use today. The invention of the chronometer that
made precise timekeeping possible became widely available in 1759, and the
sextant that helped mariners precisely measure angles between astronomical
objects and the horizon to determine latitude above and below the Equator
originated around 1791.32,33 Celestial navigation has been downgraded in
importance due to the advent of modern GNSS and other technology. It
is rarely studied except in an academic environment or used operationally
except by a few sailors and hobbyists. Despite this trend, celestial naviga-
tion remains a viable means to provide an independent source of positioning
information to confirm the accuracy of GNSS readings, especially consider-
ing its limitations in terms of degradation, DOS attacks, and spoofing events.
It can also provide a viable position reference for Inertial Navigation System
(INS) alignment.
3.3.7 Weather Instrument Use in Navigation

This chapter described satellite communications as a means to provide real-
time weather services and long-term forecasts directly to vessels at sea and
how this data is becoming increasingly important for efficient route plan-
ning, weather avoidance and better-informed voyage decision-making. In
reality, seafarers have been acutely aware of the weather for centuries and
how observations of its effects can benefit or hinder ship navigation. The
barometer has long been used to measure atmospheric pressure and how
changes and trends in this measurement can help predict changes in the wea-
ther. Accompanying the barometer is the thermometer that measures air tem-
perature, which can also help predict weather changes; and the hygrometer
and psychrometer that measures humidity and dew point, which can help
predict the likelihood of precipitation and fog. The wind vane and anemom-
eter to measure wind direction and speed can help to plan a safe route and
predict the severity of storms and hurricanes. When properly calibrated and
maintained, weather instruments can provide a seafarer knowledgeable in
their use and limitations with unique and real-time insight into conditions
directly affecting the safety of navigation that are likely to go unnoticed
by general area forecasts, especially in environments where few networked
weather sensors are available for professional weather forecasters to observe

and base their weather forecasts.
3.3.8 Other Sensor Data

Tides and currents can significantly affect a voyage, and proper planning can
use the effects of these phenomena to greatly enhance the safety and speed of
a journey. This is accomplished by anticipating water depths along the route
and avoiding shoal areas, avoiding hazardous currents and waves, especially
during the transit of harbors, inlets and areas where whirlpools and other
hazards may be prominent. Fuel efficiency can also be affected to take advan-
tage of following currents and minimizing head currents to help achieve bet-
ter fuel economy and realize cost savings and avoidance.
Predicted tide times and levels are available from many published tide
tables, and historical, current information is often available from local or
even national sources. Real-time tide and current data can also be available
over radio links and the Internet to data buoys equipped with tide gauges
to measure tides at the buoy location and current meters that measure the
speed and direction of surface and subsurface currents. An example of cur-
rent measurements is illustrated below from the Chesapeake Channel LBB
92 [cb1101] buoy located in Annapolis, Maryland, USA, measured at 0950
local time when the query was made:
Predictions (Depth: 22.0 ft). Data source: https://tidesandcurrents.noaa.gov/
map/index.html?id=cb1 101
Time (LST/L DT) Max/M in Speed (knots)

12:18 PM ebb −0.52
3:12 AM slack –
7:06 AM flood 1.06
10:54 AM slack –
2:42 PM ebb −1.02
6:36 PM slack –
8:18 PM flood 0.25
10:06 PM slack –
Today’s Currents
Speed: 0.34 kts
Direction: 35° (toward NE)
Currents measured at: 22.0 ft below the surface
Plotted data in both standard and metric units over a daily period of
between 0000 and 2359 is also available as desired.
3.4 The IMO e-N avigation Initiative

e-Navigation is defined by the IMO as:
the harmonized collection, integration, exchange, presentation and ana-

lysis of marine information on board and ashore by electronic means
to enhance berth to berth navigation and related services for safety and
security at sea and protection of the marine environment,34
and is intended to meet present and future user needs through harmonization
of marine navigation systems and supporting shore services. The development
and implementation of the concept is coordinated by the IMO as a global
collaborative effort to enhance maritime safety and efficiency by using elec-
tronic navigation. The strategy for e-Navigation implementation involves the
integration and use of information from various sources, including MF, HF
and VHF radio, GMDSS, Inmarsat, AIS, DGPS, Long-Range Identification
and Tracking (LRIT) and Enhanced Maritime Mobile Service (EMMS).35
Many technologies and services comprising e-Navigation have been dis-
cussed elsewhere in this chapter and in Chapter 7, Shore and Intership
Communications.
References
1 Chip log. Bright Hub Engineering. Albany, NY, USA. www.brighthubengineer
ing.com/seafaring/60582-what-is-a-chip-log/
2 Mariners Astrolabe. Mariners’ Museum & Park. Newport News, VA, USA.
https://exploration.marinersmuseum.org/object/astrolabe/
3 Backstaff. Mariners’ Museum & Park. Newport News, VA, USA. https://expl
oration.marinersmuseum.org/object/back-staff/
4 Using Lead Lines to Collect Hydrographic Data. National Oceanographic and
Atmospheric Administration (NOAA), USA. https://celebrating200years.noaa.
gov/transformations/hydrography/side.html
5 Bowditch. The American Practical Navigator. Chapter 1, The Art and Science
of Navigation. Pg. 1. Defense Mapping Agency Hydrographic/Topographic
Center. Bethesda, MD, USA. Pub. No. 9. 1995 Edition.
6 International Rules for the Prevention of Collisions at Sea, 1972 (COLREGS)
as Codified in the United States by Coast Guard Commandant Instruction
M16672.2D Navigation Rules –International and Inland. Rule 5, Look-out.
25 March 1995.
7 R. Glenn Wright. Innovations in Electronic Communications and Navigation.
Unmanned and Autonomous Ships. Routledge, Taylor & Francis Group.
Table 1.1, p. 7. 2020. ISBN: 978-1-138-12488-6.
8 NIST GNSS. Time and Frequency from A to Z. National Institute of Standards
and Technology (NIST), Physical Measurement Laboratory, Time and
Frequency Division. Gaithersburg, MD, USA. www.nist.gov/pml/time-and-

frequency-division/popular-links/time-frequency-z/time-and-frequency-z-g
9 Keiko Shimazu et al. Emergency Warning Services via GNSS Signals. 2020 IEEE
Aerospace Conference. Big Sky, MT, USA. Institute of Electrical and Electronics
Engineers. March 2020. DOI: 10.1109/AERO47225.2020. 9172377.
10 Brent Rose. The Best GPS Communicator to Buy for Your Next Adventure. The
Verge, VOX Media Network. 12 April 1019. www.theverge.com/2019/4/12/
18306207/best-gps-communicator-hiking-trails-garmin-spot-somewear-bivy
11 Glen Gibbons, 2007. Selective Availability: Completely Dead. Inside GNSS. 20
September 2007. https://insidegnss.com/selective-availability-completely-dead/
12 ESRI 2023. Differential GPS Explained. ESRI. Redlands, CA, USA. www.esri.
com/about/newsroom/arcuser/differential-gps-explained/
13 Alexandra Witze. Global 5G Wireless Networks Threaten Weather Forecasts.
Nature. 26 April 2019. www.nature.com/articles/d41586-019-01305-4
14 Robert J. Meurn. Watchstanding Guide for the Merchant Officer. Third
Edition. Cornel Maritime Press, 2014. Case Study: Andria Doria/Stockholm
Collision. p. 177. ISBN: 978-0-87033-631-7.
15 C. Lindstrom, R. Christensen, J. Gunther, and S. Jenkins. (2022) GPS-denied
navigation aided by synthetic aperture radar using the Range- Doppler
Algorithm. NAVIGATION, 69(3). https://navi.ion.org/content/69/3/navi.533.
abstract.
16 F. Ali, G. Bauer, and M. Vossiek, “A Rotating Synthetic Aperture Radar
Imaging Concept for Robot Navigation,” in IEEE Transactions on Microwave
Theory and Techniques, 62(7), pp. 1545– 1553, July 2014, DOI: 10.1109/
TMTT.2014.2323013
17 Short-Range Marine Radar FAQ. ELVA- 1 Microwave Handelsbolag.
Stockholm, Sweden. 28 July 2022. https://elva-1.com/blog/e-band-short-range-
marine-radar-faq
18 Sawada, R. and Hirata, K. (2023) Mapping and localization for autonomous
ship using LiDAR SLAM on the sea. Journal of Marine Science Technology.
https://doi.org/10.1007/s00773-023-00931-y
19 OS2. Ouster OS2 Long-Range High-Resolution Imaging Lidar. Ouster, Inc. San
Francisco, CA, USA. https://data.ouster.io/downloads/datasheets/datasheet-
rev7-v2p5-os2.pdf
20 Bowditch. The American Practical Navigator. Chapter 13, Radar Navigation,
pp. 213– 214. Defense Mapping Agency Hydrographic/ Topographic Center.
Bethesda, MD, USA. Pub. No. 9. 1995 Edition.
21 International Rules for the Prevention of Collisions at Sea, 1972 (COLREGS)
as Codified in the United States by Coast Guard Commandant Instruction
M16672.2D Navigation Rules –International and Inland. Rule 7, Risk of
Collision, 25 March 1995.
22 NOAA 2016. How Accurate Are Nautical Charts? NOA Coast Survey. 8 April
2016. https://nauticalcharts.noaa.gov/updates/how-accurate-are-nautical-
charts/
23 Bowditch. The American Practical Navigator. Chapter 13, Radio Direction
Finding. p. 176. Defense Mapping Agency Hydrographic/Topographic Center.
Bethesda, MD, USA. Pub. No. 9. 1995 Edition.
24 Google Bard. Response to Question Posed on 1 June 2023. https://bard.goo

gle.com/
25 Alan Grant and Dana Goward. 10 Answers about eLoran. GPS World. 11 April
2022. www.gpsworld.com/10-answers-about-eloran/
26 NATO, 2023. AIS (Automatic Identification System) overview. NATO Shipping
Centre. MARCOM Northwood, UK. https://shipping.nato.int/nsc/operations/
news/2021/ais-automatic-identifi cation-system-overview
27 MarineTraffic, 2023. Everything You Wanted to Ask about AIS. MarineTraffic
Blog., 2023. www.marinetraffi c.com/blog/ais-faq/
28 United States Code of Federal Regulations, Title 33. Chapter 1, Subchapter C,
Part 62.52 Automatic Identification System Aids to Navigation (AIS AtoN).
www.ecfr.gov/current/title-33/chapter-I/subchapter-C/part-62
29 IALA Recommendation R1016 Mobile Marine Aids to Navigation (MAtoN).
Edition 2.0, December 2020. urn:mrn:iala:pub:r1016. www.iala-aism.org/prod
uct/r1016/
30 Bowditch. The American Practical Navigator. Chapter 13, Dead Reckoning. pp.
113. Defense Mapping Agency Hydrographic/Topographic Center. Bethesda,
MD, USA. Pub. No. 9. 1995 Edition.
31 Andrey Soloviev. Inertial Error Propagation: Understanding Inertial Behavior.
Inside GNSS. 4 June 2022. https://insidegnss.com/inertial-error-propagation-
understanding-inertial-behavior/
32 Robert T. Gould. (1921) The history of the chronometer. The Geographical
Journal, 57(4), pp. 253–268.
33 Peter Ifland. Taking the Stars: Celestial Navigation from Argonauts to
Astronauts. The Mariners’ Museum. Newport News, VA, USA, 1998.
34 Strategy for the Development and Implementation of e-Navigation. International
Maritime Organization. MSC 85/26/Add.1, annex 20. London, 1998.
35 e-Navigation Strategy Implementation Plan –Update 1. International Maritime
Organization. MSC.1/Circ.1595. London, 25 May 2018.
Chapter 4
Cargo Sensors
Independent sensors unrelated to ship’s systems dedicated to track cargos and

ensure goods arrive at their destinations fresh and in good condition have
become widespread throughout the industry. Much of this trend was initi-
ated by cargo transport through containerization, which has helped simplify
the cargo handling process and standardize shipment methods. Almost every
product imaginable can be shipped using standard containers, including
clothing, textiles, furniture, appliances, machinery, mechanical and elec-
tronic equipment, and construction materials. However, there are also other
methods to transport cargo consisting of goods and products that cannot
fit within the confines of a standard container, and many ships have been
designed especially for this purpose.
Although the tracking of shipments has been one of the primary uses of
sensors, additional sensor needs exist beyond mere tracking for perishable
goods that require special handling and specialized containers to ensure tem-
perature, humidity, lighting, vibration, and shock are maintained at accept-
able levels throughout the journey. This includes fresh and frozen foods,
fruits and vegetables, meats, dairy products, liquids, beverages, medicines
and flowers. Even more unique and specialized sensors are needed for dan-
gerous goods requiring unique handling and extraordinary precautions, such
as flammable liquids, solids and gases; combustibles such as batteries; radio-
active materials, oxidizing and toxic substances.
Sometimes, it is also necessary to instrument the cargo spaces and even the
cargo itself to ensure its safety and that it is kept in good condition, to gain
additional insight into unique and undocumented behavior properties, and to
provide advance warning of possible risks to the ship in carrying such cargo.
Examples of shipments that need specialized sensors to maintain safety and
to help prevent the loss of the ship include car carriers transporting elec-
tric vehicles that use lithium-ion batteries subject to spontaneous combustion
and dry bulk carriers whose cargo contents may be prone to liquefaction.
This chapter reviews some ship configurations that carry cargo and high-
lights those with the most pressing need for sensors independent of onboard
engineering and navigation systems. Many of these sensors and their principles
Doi: 10.1201/9781003377900-4
Cargo Sensors 53
of operation have been discussed in the previous chapters on engineering

(e.g., temperature, humidity, vibration, accelerometers, etc.) and navigation
(e.g., Global Navigation Satellite System –GNSS). Therefore, a major focus
of this chapter is to consider the needs of cargos and ship configurations sub-
ject to the highest occurrence of accidents and casualties that threaten safety
in terms of loss of life, compromise of ship stability, loss of the entire ship,
environmental damage and subsequential economic impact.
4.1 Ships and Cargos

A brief glimpse is made at the different types of cargo ships that exist on the
high seas and many smaller cargo ships involved in coastal and inland ship-
ping. The most common types of cargo ships that account for the majority of
the world’s maritime trade are container ships, bulk carriers, tankers, Roll-
on/Roll-off (Ro-Ro) vessels and special cargo or multipurpose cargo ships.
Container ships can be the largest of all ships, designed to carry thousands
of standardized shipping containers that are easily loaded and unloaded. The
largest of these are the sister ships MSC Irina and MSC Loreto that entered
service in 2023, each with a length of 400 m (1,312 feet), a width of 61 m
(200 feet) and a capacity of 24,346 Twenty-foot Equivalent Units (TEUs).1
Evergreen’s Ever Alot is also of comparable size at 396 m (1,300 feet) in
length, 62 m (203 feet) in width and 24,004 TEU capacity. Many other ships
with container capacities of thousands to less than one hundred TEU service
smaller ports and communities.
Bulk carriers are ships designed to carry dry unpackaged cargoes featuring
from one to several hi-volume cargo compartments and can also be listed
amongst the largest of vessels. Their cargo includes large quantities of coal,
grain, iron and other ores. One example is MS Ore Brasil, which measures
362 m (1,187 feet) in length, 65 m (213 feet) in width and its hull depth (not
draft) is 30.4 m (115 feet).2
Tankers are also among the largest of vessels and come in various sizes,
shapes and configurations designed to carry liquid cargos that generally
include crude oil, refined oil products, gas and many types of chemicals. The
largest of these is Universal Leader, a Very Large Crude Oil Carrier (VLCC)
with a length of 336 m (1,102 feet) and a width of 60 m (197 feet).3 Another
tanker of a very different design is Shinshu Maru that carries liquid natural
gas whose length is 399.9 m (1,309 feet) and width is 48 m (157 feet).
Ro-Ro carriers transport vehicles and other wheeled cargo that can be
driven on and off the ship. These can range from small inland and coastal
ferries that also carry passengers to large vessels transporting cars, trucks,
railroad cars, cranes and other construction machinery to destinations world-
wide. The largest of these is MV Tonsberg, with a length of 265 m (860 feet),
width of 32 m (105 feet) and draft of 11 m (36 feet). This vessel has six
54 Cargo Sensors
decks and can carry heavy rolling and non-containerized cargo, breakbulk
and cars.4
General cargo ships or multipurpose vessels are flexible in their use and
can transport a wider variety of goods and cargos including containers,
bulk cargo and Ro-Ro vehicles. There are many ships in this category, and
the size leader by individual ship changes frequently. However, in 2019,
the ten largest operators by deadweight of multipurpose-project-heavy-lift
tonnage deployed a combined fleet of 476 ships with a total deadweight of
8,220,000 million tons and an aggregate lifting capability of 155,000 tons.5
Other vessels that transport specialized cargo include ships that transport
other smaller ships, refrigerated ships for transporting fish cargo, livestock
carriers, ships that transport large aircraft components such as fuselage sec-
tions, crane ships that operate at sea, open deck ships that transport giant
cranes for port use, ships that recover space rockets after launch and trans-
port them to port, and just about most anything else that can be moved from
one place to another.
4.2 Containerized Cargo

In times past, the Captain has been held liable for what happens to the ship
and environs due to being knowingly negligent in failing to take proper
precautions to transport hazardous cargo and mishandling non-hazardous
cargo. However, the situation is much less clear when operations involve a
ship transporting thousands of containers at a time. In such cases it is virtually
impossible for a ship’s crew to inspect each container and they must rely on
the accuracy of bills of lading, hazardous materials declarations and labeling
to be aware of the onboard dangers of transported cargo. It is incumbent
upon the shipper to guarantee the loading of individual containers is properly
accomplished to safeguard their contents and that they are mutually compat-
ible and not likely to come in contact, interact and combust. The problems
associated with shipping containers are well documented, and the headlines
are constantly filled with news stories about mishaps on land and at sea in
which they are the source of risk and loss.6,7,8 Some of the different sensors
for the various uses that can exist onboard container ships are described in
the following paragraphs.
4.2.1 Container Hold Sensors

Enclosed spaces within the hull designed to carry shipping containers are
equipped with sensors to confirm the security of hatch covers. In addition to
visual inspection by direct crewmember observation and video cameras to
verify closure, other sensors may confirm or contradict the visual indications
of whether a hatch cover is closed. One example is the use of pressure sensors
Cargo Sensors 55
to detect differences in air pressure between the outside and inside of the
container hold, where the same pressure may indicate open and differences
in pressure indicate closed. Magnetic sensors may also be distributed along
the hatch and frame to detect gaps and determine if the hatch is open or ajar.
Ultrasonic leak detectors may also detect changes in airflow or pressure that
may represent air or gas leaks around the hatch covers.
Additional sensors are also required in container holds to detect and sup-
press fire.9 These include temperature, smoke, and heat sensors to detect
indications that a fire may be present. Also, sensors are required to monitor
oxygen concentration within the space as part of a carbon dioxide (CO2) fire
suppression system to verify and maintain an inert atmosphere during and
after a system discharge.
4.2.2 Shipping Container Sensors

In 1967, the International Standards Organization (ISO) and International
Maritime Organization (IMO) began the process of container standardiza-
tion that resulted in the adoption of the International Convention for Safe
Containers (CSC) in 1972.10 This convention requires shipping container
doors to be marked with alphanumeric codes to assist in tracking shipments.
The codes contain a unique identifier and owner information, the dimensions
and type of container, operational marking representing maximum gross
weight and container weight (tare), and a CSC plate verifying good condi-
tion and acceptable safety. This method made possible the identification and
tracking of individual containers throughout their journey and even their
location on the ship.
Initial efforts using electronic code readers on containers increased port
efficiency somewhat but proved cumbersome by creating new needs to pos-
ition these readers accurately to achieve desired results. In 2004– 2005,
efforts were undertaken to mark shipping containers with Radio Frequency
Identification (RFID) tags to enable wireless tracking of shipping container
movements and detect events such as door openings with reporting via local
communications networks, mobile data services and satellites.11,12
In the following years, many factors have contributed to the development
of “smart containers” comprised of a wide range of sensors to provide iden-
tification of the container and its contents and also to gather information
regarding GNSS positioning, temperature, humidity, shock, and acceleration,
the presence of explosive gases, smoke and many other factors resulting in
vastly expanded data collection efforts. Sensors embedded within containers
should also help to detect container fires onboard ships that have claimed
many lives and resulted in the loss of ships and their entire cargos.13,14 The
occurrence of container ship fires are frequently reported in the news. A short
list of examples include:15
56 Cargo Sensors
• In 2018, the Ultralarge Container Ship (ULCS) Maersk Honam vessel

caught fire, killing five crewmembers.
• In 2017, the 13,800 TEU MSC Daniela was on fire for more than a week.
• The 2016 fire on the 9,000 TEU CCNI Arauco in Hamburg, Germany.
• In 2012, a fire onboard the container ship MSC Flaminia burned for six
weeks and killed three crew.
The containers collect and transmit data about their location, condition and
cargo, which can be used to optimize and improve the efficiency of container
routing. These data can also enhance security with sensors that can detect
tampering, unauthorized access, theft of contents, and cargo damage due to
shock; and promote sustainability of the shipping supply chain by tracking
the progress of the shipment and to identify potential delays. This is especially
needed due to the increasing number of containers lost at sea with resulting
pollution and their widespread dispersal across the seas as large hazards to
navigation. Smart shipping containers equipped with sensors, GNSS tracking
and other Internet of Things (IoT) devices are now replacing the simple steel
containers and discrete sensors of yesterday.
4.2.3 Parametric Rolling

A vessel behavioral phenomenon known as Parametric Rolling Movement
(PRM) can cause ships to roll at extreme angles of up to 30–40 degrees or
more with very little notice, even in low to moderate sea states. This is seen
primarily with container ships and car carriers, both of which feature hull
forms with flared fore and aft decks. Other types of vessels may be simi-
larly affected in head or following seas. Container ships may experience
exceptionally high container losses when combinations of rolling period,
wave conditions, ship’s speed and heading cause sudden and extremely
rapid increases in roll motions, even in relatively mild wave heights.16,17
Such conditions can be very hazardous and likely occur when wave lengths
approximate vessel length and the rolling period is twice the wave encounter
period.16 This can result in extreme stresses on containers and cause failure
of their securing system and loss of containers, and may possibly capsize
the ship itself.18
The Danish company Kjærulf Pedersen reports it has developed a sensor
system that can recognize the tendencies that lead to parametric rolling and
send a message to the ship’s controls so the system and crew have time to
take corrective action. Based on its inclinometer sensor product designed to
measure orientation using an accelerometer, gyroscope, and magnetometer;
digital signal processing is used to perform inclinometer calculations.19 Three
sensors placed at the bow, amidship and stern record the ship’s movements
and compare observations with the ship’s speed, acceleration and direction
to detect whether the movements are approaching a state that can become
Cargo Sensors 57
uncontrollable.20 Ship crews may then be alerted to promptly take measures

such as course changes and speed adjustments to overcome this condition.
4.3 Bulk Cargo

The Intercargo 2019 Bulk Carrier Casualty Report provided an analysis of
statistics covering 2010–2019, citing thirty-nine bulk carriers with 173 sea-
farers as having been reported lost over that period.21 Cargo shift and/or
liquefaction was identified as one of the greatest concerns and likely to be
the main reason behind eight bulk carrier casualties and 106 seafarers los-
ing their lives. Losses included six bulk carriers carrying nickel ore from
Indonesia, one with iron ore fines (assumed to be laterite) loaded with high
moisture content from Malaysia, and one with bauxite from Malaysia. These
statistics were updated in 2023, citing that in the ten years since 2013, more
than two-thirds of the 104 seafarers who perished in the 26 bulk carriers
lost at sea died due to cargo liquefaction.22 Other causes of bulk cargo ship
losses included grounding as the most common with 17 losses, flooding for
five ship losses, and four ships lost due to unknown causes accounting for
35 lives.
4.3.1 Cargo Liquefaction

A significant problem for bulk cargo ships is the fact that solid granular bulk
cargo such as crushed nickel ore, iron ore fines, bauxite fines, mineral con-
centrates, and some by-products like “red mud” can behave like a liquid dur-
ing transport sloshing around in the hold and leading to loss of stability.23,24
A major vulnerability exists when the cargo undergoes liquefaction, moves
to one side of the ship and solidifies, adversely affecting stability. Such condi-
tions can occur when moisture content and degree of saturation of the cargo
are sufficiently high, and a combination of conditions exist related to particle
size distribution, pressure within the particle pore spaces, and compaction
that can reduce friction between particles and allow them to flow and shift.
One example attributed to cargo liquefaction is the January 2015 loss of
Bulk Jupiter and 18 crewmembers carrying 46,400 tons of bauxite that sank
off the coast of Vietnam, with only one survivor.25 Unfortunately, there are
also many other examples.
To help mitigate the risk, international regulations and guidelines such as the
International Maritime Solid Bulk Cargoes (IMSBC) code provide standards
and procedures for the safe transportation of solid bulk cargoes, including
testing and moisture limits, certification of cargoes, monitoring moisture
content and ensuring suitable ship stability calculations are performed before
loading and during the voyage. Cargos known to liquify are listed as Class
A under the code. Each type of solid cargo now has a Transportable Moisture
Limit (TML) rating 10% beyond the product’s Flow Moisture Point (FMP),
58 Cargo Sensors
the percentage of moisture in the product at which it may begin to behave

like a liquid. Should the Moisture Content (MC) of the cargo exceed TML,
it is not considered safe or suitable for shipment. Additional approaches to
ensure safety during transport include installing sensors inside cargo holds
to monitor moisture content, humidity and water pressure within the cargo.
Lasers may also be used to monitor changes in cargo position to provide
advance warning of unsafe conditions.24
However, the IMSBC code does not necessarily list all cargos that may
liquefy. P&I Club Gard recently identified cargos such as fluorspar, certain
grades of coal, pyrites, mill scale, sinter/pellet feed, etc., as also being able
to liquefy and provide many references on the subject.26 In their warning,
they cite one incident where a vessel carrying over 1,900 tons of soil from
a landfill listed eventually sank due to wind and waves, causing the cargo
to behave like liquid during transit.27 Another incident was a cargo of cal-
cium carbonate that caused a vessel to develop a list shortly after leaving the
load port. The cargo’s moisture content was greater than 30%, exceeding the
TML of 24% and Flow Moisture Point (FMP) of 26.7%, where the material
consisted entirely of particles less than 2 mm.
4.3.2 Explosion
An explosion took place, and four crew members were injured on board the
120,600-dwt bulk carrier CSSC Cape Town as it entered British Gibraltar
Territorial Waters in the Bay of Gibraltar on 19 February 2021.28 The ship,
built in 2020, was loaded with 112,365 metric tons of coal at the Curtis Bay
Coal Terminal in Baltimore, Maryland, USA. According to the report, the
explosion appeared to be in the area of the vessel’s forecastle, and the cause
of the explosion is not known. No fire was reported.
Coal is classed as a non- dangerous cargo under the Convention on
Facilitation of International Maritime Traffic. However, coal dust suspended
in the air can be highly explosive and very susceptible to spontaneous com-
bustion because it has much more surface area per unit weight than lump
coal. For an explosion to occur, five simultaneous elements must be pre-
sent: fuel, heat, oxygen, suspension and confinement.29 A fire may occur
should fuel, heat, oxygen and confinement conditions exist in proper quan-
tities. However, an explosion could occur with the suspension of burning fuel
as may result from the introduction of a sudden blast of air.
The primary sensors for determining coal dust concentrations are cumber-
some and involve frequent maintenance and replacement of filters. The dust
can easily block and pollute photodiodes of light-scattering dust concentra-
tion optical sensors. New research in electrostatic induction coal dust con-
centration sensors has yielded high detection accuracy for coal dust and may
be suitable for future use in bulk cargo ships.30
Cargo Sensors 59
4.3.3 Unpreparedness to Fight Fires

Numerous opportunities can lead to shipping companies and crewmem-
bers being ineffective in fighting ship fires. The July 2020 fire aboard the
amphibious assault ship USS Bonhomme Richard (LHD 6) at Naval Base
San Diego while it was moored pier side for maintenance is a good case
study for such events. In a report released by the U.S. Navy describing the
complete loss of the vessel, four categories of causal factors allowed for the
accumulation of significant risk and led to an ineffective fire response.31 These
included the material condition of the ship, the training and readiness of
the ship’s crew, the integration between the ship and supporting shore-based
firefighting organizations, and the oversight by commanders across multiple
organizations. The command investigation also concluded that “a lack of
familiarity with requirements and procedural noncompliance at multiple lev-
els of command” contributed to the loss of the ship.
This incident did not involve a civilian bulk cargo ship, and there were
many heroic acts of bravery, ingenuity and resourcefulness in the actions of
sailors across the San Diego waterfront and others who had a role in the fire-
fighting response. However, lessons learned from this naval equivalent must
highlight the fact that gross incompetence and negligence at the highest levels
resulted in most of the onboard smoke and fire detection sensor capabilities
and 87% of the ship’s fire stations remaining in an inactive equipment main-
tenance status that prevented the ship’s aqueous film forming foam sprinkling
system from being used to fight the fire.32
4.4 Liquid Tank Cargo

Cargo monitoring is an integral part of the engineering sensor suite for most
ships built to specifications for specific purposes, such as oil tankers and
Liquid Natural Gas (LNG) carriers. These sensors are generally fixed systems
that aim to detect unsafe and explosive gas–air mixtures in cargo tanks and
holds, ballast tanks and void spaces before they reach the Lower Explosion
Limit (LEL). However, accidents with catastrophic results still occur. The fol-
lowing paragraphs discuss the unique aspects of some of these accidents to
illustrate potential additional uses of sensors that may help to avert accidents
in the future.
4.4.1 Cargo Tank Overheating

The UK Marine Accident Investigation Branch (MAIB) report on the inves-
tigation of the 28 September 2019 cargo tank explosion and fire on board
the chemical tanker Stolt Groenland while alongside in the port of Ulsan,
Republic of Korea, found that heated cargo tanks in nearby compartments
60 Cargo Sensors
likely caused dangerous warming of a temperature-sensitive cargo of styrene

monomer.33 Details of this investigation cited in the report include:
The rupture of the styrene monomer tank resulted from a runaway poly-
merisation that was initiated by elevated temperatures caused by heat
transfer from other chemical cargoes. The elevated temperatures caused
the inhibitor, added to prevent the chemical’s polymerisation during the
voyage, to deplete more rapidly than expected. Although the styrene
monomer had not been stowed directly adjacent to heated cargo, the
potential for heat transfer through intermediate tanks was not fully appre-
ciated or assessed. Critical temperature limits had been reached before the
vessel berthed under the road bridge in Ulsan. The tanker’s crew did not
monitor the temperature of the styrene monomer during the voyage, and
therefore were not aware of the increasingly dangerous situation.
This accident indicates a need for the use of one or more smart sensors cap-
able of performing the task supposed to be accomplished by the crew to
monitor the temperature of the styrene monomer and the environment in
the immediate vicinity and to promptly analyze and report the findings to
crewmembers. Several environmental characteristics, such as temperature,
trace gases and other phenomena appropriate to styrene monomer, could
be monitored simultaneously using multi-sensor elements within one instru-
ment enclosure. In addition to monitoring specific environmental character-
istics, integrating machine learning and deep-learning artificial intelligence
(AI) technologies into the sensors can discern interrelationships between the
characteristics not readily detectable through human senses or understand-
ing by examining data continuously over periods to detect evidence of trends
and potential pending events. Such capability is appropriate in preventing
this accident since the cargo paperwork called for maintaining tank tempera-
ture below 30°C. The tank reached 50° C three days before the event, 65°
C the day before, then rapidly passed 90° C shortly before the explosion.
The report identifies a similar spontaneous heating incident in August 2019
affecting a styrene cargo aboard the tanker Stolt Focus. In this instance, the
crew added seawater to the cargo, then diluted it with benzene and added an
inhibitor to bring the situation under control.
4.4.2 Explosion and Fire

The Aframax-class crude oil tanker MT Pablo, sailing under the flag of
Gabon, exploded and caught fire off the coast of southern Malaysia on 1
May 2023 while en route to Singapore with 28 crew members.34,35 Twenty-
five crewmembers were rescued, and three were killed. Built in 1997 and
capable of carrying 700,000 barrels of crude oil, the ship was reported to
be nearly empty after offloading its cargo in China. Bloomberg reports the
Cargo Sensors 61
Marshall Islands-registered company that owns the ship does not own any
other ships, and no trace of insurance could be found. This highlights the
risk of a growing dark fleet of aging vessels transporting sanctioned oil as the
result of purchases of hundreds of old tankers by undisclosed buyers.
The cause of the fire is unknown, but since it appears to have been nearly
empty, there is a great possibility that vapors from the remains of the oil
cargo played a part in this accident. However, the conditions leading up to
this accident likely relate directly to the use of old ships subject to the most
pervasive types of structural problems of corrosion and fatigue cracking
known to be major threats to the structural integrity of aging vessels, espe-
cially tanker structures and bulk carriers that operate beyond their design
service life.36 While modern sensor technology may have detected concentra-
tions of explosive vapors before the explosion, their installation and use on
an old crude oil tanker is not very probable nor likely to make them safer.
4.4.3 Another Explosion and Fire

An explosion occurred on 1 July 2023 in the tanker Sagar Nandini-2 while
anchored in the river Sugandha near Jhalakathi, Bangladesh.37 According to
the report, the ship was laden with gasoline (petrol) and diesel at the Padma
Oil Depot in Jhalakathi town when the engine room exploded, followed by a
fierce fire that engulfed the vessel. It was extinguished before it spread beyond
the engine room, and no damage occurred to the oil since the chamber con-
taining the oil did not catch fire.
While the exact cause of this accident had not been reported, the spontan-
eity of the explosion and the fact that the vessel was at anchor suggests sev-
eral possibilities. These include the ignition of fuel vapors from empty tanks
or flammable gas such as methane or hydrogen due to lack of proper ventila-
tion, electrical malfunction leading to sparks or ignition of nearby combust-
ible materials with subsequent fuel or gas ignition, chemical reactions caused
by reactive incompatibilities and coming into contact or being improperly
stored; and human error or sabotage.
The likely causes of the explosion and ensuing fire initially appear to be
unrelated to the ship’s function as an oil tanker and might have occurred on
any ship. Even so, the explosion suggests that the accumulation of fuel or
combustible gas would be capable of being detected using one or more cor-
rectly installed and placed smart sensors before igniting a spark. The elem-
ents that might cause an explosion were not detected, suggesting a lack of
operating sensors for such conditions or human error resulting from ignoring
sensor indications and warnings. Sabotage may also be a possibility where
human activity sensors and cameras, if installed, could have detected the
presence of one or more intruders and recorded their actions or the exist-
ence of one or more unauthorized devices within the engine room before the
explosion.
62 Cargo Sensors
4.4.4 Combustion Caused by Welding

In many instances concentrations of leaking gases, liquids and combustible
dust may appear at unexpected locations across the ship where fixed moni-
toring equipment is not installed or has become deactivated through routine
maintenance procedures where gases released during loading, unloading, or
cleaning may go undetected.
One such case is an accident where the oil tanker Smooth Sea 22 exploded
on 17 January 2023 while docked on the Mae Klong River in the Muang
district of Thailand.38 The explosion killed one welder and left four welders
injured. As in the previous example, the cargo may or may not have directly
participated in the explosion. However, the crewmembers, temporary work-
ers performing the welding and the ship itself were all placed at risk. In such
cases, it may prove extremely valuable to scan areas with portable sensors
before work commencement and where workers are present to temporarily
install smart sensors capable of detecting explosive dust and vapors that may
have accumulated outside of their normal confines.
4.5 Ro-Ro Ships

Ro-Ro carriers are subject to two threats in particular. One threat, Parametric
Rolling Movement (PRM), is shared with container ships and is described in
a preceding paragraph. The other major threat is fire onboard Ro-Ro carri-
ers that transport vehicles fueled by highly combustible gasoline (petrol) or
somewhat less combustible diesel fuel. This issue is highly problematic with
these vessels, where the compact loading of cargo described as “a worst-case
parking garage” makes it easy for the fire to spread, and firefighting becomes
much more difficult.39 Added to the equation is the diversity and types of
cargo, which make it hard to determine what type of fire needs to be fought,
and openings on the decks that make detection and localization difficult.
Vehicles are stowed 10–15 cm apart during voyages, making access challen-
ging. Typically, fire can spread between vehicles within an alarming five to
ten minutes, heightening the importance of identifying the fire’s location as
early as possible.40
The names of some of these car carrier ships are familiar as they have been
prominent in the news:
• Grande Costa d’Avoria when in July 2023, while completing used vehicle
loading operations in Newark, New Jersey, USA, a fire started on deck
number ten and burned for more than six days.41
• Höegh Xiamen burned for eight days in June 2020, started by a fire caused
by an electrical fault from an improperly disconnected battery in a used
vehicle while in the harbor in Jacksonville, Florida, USA.42
Cargo Sensors 63
• Grande America when in March 2019, it burned and sank in the Bay
of Biscay off France when a container that was part of the cargo caught
fire.43
The list of names continues with Euroferry Olympia (2022), Felicity Ace
(2022), Eurocargo Trieste (2019), Grande Europa (2019), Golden Ray
(2019: capsized) and several others.
Conventional smoke and heat fire detectors require high threshold lev-
els to trigger a warning and are not sensitive or accurate enough to detect
slight thermal changes. Smoke may be rerouted away from smoke detectors
by onboard extractor fans, making the detection harder.42 At present, cargo
holds are routinely inspected by crewmembers using predefined walkways
between the vehicles to supplement fire detection systems. This is a lengthy
and manual process and not always effective, especially when fire risks are
not always visible to the human eye. Using infrared cameras to detect heat
signatures may help in this endeavor.
One effort currently underway to help resolve these problems is the
Swedish Legislative Assessment for Safety Hazards of Fire and Innovations
in Ro-Ro Ship Environment (Lash Fire). This initiative is intended to inves-
tigate these problems and provide a recognized technical basis for revising
international IMO regulations to greatly enhance fire prevention and ensure
independent management of fires on Ro-Ro ships in current and future fire
safety challenges.44
4.5.1 Fires Caused by Lithium-I on Batteries

A new threat has emerged due to transporting Tesla and many other elec-
tric vehicles powered by lithium-ion batteries that are well known to self-
combust from thermal runaway and result in extremely difficult-to-extinguish
fires. Statistics from the U.S. Bureau of Transportation Statistics and the U.S.
National Transportation Safety Board found that hybrid vehicles had the
most fires per 100,000 vehicles at 3,474.5; gasoline vehicles were next with
1,529.9 fires per 100,000 vehicles and fully electric vehicles next with 25.1
fires per 100,000 vehicles.45,46,47 However, electric vehicles can require a tre-
mendous amount of effort from emergency personnel to extinguish the blaze.
Lithium-ion batteries also burn hotter and much longer than gasoline and
may require tens of thousands of gallons of water to extinguish.
The Safetytech Accelerator challenge was performed in partnership with
Wallenius Wilhelmsen, one of the world’s largest global vehicle carrier opera-
tors.42 This initiative sought sensor solutions to provide early, real- time
detection of significant thermal increases associated with transported vehi-
cles with few or no false positives. The solution must also identify the exact
location of the heat change on a ship’s deck and accurately communicate
64 Cargo Sensors
the change in real-time through a warning panel. A three-day pilot project

provided insight into how a sensor network can be deployed and maintained
with reliable results during harsh sea conditions. A Ro-Ro vessel with a typ-
ical deck area between 55,000 and 65,000 square meters for cargo would
require about 40,000 sensors to provide complete coverage on a ship using
a wireless mesh network. Encasing the sensors in robust packaging for the
marine environment and fine-tuning their accuracy would cost around 1 USD
per node to install, excluding maintenance costs. Costs could be minimized
using five-year batteries in the temperature sensor nodes, assuming a typical
ship service interval of around five years. Further thermodynamic tests that
simulate a vehicle fire more realistically will need to be performed before this
technology can be deployed across the fleet.
However, the threat of lithium-ion batteries is not limited to within cargo
spaces. The U.S. National Transportation Safety Board (NTSB) in their
report determined a November 2022 fire on the oil tanker S-Trust was
caused by the thermal runaway of a cell within a handheld radio’s lithium-
ion battery.48 The event resulted in extensive damage throughout the bridge,
including significant smoke and thermal damage, amounting to $3 million
in total.
4.5.2 Refrigerator Unit and Other Fire Sources

Before concern regarding lithium- ion batteries became prominent, the
European Maritime Safety Agency (EMSA) FIRESAFE final report concluded
that fires originating in the cargo spaces (90%) far outweighed ship equip-
ment (5.3%) and other causes (4.7%).49 The fleet at risk included in the ana-
lysis consisted of 490 ships selected based on criteria as agreed by EMSA. SP
Technical Research Institute of Sweden AB, Bureau Veritas, and Stena Rederi
AB participated in this study.
One of the most significant contributors to fire onboard Ro-Ro ships has
been the refrigerator units (33%) that keep containers cool, followed by
vehicle drive trains (24%) and truck cabs (17%). Conclusions determined
that in most cases, a fire starts as a small smoldering fire and grows very
quickly. Also, the development of fires is highly dependent on ventilation
and available fuel, and quick response to fire is the most important safety
precaution.
4.5.3 Fuel Spill

Transport Malta released its report on an accident onboard the Ro-Ro ship
Eurocargo Trieste when it experienced a fire in its engine room, most likely
caused by fuel spilling accidentally onto a hot surface.50 Once started, the fire
quickly spread due to the presence of combustible material, leaking heavy
fuel oil and lubricating oil around the engine and its bilges.
Cargo Sensors 65
Additional findings cited in the report included:
• The smoke detectors, although operational, failed to alert the crew of the
existence of a fire immediately as they probably had been silenced for a
short period of time while the vessel was in port.
• Combustible materials in the form of leaked fuel, leaked oil, braided PVC
pipes (to direct the leaks), plastic containers to collect drained oils, oil in
the bilges and the vicinity contributed to the propagation of the fire.
• It is highly likely that the leak in the CO2 system compromised its
effectiveness.
• Evidence indicated that the doors to the fuel oil modules and separator
rooms were open.
• The delay in stopping the port main engine is likely to have contributed
to the fire taking hold.
These findings indicate multiple contributory causes, many of which

were human in nature such as using plastic containers to collect leaking oil
and silencing smoke detectors, thereby exhibiting neglect of proper safety
procedures.
Using sensors to enhance safety and situational awareness can only assist
seafarers when the sensor indications are observed and not ignored, as in
the case of silencing smoke detectors. However, it appears the leak of CO2
activated the smoke detectors and compromised the CO2 fire suppression
system, making it less effective. This sequence of events was likely confusing,
and mistakes may unknowingly have been made. Had appropriate sensors
existed within the CO2 system, this leak may have been detected before the
system needed to suppress the fire.
However, one of the most telling of the findings was:
The decision-making process of the master would have been very complex,
involving at least cues (possibly conflicting), technological data, informa-
tion from fellow crew members, interpretation of that data and a decision
to act, either in one way or another.
This finding supports an argument for using automation and smart sensors
to aid decision-making.
4.5.4 Improper Stability

The capsize of the car carrier Golden Ray in 2019 while departing the port
of Brunswick, Georgia, USA, has been attributed to incorrect calculations
about the vessel’s stability, according to the U.S. National Transportation
Safety Board (NTSB).51 In the report, the chief officer was cited as making
an error entering ballast quantities into the stability calculation program.
66 Cargo Sensors
This led to his incorrect determination of the vessel’s stability and resulted
in Golden Ray having an insufficient righting arm to counteract the forces
developed during a turn. The ship caught fire and was ultimately scrapped
in place at the accident site. The NTSB also found that two watertight doors
had been left open for almost 2 hours before the accident and that no one on
the bridge ensured the doors were closed before departure. Nothing appeared
amiss until the ship began to heel rapidly to port during a 68-degree turn to
starboard.
Although human error is the cause of this accident, there were no backup
procedures in place to identify and correct the error. Smart sensors trained
in the behavior patterns of the vessel may have detected subtle differences
between actual vessel weight, mass, momentum and acceleration forces from
nominal before initiating the turn —but maybe not. Indeed, had simple sen-
sors been in place to detect the open doors, the vessel may not have flooded
nearly as rapidly.
4.6 Other Ship Types

Many other types of cargos are carried onboard numerous ships of diverse
designs. Their sensor requirements to maintain ship stability and detect other
threats to the cargo or ship are common with the previously discussed vessel
types. Where unique sensor requirements exist due to the nature of the cargo
and resulting ship features, solutions are often devised and invented based on
specific needs and applications. However, the same basic principles apply to
all vessels.
References
1 Zahra Ahmed, Top 20 World’s Largest Container Ships in 2023. Marine
Insight. 11 April 2023. www.marineinsight.com/know-more/top-10-worlds-
largest-container-ships-in-2019/
2 Rishab Joshi. Top 5 Biggest Bulk Carriers in the World. Marine Insight. 6 June
2022. www.marineinsight.com/types-of-ships/biggest-bulk-carriers/
3 TMP Staff. Top 10 Largest Tanker Ships in the World. The Maritime Post.
3 February 2022. https://themaritimepost.com/2022/02/top-10-largest-tan
ker-ships-in-the-world/
4 Saurabh Sinha. Top 10 Biggest Ro-Ro Ships in the World. 14 March 2022. www.
marineinsight.com/types-of-ships/top-10-biggest-roro-ships-in-the-world/
5 Breakbulk’s Top Ten. The Maritime Executive. 10 April 2019. www.maritime-
executive.com/article/breakbulk-s-top-ten
6 ECIB. ONE Apus Container Loss: Then and Now. Expeditors Cargo Insurance
Brokers 28 January 2021. https://insider.ecibglobal.com/blogs/one-apus-contai
ner-loss-then-and-now
Cargo Sensors 67
7 Andrew Gray. Tackling the Scourge of Container Ship Fires. 13 November,

2020. Safety4Sea. https://safety4sea.com/tackling-the-scourge-of-container-
ship-fires/
8 Mike Schuler. ZIM Kingston Cargo Fire Stabilized and Ship Held Overnight.
25 October 2021. https://gcaptain.com/zim-kingston-cargo-fire-stabilized-
canadian-coast-guard/
9 ABS. Guide for Fire-Fighting Systems for Cargo Areas of Container Carriers.
May 2022. American Bureau of Shipping 7.7 FBC Notation (1 October
2019), p. 13.
10 International Convention for Safe Containers (CSC), 1972. International
Maritime Organization. Maritime Safety Committee Resolution MSC.355(92).
1 July 2014.
11 George V. Hulme. RFID Helps to Track Cargo Containers. Information Week.
6 April 2005. www.informationweek.com/software/rfid-helps-to-track-cargo-
containers
12 Martin LeMonica. IBM and Maresk to Equip Cargo with Sensors. 20 September
2005. CNET.com. www.cnet.com/culture/ibm-maersk-to-equip-ship-cargo-
with-sensors/
13 Allianz. Industry’s Struggle with Container Ship Fires Continues. August 2018.
Safety and Shipping Review. www.agcs.allianz.com/news-and-insights/expert-
risk-articles/container-ship-fires.html
14 ING, 2021. Sensor Ships: Why Smart Containers Are the Future of Shipping.
Environmental Finance. 1 November 2021. www.environmental-finance.com/
content/focus/sustainable-transformation-progress-not-perfection/articles/sen
sor-ships-why-smart-containers-are-the-future-of-shipping.html
15 Allianz. Industry’s Struggle with Container Ship Fires Continues. Allianz Global
Corporate & Specialty. August 2018. www.agcs.allianz.com/news-and-insig
hts/expert-risk-articles/container-ship-fires.html
16 Capt. Mark Wood. Parametric Rolling: What It Is and How to Avoid It?
Maritime Executive. 28 April 2022. https://themaritimepost.com/2022/04/par
ametric-rolling-what-it-is-and-how-to-avoid-it/
17 Circular. 1228 –Revised Guidance to the Master for Avoiding Dangerous
Situations in Adverse Weather and Sea Conditions. 11 January 2007. Annex,
Revised Guidance to the Master for Avoiding Dangerous Situations in Adverse
Weather and Sea Conditions. International Maritime Organization (IMO).
18 Karan C. What Is Parametric Rolling in Container Ships? Marine Safety. 6
June 2021. www.marineinsight.com/marine-safety/what-is-parametric-rolling-
in-container-ships/
19 Datasheet. Inclinometer Type KPM 100. Kjærulf Pedersen a/s. www.kp-as.com/
wp-content/uploads/KPM100-datasheet.pdf
20 Sensor Company Builds System to Warn of Parametric Rolling Danger. The
Maritime Executive. 5 September 2022. https://maritime-executive.com/article/
sensor-company-builds-system-to-warn-of-parametric-rolling-danger
21 Intercargo. Bulk Carrier Casualty Report, 2019. International Association of
Cargo Ship Owners. 1 May 2020. www.intercargo.org/bulk-carrier-casualty-
report-2019/
68 Cargo Sensors
22 Paul Bartlett. Cargo Liquefaction Greatest Cause of Deaths in Dry Bulk Sector.
31 July 2023. www.seatrade-maritime.com/dry-bulk/cargo-liquefaction-great
est-cause-deaths-dry-bulk-sector
23 Susan Gourvenec. The Cargo Ships That ‘Liquefy’. BBC. 16 September 2018.
www.bbc.com/future/article/20180905-the-cargo-ships-that-liquefy
24 North P&I Club. The Dangers of Cargo Liquefaction in a Nutshell. As
described in ShipNerd. 30 May 2022. www.shipnerdnews.com/the-dangers-of-
cargo-liquefaction-in-a-nutshell/
25 IMO: Bauxite Liquefaction Sank Bulk Jupiter. Offshore Energy Today. 21
September 2015. www.offshore-energy.biz/imo-bauxite-liquefaction-sank-
bulk-jupiter/
26 Gard P&I Club. Cargo Liquefaction. 27 September 2023. www.gard.no/web/
topics/article/20651747/cargo-liquefaction
27 Mike Schuler. Gard Warns of Liquefaction Risks with Unlisted Cargoes. gCap-
tain. 13 July 2023. https://gcaptain.com/gard-warns-of-liquefaction-risks-of-
unlisted-cargoes/
28 TMP Staff. Explosion Aboard Chinese Bulker CSSC Cape Town in Gibraltar
Bay Leaves Four Injured. The Maritime Post. 22 February 2021. https://them
aritimepost.com/2021/02/explosion-aboard-chinese-bulker-cssc-cape-town-in-
gibraltar-bay-leaves-four-injured/
29 Clete R. Stephan, P.E. Coal Dust Explosion Hazards. Mine Safety and Health
Administration. Pittsburgh, PA, USA. https://ncsp.tamu.edu/reports/MSHA/
coaldust.pdf
30 Jiange Chen, Dewen Li, Guoqing Liu, Yanzhu Li, Anran Zhang, Siyuan Lu,
and Mi Zhou. Development of a Coal Dust Concentration Sensor Based on
the Electrostatic Induction Method. American Chemical Society. ACS Omega
2023, 8, 14, 13059–13067. 28 March 2023. https://doi.org/10.1021/acsom
ega.3c00319
31 Navy Releases Extensive Bonhomme Richard Fire Report, Major Fires Review.
Vice Chief of Naval Operations Public Affairs. 20 October 2021. www.navy.
mil/Press-Offi ce/News-Stories/Article/2816283/navy-releases-extensive-bonho
mme-richard-fire-report-major-fires-review/
32 USNI. Navy Investigation into USS Bonhomme Richard Fire, Major Fires
Review. US Naval Institute. 20 October 2021. https://news.usni.org/2021/10/
20/navy-investigation-into-uss-bonhomme-richard-fire-major-fires-review
33 UK Marine Accident Investigation Branch (MAIB) inquiry into the 2019 blast
aboard the chemical tanker Stolt Groenland. Report 9/2021. Southampton.
Synopsis. July 2021. https://assets.publishing.service.gov.uk/media/60f93e2cd
3bf7f044c51590b/2021-09-StoltGroenland-Report.pdf
34 Heather Chen, Irene Nasser and Teele Rebane. Oil Tanker Catches Fire Off
Malaysian Coast, Three Crew Missing. CNN. Mon, 1 May 2023. www.cnn.
com/2023/05/01/asia/malaysia-coast-oil-tanker-fire-rescue-intl-hnk/index.html
35 Alex Longley and Yongchang Chin. An Oil Tanker Ablaze in the South China
Sea Is a Global Problem. Bloomberg. 6 May 2023. www.bloomberg.com/news/
articles/2023-05-07/an-oil-tanker-ablaze-in-the-south-china-sea-is-a-global-
problem
Cargo Sensors 69
36 Unyime O. Akpan, T.S. Koko, B. Ayyub and T.E. Dunbar. (2002) Risk assess-
ment of aging ship hull structures in the presence of corrosion and fatigue.
Marine Structures, 15(3), 211–231. ISSN 0951-8339. https://doi.org/10.1016/
S0951-8339(01)00030-2
37 TBS. Explosion on Oil Tanker in Jhalakathi Leaves Five Burnt, Four Missing.
The Business Standard. 1 July 2023. www.tbsnews.net/bangladesh/oil-tanker-
catches-fire-after-explosion-jhalakathi-5-burnt-4-missing-658470
38 Marine Insight. Oil Tanker Explodes in Thai Waters. Marine Insight News
Network. 18 January 2023. www.marineinsight.com/videos/video-oil-tanker-
explodes-in-thai-waters-blows-up-workers-leg-500-m-away/
39 Franz Evegren. Fire in Vehicles Onboard Ships. RISE Fire Safe Transport.
Research Institutes of Sweden. www.ri.se/sites/default/files/2020-12/FIVE_Fire
in vehicles onboard ships_Evegren_WEB_201215.pdf
40 Safetytech. Startup Deploys Wireless Sensors across Ship’s Cargo Hold to
Predict Fire. Safetytech Accelerator, London, UK. 3 March 2020. https://safety
techaccelerator.org/case-studies/safetytech-startup-deploys-wireless-sensors-acr
oss-ships-cargo-hold-to-predict-fire/
41 Mikhail Voytenko. Grimaldi’s Ro-Ro Major Fire at Newark, 2 Firefighters
Died. FleetMon. 6 July 2023. www.fleetmon.com/maritime-news/2023/42341/
grimaldis-ro-ro-major-fire-newark-2-firefi ghters-d/
42 NTSB. Failure to Properly Disconnect and Secure Vehicle Batteries Led to Fire
Aboard Vehicle Carrier Höegh Xiamen. US National Transportation Safety
Board. 16 December 2021. www.ntsb.gov/news/press-releases/Pages/mr20211
216.aspx
43 FreightWaves Staff. Grimaldi Confirm Grande America Fire Started in
Container Cargo. FreightWaves. www.freightwaves.com/news/maritime/grima
ldi-confirms-grande-america-fire-started-in-cargo-container-ezkx7-g33nr-
gn3cf
44 Legislative Assessment for Safety Hazards of Fire and Innovations in Ro-Ro
Ship Environment (Lash Fire). www.lashfire.eu
45 Chris Teague. How Much Should You Worry About EV Fires? Autoweek
News. October 2022. www.autoweek.com/news/a38225037/how-much-you-
should-worry-about-ev-fires/
46 Traffic Safety Facts Annual Report Tables. National Highway Traffic Safety
Administration. U.S. Department of Transportation. https://cdan.dot.gov/
tsftables/tsfar.htm#
47 NTSB. Safety Risks to Emergency Responders from Lithium-Ion Battery Fires
in Electric Vehicles. U.S. National Transportation Safety Board. www.ntsb.gov/
safety/safety-studies/Pages/HWY19SP002.aspx
48 National Transportation Safety Board. Fire aboard Tank Vessel S-Trust, Report
MIR-23-23. 25 October 2023. www.ntsb.gov/investigations/AccidentReports/
Reports/MIR2323.pdf
49 J. Wikman, F. Evegren, M. Rahm, J. Leroux, A. Bruillard., M. Kjellberg, L.
Gustin, and F. Efraimsson. FIRESAFE: Study Investigating Cost Effective
Measures for Reducing the Risk from Fires on Ro- Ro Passenger Ships.
Lisbon: European Maritime Safety Agency. 2016.
70 Cargo Sensors
50 Safety Investigation into the Engine-Room Fire on Board the Maltese Registered
Ro-Ro Cargo Vessel Eurocargo Trieste. Transport Malta. Marine Safety
Investigation Unit. Marine Safety Investigation Report No. 21/2020. Final.
www.iims.org.uk/wp-content/uploads/2020/11/Transport-Malta-MV-Euroca
rgo-Trieste-Safety-Investigation-Report.pdf
51 NTSB. NTSB Determines Inaccurate Stability Calculations Caused Capsizing
of Vehicle Carrier Golden Ray. U.S. National Transportation Safety Board. 14
September 2021. www.ntsb.gov/news/press-releases/Pages/NR20210914b.aspx
Chapter 5
Crewmember and Passenger

Sensors
Probably one of the most significant areas in which advances in sensor tech-
nology have taken place is the result of increasing interest by the shipping
industry in the instrumentation of crew and passenger spaces with sensors
that not only identify personnel but also track and monitor their movements
and well-being while in port and throughout their voyages. This is being
accomplished due in part to digitalization in maritime shipping through Port
Community Systems (PCSs) and other means to connect different govern-
ment, public and private organizations for the secure and intelligent exchange
of information.
5.1 A Burst of Non-t raditional Ship Sensors

The variety and diverse requirements of sensors are as wide-ranging as the
passengers and crewmembers that ships carry, the personnel who maintain
shipboard systems and others who may or may not have legitimate business
onboard. Company- provided crewmember credentials have been supple-
mented with government-issued identification cards with biometric features
allowing access to port and ship infrastructure. Once aboard, additional sen-
sors are used that further verify identity. Likewise, passenger credentialling
has changed and now contains vastly greater details to enhance identifica-
tion, provide access to services and products, and reduce the risk of fraud and
unauthorized access to vessels. Long past are the days when guests could be
entertained in passenger staterooms before the ship’s departure.
Cruise ships integrate fire and smoke detectors into voyager cabins that
can now be supplemented with floor, wall and shower sensors to detect occu-
pancy, movement, and even the occurrence of accidents such as slips and
falls by the elderly. On some vessels, facial recognition is used to open cabin
doors, and wearable devices containing passenger- unique identity infor-
mation are read by thousands of sensors on the ship and in port to enable
hospitality and victualling services and speed the boarding and debarkation
processes. These developments are also being adapted to meet the needs of
merchant mariners and ship crews.
Doi: 10.1201/9781003377900-5
72 Crewmember and Passenger Sensors
Data from personnel sensors can be interpreted in combination with ship

engineering and navigation sensor data to form a more complete and com-
prehensive picture to enhance situational awareness and overall vessel and
human safety. This chapter describes many of the different types of sensors
that can be found on ships and their methods to monitor human activity.
Note that some of these methods are complementary in nature and cap-
ability, and using more than a few different sensor types on the same ship is
highly unlikely.
5.2 Identification Cards

Many countries have developed their own versions of port security card
programs of standardized identification systems for workers in sensitive
transportation-related industries to help ensure that only authorized indi-
viduals with proper credentials can enter restricted areas and carry out their
work duties. The sensors and electronic devices that implement these systems
and the credentials they examine are identified by different names such as
Transportation Worker Identification Card (TWIC), Marine Transportation
Security Clearance (MTSC), Maritime Security Identification Card (MSIC),
Port Security Identification Card (PSIC) and Maritime Security (MARSEC)
Card. However, their purpose remains the same: to enhance security while
allowing access to certain restricted areas within ports and harbors, including
the vessels within them. The information contained on and encoded in iden-
tification (ID) cards generally include the cardholder’s full name, their color
photograph, date of birth, card expiration date, a unique serial number, card
issue date and various security features such as holograms, watermarks and
encrypted information containing biometric data including fingerprints and
other embedded elements to prevent counterfeiting and ensure authenticity.
Passengers on cruise ships are also provided with identification cards
to verify individual personal identities when embarking, disembarking, or
reboarding the ship at different ports of call. These cards allow access to
the vessel and are used to open their cabin and stateroom doors and to gain
access to the many services and facilities on the ship. Specific information
contained on and within the ID card vary depending on the cruise line and
their systems. In general, it contains the passenger’s name and photograph,
assigned cabin or stateroom number, cruise ship name or logo and expiration
date corresponding to the duration of the voyage. A unique passenger identi-
fication number is also assigned, reflecting booking information and onboard
accounts to enable purchases such as drinks, dining, entertainment, ship and
shore tours, spa services and merchandise. Charges are recorded against the
account registered to the card, with the total bill settled by the passenger
at the end of the cruise. The card also assists in accounting for passengers
during emergency procedures. Large numbers of ID card sensors can exist
on individual ships, extending to literally thousands on cruise ships that can
Crewmember and Passenger Sensors 73
track movements and activities throughout the ship on a continuous, round-

the-clock basis over spans of days, weeks and even months. Radio Frequency
Identification (RFID) technology may also be integrated into ID cards to
enable data transfer by passing close to a sensor without requiring contact.
While professional crewmembers, maintenance workers and passengers
aboard most vessels must comply with government-mandated identification
card requirements, notable exceptions exist on ferries and other relatively
short-term transports where such practices are not logistically justified or
economically feasible. However, in many cases, the vehicle driver and indi-
vidual walk-on passengers may be required to show government-issued iden-
tification before boarding.
5.3 Closed Circuit Television (CCTV)

Video monitoring has been prominent in previous chapters for surveillance
and maintaining security and safety in the engine room, engineering spaces,
gangways and other sensitive areas including various entrances, exits, corri-
dors, decks, machinery and storage areas. For navigation, video cameras are
used for collision avoidance, aid to navigation (AtoN) detection, visual pilot-
ing and maneuvering while docking. They are also used for monitoring cargo
spaces, verifying hatch, vent and door closure; and maintaining security for
the bridge, the entire ship and its perimeter. It is, therefore, not surprising
that video monitoring’s greatest role is observing and tracking human activity
throughout the ship.
Merchant ships with 20 or so crewmembers are dwarfed compared to the
gigantic cruise ships currently being placed into service, capable of carry-
ing over 8,000 crew and passengers. These ships feature 20 or more decks
containing many restaurants and lounges, retail stores, child care facilities,
swimming pools, rock climbing walls, roller coasters, waterparks, casinos,
entertainment forums and other amenities; in addition to machinery spaces,
ship’s stores, food preparation; waste treatment, disposal and refuse hand-
ling; and other functions and services necessary to accommodate the needs
of a small city.1
Traditional CCTV systems help monitor for suspicious activity such as
people loitering, to deter crime and to investigate accidents or disputes
by enabling the analysis of the sequence of events, identifying individuals
involved and determining the cause of an incident. It can also be used to
monitor for safety hazards such as people in unauthorized and high-risk
areas like the deck during a storm and people not following proper safety
procedures related to the wearing and use of personal protective equipment
(PPE) and safety equipment, hard hats, or life jackets.2,3 CCTV can also be
instrumental in demonstrating compliance with and detecting violations of
environmental regulations, monitoring lifeboat and fire drills, and for train-
ing purposes. However, given the large numbers of people on cruise ships, one
of its most important uses is the monitoring of crowded areas like swimming
pool decks, dining halls, theaters, or entertainment venues to observe crowd
density, detect potential issues before they unfold or help stem their effects,
and ensure the smooth flow of passengers, especially during emergencies or
evacuation. Many such cameras also have embedded microphones capable of
recording human conversations and other sounds in the immediate vicinity
and allowing two-way conversations to be carried on through the camera.
Introducing smart video cameras into the ship environment with motion
detection and embedded intelligence through machine learning and deep-
learning AI provides even more capabilities than traditional video cameras.4
These capabilities include processing images using facial and object recog-
nition techniques to identify specific people, objects and human activities
such as walking, talking and running. Video cameras incorporating thermal
imaging capabilities can also distinguish healthy crewmembers and passen-
gers from those exhibiting fevers and possibly suffering from communicable
diseases. Video footage may also be stored in the Cloud, allowing company
officials and other users to access it from anywhere.
5.4 Facial Recognition

Facial recognition is a method whereby software processes imagery to iden-
tify a human face through biometrics and maps facial features from a photo-
graph or video for comparison with a database of known faces to find a
match.5 The primary use on ships involves real-time identity verification by
comparing the photograph on the crewmember or passenger identity card
with an image retrieved from a live camera image database obtained while
boarding or disembarking the vessel. However, many other uses onboard
ships and in restricted and secure areas of ports and harbors are possible
using this technology.
Facial recognition involves establishing a baseline image using a photo-
graph from which facial features are extracted, including eye position and
shape of the nose, mouth, jawline and other facial landmarks converted
into a digital mathematical representation.6 Comparisons are made against
a database of face image representations obtained from known individuals.
A matching process driven by one or a combination of statistical analysis,
machine learning, and deep-learning AI techniques is accomplished to a cer-
tain threshold of similarity whereby an identity is considered verified as a
match, or not verified. Notification to an operator or computer-based system
is then generated for further action in providing access, completing a trans-
action, or other processes.
Facial recognition is also used to enhance security by constantly monitor-
ing individual people within the ship’s spaces to rapidly determine whether
they are authorized to be in a particular location or not. Should unauthor-
ized personnel be detected, such as a passenger entering crew-only spaces, an
immediate alarm can be raised identifying the intruder and location at which
the intrusion took place so that corrective action may be taken. Further use
can be made to track movement and activity around a ship to identify loca-
tions visited and the times visits occurred, which is especially useful for deter-
mining the popularity of cruise ship venues and for tracking an individual’s
movements while investigating missing person reports on ships.
The accuracy of facial recognition has improved dramatically in recent
years. According to data from the U.S. National Institute of Standards and
Technology (NIST), the top 150 algorithms are over 99% accurate across
black male, white male, black female and white female demographics.7 For
the top 20 algorithms, the accuracy of the highest-performing demographic
versus the lowest varies between 99.7% and 99.8%. Unexpectedly, white
male is the lowest-performing of the four demographic groups for the top 20
algorithms. For 17 of these algorithms, accuracy for a white female, black
male and black female are nearly identical at 99.8%, while they are least
accurate for the white male demographic at 99.7%. Detailed statistics for
other demographics were not easily available. Facial recognition may be sty-
mied by face masks and plastic surgery. However, this technology appears
resilient and can still discern unique patterns sufficient to positively identify
people using partial patterns.
5.5 Radio Frequency Identification (RFID)

RFID technology can be incorporated into individual crewmember and pas-
senger ID cards and other wearable devices to supplement printed data as a
further precaution to enhance physical and cybersecurity and to help prevent
fraud and ID card counterfeiting.8 This technology operates by storing, and
hopefully encrypting, not more than a few thousand bytes of data containing
identification numbers, personal information, and access control data and
other information within an integrated circuit (IC, or chip) residing in the
ID card along with a small antenna that can receive queries and respond by
transmitting the data it contains. When an ID card is brought near the RFID
reader the reader transmits electromagnetic field emissions that induce a cur-
rent in a passive antenna providing power to the ID card chip, which then
replies by sending the encoded data back to the reader.
RFID technology supports various uses, including granting or denying
access to restricted areas, tracking crewmember and passenger movements
throughout the ship, and accounting for working hours. In addition, RFID
can be used to track the movement of inventory and the location of assets on
the ship.
ID cards with embedded RFID technology offer several advantages over
traditional identification cards since they are more difficult to counterfeit and
easier to use and can be used for a wider variety of applications than trad-
itional ID cards. Disadvantages include their ability to be hacked to retrieve
the information they contain, but this vulnerability can be reduced through
data encryption. They can also be more expensive than traditional ID cards.
RFID operating frequencies cover three different bands: Low Frequencies
(LF), ranging from 30 to 300 kHz; High Frequency (HF), ranging from 3 to 30
MHz, with 13.56 MHz as the dominant frequency; and Ultrahigh Frequency
(UHF), ranging from 300 MHz to 3 GHz.9 The transmission range for pas-
sive devices is typically one to five meters or more. Actual ranges will vary
depending on the specific tag, reader and environment.
The cost of passive RFID technology continues to decrease, and their
utility is increasing with greater memory capacity and two-way read/write
capabilities, tending more toward multisurface UHF systems. One study in
the proper application of this technology demonstrated the test results of
various multisurface UHF systems from different manufacturers for their
readability under varying conditions, such as orientation with respect to the
reader, distance from the reader and materials used for embedding them.10
Of ten different products tested, none scored 90% or higher for either read-
ability or precision, so there is ample opportunity to continue enhancing this
technology. Antenna orientation and distance to the reader appear to be the
greatest contributors to error and inaccuracy. However, some tags seemed to
place well on every metric given.
During the period between 2010 and 2020, many enhancements to RFID
technology have taken place to improve their performance and accuracy.11
This includes increased reader sensitivity and read rate, increased transmit
power to enhance range, improved processing power with some readers hav-
ing built-in microprocessors, better antenna designs and additional connect-
ivity options for transferring data.
5.6 Eye Scanning

Optical scanners developed specially to take an image of the eye’s retina and
the iris can be found on ships to grant or deny access to secure areas. Neither
glasses, contact lenses, nor changes in ambient light conditions appear to
affect the accuracy or viability of these methods. Scanning of the retina,
which is the light-sensitive tissue at the back of the eye, is a biometric tech-
nique similar to facial recognition that uses the unique patterns formed by
the blood vessels to identify them. The retina and the patterns of its blood
vessels are unique to each individual, and such that the patterns they form
are often so complex that they are different even between identical twins.12
The process used involves a person looking into a scanner that emits a
beam of low-intensity infrared light into the user’s eye and traces a circular
path on the retina. A high-resolution camera records the reflected light to
create an image that is then processed to form a digital mathematical model
representing a unique identifier for that individual. Comparisons are made
against a database of retina models obtained from known individuals using
a matching process driven by one or a combination of statistical analysis,

machine learning, and deep-learning AI techniques to a certain threshold of
similarity whereby an identity is considered verified as a match or not verified.
A study by Bayometric, a global supplier of biometric security systems,
found that retinal scans had no false matches while comparing over two mil-
lion retinal maps. In contrast, fingerprint scanners had more false matches
and a lower success rate overall.13
Scientists have conducted several studies showing that the human retina
can change over time for various reasons, including disease and injury.14
A common cause of change can be attributed to type 1 and 2 diabetes (dia-
betic retinopathy) that can damage and burst blood vessels of the retina and
thereby invalidate previously acquired retina scan models. Macular degen-
eration may also cause similar effects in the retina. Researchers have also
used human stem cells to grow new retinal cells to replace those destroyed
by other medical conditions, which can also change the characteristics of the
retina.15
Another type of identity verification uses optical scans of the eye’s iris,
which is the colored part of the eye that contains muscles that control the
pupil or the small black opening that lets light into the eye. These same stud-
ies have also found that a healthy iris should remain unchanged throughout
a person’s lifetime. They claim it is impossible to find two completely iden-
tical patterns, even in twins. They also note that operations performed on the
eyes, removal of cataracts, or implantation of corneal implants do not change
the iris’s characteristics. In addition, a blind person can also be identified
using the iris of their eye. Still, injury from burns, punctures, or strikes with
a blunt object and some medical conditions may change the iris.
Iris scanning uses a device very similar in function to retina scanners but
targets different parts of the eye. The same basic techniques are used where
low-intensity infrared light is aimed into the user’s eye and traced across the
iris. A high-resolution camera records the reflected light to create an image of
the iris that is processed to form a digital mathematical model that represents
a unique identifier for that individual.
5.7 Fingerprints
Fingerprints are also used to verify identity because they are unique to each
individual, and this method is one of the earliest forms of biometric sens-
ing for identifying people. No two people have the same fingerprints, not
even identical twins. This makes fingerprints a very reliable way to identify
someone.
Multiple types of sensors may be used for fingerprint identification,
including optical, capacitive and ultrasonic scanners. Once scanned, a finger-
print is converted into a digital mathematical model representing a unique
identifier for that individual and then stored in a database. Ultrasonic sensors
create a 3D fingerprint image that is more secure and accurate than trad-
itional capacitive fingerprint readers. Fingerprint identification uses pattern
recognition techniques similar to facial recognition and eye scanning.
The U.S. National Institute of Standards and Technology (NIST) has
evaluated several commercially available contactless fingerprint scanning
technologies, allowing users to compare their performance to conventional
devices that require physical contact between a person’s fingers and the scan-
ner.16,17 The study results show devices requiring physical contact remain
superior to contactless technology at matching scanned prints to images in a
database. However, when contactless devices scan multiple fingers on a hand,
it improves their performance. Contactless devices that scanned multiple fin-
gers also seldom made “false positive” errors that incorrectly matched one
person’s print with another’s record.
Other factors that may make fingerprint matching less reliable include dam-
age to the finger caused by cuts, burns, or other injuries that have occurred
since the baseline database model was established, which may reduce or
invalidate potential matches. This may also occur should dirt, grease, or other
materials obscure the fingerprint; if it is smudged, blurry, or distorted in any
way, or if combined with or overlaps fingerprints from other individuals.
5.8 Body Temperature Sensors

Earlier in this chapter, it was mentioned that video cameras incorporating
thermal imaging capabilities can distinguish healthy crewmembers and pas-
sengers from those exhibiting fevers. Also, present on ships are contactless
infrared temperature sensors that, when aimed at a person’s forehead, can
instantly measure a person’s body temperature. These are generally operated
by individuals present on a gangway, in a door opening, or in other locations
that can restrict personal movement to one individual at a time.
At present, temperature measurements of individuals are accomplished
with little or no notice except when it is obvious that a person aims a sensing
device straight at someone. Large-scale, real-time measurements of individ-
uals’ temperatures for large groups are only possible using wearable tempera-
ture sensors such as those within fitness monitoring and medical devices that
can communicate using Wi-Fi or Bluetooth wireless connections. Except pos-
sibly for ships with few crewmembers, such monitoring is unlikely to become
widespread in the foreseeable future due to cost, logistics and privacy issues.
5.9 Space Occupancy Detection

There are many occasions where it is advantageous to determine whether
a person or people occupy a room or other space on a ship. This includes
the security of physical spaces, detection of motion within secure areas and
cargo holds where people are generally not present, or in a passenger ship
stateroom to determine opportune times for housekeeping tasks based on a

person’s habits. Another use is for energy efficiency to turn off room lights,
reduce heating temperatures, or increase cooling temperatures when a space
is unoccupied.
CCTV is the most accurate sensor for person detection, which was dis-
cussed previously in this chapter. However, other sensors are less intrusive
and often just as effective while maintaining privacy. One of the most com-
mon is passive infrared (PIR) sensors that detect the heat emitted by people
and other warm-blooded animals. In rooms with an abundance of glass or
metal may be found where PIR sensors may be less effective, an alternative
is the use of microwave-emitting sensors that measure reflected signals to
detect the presence of people. Still, another alternative sensor uses ultrasound
signals to detect the presence of people, especially in areas such as rooms sus-
ceptible to dust or smoke.
5.10 Accident (Slip and Fall) Detection

The maritime environment is a hazardous one at best, where slippery surfaces
and rolling motions due to changing sea states can cause the most experi-
enced mariners, let alone passengers, to slip and fall that may result in severe
injury or even death.18 Of those 65 and older, who comprise many passengers
on ships, more than one in four people fall each year.19 Although many such
occurrences can be prevented by taking proper precautions beforehand, there
is no sure prevention method.
The most useful sensor for detecting a slip or fall is a video camera that
happens to be aimed at the victim and is being monitored by an operator
when an actual fall occurs. However, the likelihood of such combinations of
a camera being in the right place and looking at the right time; and a person
viewing the camera at the precise time to observe the fall, understand the
significance of what they are looking at, and take appropriate action is min-
imal. Smart cameras can be trained to detect falls and report independent of
human activity through proper training using machine learning and deep-
learning AI. Still, such applications are relatively new and untested.
Advances in the senior citizen healthcare industry have led to the develop-
ment of wearable technology that can assist people in alerting others that a
fall has occurred by the victim pressing a button and communicating directly
to a response center. An alert can also be accomplished through automated
alarm generation without any assistance from the victim if incapacitated.
Such capabilities can also be afforded to seafarers through wearable devices
that monitor for extreme conditions related to velocity, acceleration and
pressure outside of nominal values that may be consistent with a fall.
Types of sensors that can be used to detect falls include accelerometers that
measure the sudden change in acceleration, gyroscopes that measure the sud-
den change in angular velocity, and pressure sensors that measure a sudden
increase in pressure that occurs when a person falls. Also included are inertial
measurement units (IMUs) that use sensor fusion by combining measure-
ments made by accelerometers, gyroscopes and other sensors to determine
the overall motion of a moving object with behavioral characteristics that
may indicate a slip or fall has occurred.
5.11 Man Overboard Detection

Many devices are available to assist in the detection of a man overboard
(MOB) event and recovery of individual crew members or passengers using
a transmitter attached to the individual in the water. These maritime sur-
vivor locating devices can be used by both crewmembers and passengers and
include two basic types: A personal Automatic Identification System (AIS)
beacon (PAB) and a personal locator beacon (PLB).20
The PAB transmits the wearer’s Global Navigation Satellite System (GNSS)
position to any vessels in the area up to a 5-mile range via VHF radio Digital
Select Calling (DSC) alarm on 161.975 MHz to AIS receivers and AIS-enabled
plotters. However, the PLB is a portable device attached to the wearer that
sends a distress signal on 406 MHz to the COSPAS-SARSAT satellite net-
work for use by search and rescue (SAR) teams to locate the person in dis-
tress. A PLB does not give a localized AIS signal that a vessel could receive.
The best devices include both AIS and PLB capabilities. Both types of
devices designed for maritime use can also feature an integrated high-intensity
strobe light to assist rescuers in locating the wearer in the water, even in low
light conditions.
Cruise ships with many thousands of crewmembers and passengers would
find the cost of equipping each with PAB or PLB devices prohibitive, and
doing so would likely result in hundreds of false alarms as untrained pas-
sengers could inadvertently trigger these devices. However, the U.S. Cruise
Vessel Security and Safety Act of 2010 requires passenger vessels operating
in the United States to “integrate technology that can be used for capturing
images of passengers or detecting passengers who have fallen overboard, to
the extent that such technology is available”.21 The International Standards
Organization (ISO) 2020 subsequently developed ISO 21195:2020 that spe-
cifies technical requirements for systems designed to detect a person who
has gone overboard from a ship which specifically excludes MOB detection
systems that require the passengers or crew to wear or carry a device such
as a PAB or PLB to trigger a MOB event.22 Several sensor technologies are
envisioned to implement ISO 21195:2020 including Radar, infrared, and
video imaging, as well as using a combination of technologies to improve the
accuracy of MOB detection.23 Specific technologies for useful MOB detection
depend on several factors, including ship size, operating environment and
budget. However, all MOB detection systems must meet the requirements of
ISO 21195:2020 to be used on ships.
References
1 Molly Bohannon. The ‘Icon of the Seas’ Will Soon Be the World’s Largest
Cruise Ship. Forbes Magazine. 12 July 2023. www.forbes.com/sites/mollyb
ohannon/2023/07/12/the-icon-of-the-seas-will-soon-be-the-worlds-largest-cru
ise-ship---heres-how-much-it-costs-to-get-aboard/
2 The use of CCTV Cameras on Ships. Monarch Group. 27 November 2019.
www.monarchglobal.net/post/the-use-of-cctv-cameras-on-ships
3 Staff. More Shipowners Use Video Cameras for Security, Safety and Operations.
Professional Mariner. 27 May 2016. https://professionalmariner.com/more-shi
powners-use-video-cameras-for-security-safety-and-operations/
4 How Do Smart Cameras Work? Technology Org. 18 July 2019. www.technol
ogy.org/2019/07/18/how-do-smart-cameras-work/
5 What Is Facial Recognition? How Facial Recognition Works. Internet of
Things. Norton. 20 August 2021. https://us.norton.com/blog/iot/how-facial-
recognition-software-works
6 What Is Facial Recognition –The 2023 Ultimate Guide for Facial Recognition
Technology Guides & Tips. FaceMe. 14 March 2023. www.cyberlink.com/fac
eme/insights/articles/204/Facial-Recognition-at-the-Edge-The-Ultimate-Guide
7 Patrick Grother, Mei Ngan, Kayee Hanaoka, Joyce C. Yang, and Austin Hom.
Ongoing Face Recognition Vendor Test (FRVT). Information Technology
Laboratory, National Institute of Science and Technology (NIST), US
Department of Commerce. 16 June 2023. www.nist.gov/programs-projects/
face-recognition-vendor-test-frvt-ongoing
8 The Use of RFID for Human Identification. Draft Report. U.S. Department
of Homeland Security (DHS), Emerging Applications and Technology
Subcommittee. Version 1.0. www.dhs.gov/xlibrary/assets/privacy/privacy_ad
vcom_rpt_rfid_draft.pdf
9 ACD. Understanding RFID and RFID Operating Ranges. Advanced Controls
and Distribution. 27 March 2017. https://blog.acdist.com/understanding-rfid-
and-rfid-operating-ranges
10 Aldo Minardo, Joshua Bolton, Erick Jones, Raghavendra Kumar Punugu,
Ankan Addy, and Samuel Okate. Performance and benchmarking of multisur-
face UHF RFID tags for readability and reliability. Journal of Sensors. Hindawi.
05 September 2017. https://doi.org/10.1155/2017/3467593
11 Suzanne Smiley. RFID Failed You in the Past? It May Have Improved
More Than You Think. 19 May 2020. www.atlasrfi dstore.com/rfid-insider/
why-try-rfid-again/
12 Alexa Saul. Retinal Security Scans: How Accurate Are They? Arizona Retina
Project. 31 October 2018. https://azretina.sites.arizona.edu/index.php/
node/379
13 Mary Clark. Iris Recognition Scanners vs. Fingerprint Scanners: Compare and
Contrast. Bayometric. www.bayometric.com/iris-recognition-scanners-vs-fing
erprint-scanners/
14 What Are Iris and Retina Scanners, and How Do They Work? RecFaces. https://
recfaces.com/articles/iris-scanner
15 David Turbert. What Parts of the Eye Can Be Transplanted? 13 January 2022.
American Association of Ophthalmometry. www.aao.org/eye-health/treatme
nts/transplantation-eye
16 NIST Study Measures Performance Accuracy of Contactless Fingerprinting
Tech. 19 May 2020. www.nist.gov/news-events/news/2020/05/nist-study-
measures-performance-accuracy-contactless-fingerprinting-tech
17 John Libert, John Grantham, Bruce Bandini, Kenneth Ko, Shahram Orandi,
and Craig Watson. Interoperability Assessment 2019: Contactless-to-Contact
Fingerprint Capture. NISTIR 8307. National Institute of Standards and
Technology. https://doi.org/10.6028/NIST.IR.8307
18 Maritime Slip and Fall/Trip and Fall Injuries. Maritime Injury Guide. www.
maritimeinjuryguide.org/maritime-accidents-injuries/maritime-bodily-injuries/
slip-fall-trip-fall/
19 Facts About Falls. U.S. Centers for Disease Control. www.cdc.gov/falls/
facts.html
20 Fox Morgan. Best Personal Locator Beacons and AIS Units. Yachting World
Magazine. 18 July 2023. www.yachtingworld.com/yachts-and-gear/best-perso
nal-locator-beacons-and-ais-units-top-options-for-boating-137237
21 U.S. Cruise Vessel Security and Safety Act of 2010. Public Law 111–207. 27
July 2010, 124 Stat. 2243. 111th Congress. www.congress.gov/111/plaws/publ
207/PLAW-111publ207.pdf
22 ISO 21195:2020. Ships and Marine Technology –Systems for the Detection
of Persons While Going Overboard from Ships (Man Overboard Detection).
www.iso.org/standard/76051.html
23 Elliot Gardner. Setting International Standards for Man Overboard Systems.
Ship Technology. 27 May 2018. www.ship-technology.com/analysis/setting-
international-standards-man-overboard-systems/
Chapter 6
Artificial Intelligence in Sensor

Systems
Throughout the proceeding chapters the use of artificial intelligence (AI) has
featured prominently in discussions on smart sensors and sensor data analyt-
ics to detect objects, trends and events of human interest with respect to the
roles in which seafarers are engaged. This includes the jobs performed by
members of the engineering and deck departments in operating and guiding
the ship to ensure cargo is loaded, transported and delivered to its destination
safely and efficiently; and the hospitality department to safeguard the com-
fort and well-being of passengers. Critical to this concept is how the various
systems and sensors of the ship itself can aid and enhance seafarer situational
awareness to improve job and ship safety, performance and effectiveness. AI
can provide significant advantages towards achieving this goal, but the tech-
nology is not without controversy.
Some argue that AI is the panacea to overcoming human frailties by even-
tually replacing highly trained and experienced seafarers with automatons
that perform their jobs flawlessly, never get bored or need to take breaks or
sleep, can work around the clock and do not need to be paid.1,2,3 Supporting
this position are statistics that human error is estimated to be responsible for
between 76% and 94% of marine casualties.4,5 Others think that AI brings
the risk of humans losing control to the automaton within the machinery it
operates; that external forces can use AI to infiltrate, manipulate, disable and
even hijack shipboard systems; and that AI can become superintelligent and
capable of malicious behavior to cause harm to those onboard and imperil
the ship.6,7 The benefits of using AI to alert engineering and navigation
watchstanders to the development of significant events, conditions, trends
and situations can and have been repeatedly demonstrated in enhancing effi-
cient ship routing and fuel savings, collision avoidance, reducing workload,
predicting failures and maintenance actions, and detecting objects and people
in the water.8,9 However, the lack of a clear understanding as to exactly how
AI systems (especially deep-learning AI) achieve the conclusions they reach,
inherent biases in the methods and datasets used to train them, the lack of
metrics to verify their development processes and validate their performance,
Doi: 10.1201/9781003377900-6
84 Artificial Intelligence in Sensor Systems
and the uncanny ability of hackers to penetrate all levels of security precau-
tions lends credence to doomsayer theories.
Many definitions of the term “artificial intelligence” refer to the ability of
a machine to simulate, replicate, or even improve upon intelligent behavior
exhibited by various forms of life that range from fruit flies to human beings.
Since we often fail miserably in discerning how human beings or even insects
think and reach conclusions, the best we can do is to compare the results
achieved by an AI-based solution against human results for a given, spe-
cific and well-defined task. Apparently, AI can be very good at what it does.
Spectacular results have been achieved for many years now by AI gaming
applications in outwitting human opponents. ChatGPT has successfully
passed academic and qualification examinations for a variety of occupations
that include medicine, law, accounting and sommelier.10,11 Internet searches
thus far have not yet revealed published records on the success of AI in pass-
ing merchant mariner examinations.
Despite the successes of AI cited in the previous paragraph, there is no assur-
ance that such achievements can be translated from an academic environment
directly into practical, reliable and widespread use in the real maritime world
in much the same manner that a fresh maritime academy graduate is not
yet qualified nor sufficiently experienced to stand watch let alone command
a ship. However, much research and experimentation have advanced ship
autonomy and navigating unmanned voyages along predetermined routes.
As applied to ship sensors, there is no well-defined general methodology to
ensure AI-based software has been developed correctly or will properly and
adequately perform its intended function. Each situation is evaluated on its
own merits based on outcomes rather than general and repeatable scientific
methods. The question is, is that enough?
This chapter sheds light on how AI may be utilized in sensor-based ship-
board systems, what constitutes AI and how these technologies are developed.
Benefits of their use and precautions that should be taken when developing,
installing and implementing these systems are also discussed. Finally, asser-
tions are made on how these systems may affect seafarers in helping to per-
form their jobs and enhance safety, how crewmember jobs may change or
be eliminated and how seafarer training practices may better adapt to future
situations.
6.1 Smart Sensors

A revolution in engineering sensor design has emerged with AI providing
greater awareness of and insight into the properties of fluids, flows, and
materials; the dynamics and processes used in their handling and changing
characteristics at various states and points in time. Individual sensors can
monitor various ship components and machinery in real time, learn nominal
performance patterns and detect variations and anomalies in the data that
Artificial Intelligence in Sensor Systems 85
are not necessarily perceived nor detected by humans which can help pre-
dict equipment failures. A traditional sensor samples the physical environ-
ment and then passes the measurement to a system or operator for analysis
and action. However, a smart sensor samples the physical environment and,
upon detecting specific inputs, uses built-in computer resources to perform
predefined functions and then process and analyze the data before passing
the data and analytical results to a user or an automated process.12 The
results may alert appropriate crewmembers in sufficient time for them to
initiate recommended and appropriate actions, and some can even perform
diagnostics and restorative actions on their own. The aggregation of smart
sensor data from multiple systems can greatly contribute toward improving
overall vessel efficiency and performance. Such capabilities have changed
the face of engineering departments such that most engine rooms and many
related functions are completely automated. In-person attendance is gen-
erally required only on a sporadic basis and when alarms and exceptional
situations occur.
New generations of smart sensors can make measurements with greater
precision and at higher resolutions. They are fortified with embedded signal
processing capabilities that greatly reduce noise and signal loss, with micro-
processors that analyze data characteristics at the sensor itself and then
communicate their results and enable adjustments to be made to improve
the efficiency and performance of the monitored systems. Sensors used in
AI-based navigation systems can analyze historical and real-time data that
include weather conditions, sea currents and traffic patterns to help optimize
ship routes. The fusion of multiple sensor data such as Radar, Sonar, visible
light and infrared cameras can help detect, identify and track close-by objects
in the ship’s vicinity that pose hazards, including other vessels, off-station
buoys or floating debris for early detection and identification of potential
collisions or obstacles allowing the ship’s crew to take appropriate actions.
Shipboard sensors and sensors contained within shipping containers can iden-
tify sea states that increase their chances of being lost overboard, and sensors
contained within bulk cargo can help determine whether cargo liquefaction
is likely to occur that may destabilize ships. On the nefarious side, unsoli-
cited and unexpected deliveries of smart watches containing smart sensors
to many U.S. sailors and servicemen (and possibly merchant mariners), when
turned on, auto-connect to Wi-Fi and cell phones. Without being prompted,
access is gained to many user data, cameras, microphones and tracking infor-
mation while simultaneously inserting malware into these devices.13
Benefits of smart sensors include instances where previously required dis-
crete sensor infrastructure components were installed throughout the ship;
they are now frequently encapsulated entirely within the sensor itself, result-
ing in sensor data transmission path distances reduced from meters across the
ship to mere millimeters. Shorter signal path distances result in faster data
propagation rates, decreased data processing times and reduced chances of
injecting unwanted noise and interference into the data, especially when the
entire process occurs within the shielded sensor enclosure. Accompanying the
sensor are all required signal processing elements combined with computer-
based sensor analytics and communications hardware needed to transfer
sensor-derived information and analytical results, rather than just the sensor
data itself, but also throughout the vessel itself and even worldwide to remote
monitoring and control facilities. Further enhancing smart sensor systems are
capabilities for self-calibration and integrated diagnostics that lower mainten-
ance costs and improve system reliability by reducing cabling and connector
requirements and simplifying remove and repair actions to replacement of
individual modules. Additional efficiencies are achieved by taking advantage
of distributed computing at the sensor level, enabling rapid adjustments to
processes and adapting to changing conditions as they occur. Sensor mini-
aturization also provides opportunities for including multi-sensor elements
within one enclosure, enabling the measurement of several phenomena with
one sensor assembly rather than multiple assemblies.
The greatest benefit of smart sensors is probably achieved by taking full
advantage of modern microprocessor computing power to examine sen-
sor data in new ways that provide a greater understanding of the phenom-
enon being measured. This includes the integration of machine learning and
deep-learning AI technologies into the sensors themselves that can discern
interrelationships between flows, levels, temperatures, and other meas-
ured characteristics that are not readily detectable through human senses
or understanding. This can be achieved using simultaneous sensor measure-
ments across different sensors, by examining data from one sensor over peri-
ods of time to detect evidence of trends and events contained in the data, and
with measurements from multiple sensors over time to discern even broader
relationships. Having computing power distributed among multiple sensors
can effectively achieve at minimal expense what in previous times could only
be accomplished using supercomputers at an extraordinary cost. The fol-
lowing paragraphs provide a discussion of what machine learning and deep-
learning AI technologies measure and how AI is integrated into sensors and
sensor systems.
6.2 Sensor Data Comprising Multiple Perspectives of

the Same Subject
The physical properties of things can take many dimensions, each providing
a different perspective of behavior or consistency. From one point of view,
sensor data may be considered a sample of all relevant characteristics of a
measured phenomenon at any single point in time. These characteristics may
represent one or more frequency, amplitude, pulse width, resistance, cap-
acitance, or other electrical properties; or a physical property such as level,
height, intensity, temperature, and humidity as converted from electrical
properties. Commonly measured characteristics of a phenomenon taken as a

snapshot at one point in time include multiple signals of interest at different
frequencies, much like the different components of a Radar or Sonar signal.
Such measurements are considered in the frequency domain and represent all
relevant signal properties of the phenomenon being measured at that time.
When this same measurement (snapshot) is repeated multiple times during a
given time period, this sequence is considered in the time domain, whereby
changes to the signal as they occur over time may be revealed.
Frequency and/or time domain measurements can also be converted into
imagery consisting of multiple thousands of picture elements, or pixels, on a
display screen relating to the position, area and size of things. This data are
represented in the spatial domain, which reflects a translation of a spatial
quantity into a pixel representation. This can occur by capturing an image of
a scene or object directly onto pixels, each containing an impression of the
qualities of a small portion of the overall image. The original scene or object
is reconstructed by reproducing the pixel impressions onto a display. Such is
the case for digital and infrared cameras and other visual sensors. Direct-to-
pixel domain imagery is limited based on the size and resolution of the sen-
sor and can be enhanced using optical magnification and greater numbers of
smaller pixels, and by using image filtering and software analytics.
Radar, Sonar, and Lidar images are created using an entirely different pro-
cess involving one or more transducers (antennas) that transmit and/or receive
signals. These signals are subsequently converted into different domain repre-
sentations. An example of Radar is provided in Figure 6.1, where received
waveforms are analyzed in the frequency and time domains (b and c) and
processed to create a pixel domain representation (a).
Figure 6.1 Radar Image in Pixels, with Associated Radar Chirp Waveform Variation
over Time.
Source: R. Glenn Wright. Intelligent Autonomous Ship Navigation using Multi-
S ensor
Modalities. 12th International Conference on Marine Navigation and Safety of Sea
Transportation. June 2017. Gdynia, Poland.
These electronic signals containing highly complex waveforms across

many frequencies are projected through the air or water onto a scene, where
the signals are then modified through reflection and absorption based on the
physical and electrical characteristics of the environment and the objects pre-
sent within the scene. A portion of the transmitted signals are reflected to and
received by the antenna or transducer, which are then analyzed as a function
of changes that occur over time and detected in the frequency spectrum of the
signal. Analysis of time and frequency domain signals is performed to acquire
the information necessary to subsequently create a pixel domain image for
display in a manner unique and customary to Radar, Sonar, and Lidar.
While this indirect approach has proven to be highly accurate and reliable,
it can result in great variability in how the targets and scene are displayed to
the user based on signal resolution and manufacturer user interface design
preferences. For example, a target may be represented as a “blip” on a radar
screen, and forward-looking sonar (FLS) can paint a 3D surface model of
bottom terrain while Lidar systems can display a highly accurate model of
the terrain and quayside environment.
Unlike imaging sensors used in cameras the information contained within
the received signal from Radar and similar active sensors in the time and
frequency domains creates the resulting pixel domain image based upon the
properties of the waveforms being transmitted, the gain and resolution of
the transducer elements, the sensitivity of the receiver and the capabilities of
the software to analyze the reflected signals. The ability to actively interrogate
targets using a wide range of complex waveforms provides great flexibility
to analyze their reflected signal properties across all data classes. Dynamic
adjustment of waveform signal characteristics in real time based upon target
properties and greater capabilities in analyzing time and frequency domain
datasets continue to result in the retrieval of much greater information con-
tent than has ever been possible. Recent examples include the acquisition of
swath bathymetry from FLS and other scientific data from high-resolution
side scan sonar imagery.14,15 Similar advances have also occurred in other
maritime applications including improvements in solid-state Doppler radars.
6.3 Machine Learning and Deep-L earning AI

Machine learning is a component of AI and computer science which focuses
on the use of data and algorithms to imitate the way that humans learn,
gradually improving its accuracy.16 Traditional algorithms must be written so
that their response to every conceivable scenario is anticipated and planned.
In contrast, AI is taught or learns on its own the principles involved in a par-
ticular application and works out its own responses.17 The problem is that
AI-based systems might respond in unanticipated ways, such as inventing
new moves in the game of chess previously unseen by human grandmas-
ters.18 Much depends on the breadth of the application and the methods and
datasets on which the system is trained. It is conceivable that an AI-based

system trained to detect and identify navigation buoys could misidentify an
oblique view of a small boat painted red with a white number on the side as
a navigation buoy.
Various ways to implement machine learning meet the broad definition of
allowing computers to learn, including statistical methods, algorithms and
neural networks. Statistical methods are based upon mathematical concepts
and algorithms in terms of formulas, models, inferencing and other tech-
niques to analyze numerical data and imagery in determining classes, types
and variations within and their features and characteristics. Tools to assist
in this process include various types of regression techniques, decision trees,
and modeling techniques used for predicting and forecasting values, classi-
fication, prediction, clustering, and pattern recognition.19 Depending on the
problem's complexity, more sophisticated models and computational archi-
tectures may be needed to achieve accurate solutions. This would include
artificial neural networks utilizing weighted node structures with multiple
visible and hidden layers rather than algorithms. Machine learning tech-
niques do not necessarily require large amounts of training data or long peri-
ods of time to produce acceptable results, and the data does not have to be
well structured or organized in many cases. Training may be accomplished
by mimicking human methods through supervised learning and, in areas
where humans are less skilled in performing these tasks, some machine learn-
ing architectures can be left to determine trends and outcomes on their own
without human direction through unsupervised learning.
Machine learning using statistical and neural network processes can
determine similarities between live sensor data and imagery and those cor-
responding to training data sets to detect and identify trends, events and
objects as well as determine sensor performance and possible sensor failure
or pending failure. One way to accomplish this is to examine sensor data
features and characteristics found outside two and three standard deviations
by applying a mask representing qualitative tolerances from nominal values
that correspond to nonlinearity, noise, or other aberrations expressed as
a percentage of nominal. Other statistical analysis approaches can be fol-
lowed to reliably and consistently analyze sensor signals considering the
ability to detect, classify and identify unique features and characteristics
associated with the various data, imagery, waveforms and signal types.
Through such means, a wide range of sensors can be supported, providing
the greatest applicability across ship and Unmanned Surface Vehicle (USV)
remote operations.
The acquisition and analysis of numerical data, imagery, waveforms and
other forms of complex sensor signals generally focus on using neural net-
works designed to meet the specific requirements of each sensor and the
objects, events and characteristics of phenomena being sensed. This is espe-
cially relevant in cases where complex sensor data exist in the spatial (data
that exists in various forms of imagery), frequency (data exist within and
across the frequency spectrum) and time (data that change over time) domains
defy description in natural language and where numerical and/or algorithmic
analysis is burdensome and falls short of comprehensive and complete por-
trayal of sensor target behavior and characteristics.
Many, even hundreds, of viable neural network architectures may be used
for sensor data processing and data analytics, each dedicated to solving spe-
cific types of problems and for use in specific applications. One of the more
commonly used forms is the Convolutional Neural Network (CNN), capable
of detecting numerical data, signal, and imagery features and characteristics
needed to identify particular classes, types and variations that may reflect
individual objects, trends and events. When using the supervised learning
method of training, the relevant features present within the data or imagery
are manually extracted and applied in the CNN training process. The advan-
tage of directed learning lies in the reduced time and number of datasets and
images needed for training.
Deep-learning AI is also used to train a CNN to detect sensor signal fea-
tures and characteristics to identify trends, objects and events. In contrast to
selecting specific signal and image features and characteristics of interest as is
accomplished through machine learning, with deep learning, the raw images
are fed directly into a deep neural network where the features are learned
automatically. This method generally requires hundreds of thousands to mil-
lions of images to achieve the best results, depending on the complexity of the
imagery. The advantage of deep learning is in the ability of a CNN to detect
hidden relationships present within imagery that may not be readily discern-
ible to someone and prevent their use of the directed learning method. Both
methods of neural network training can complement statistical methods for
analyzing analog, digital and complex sensor signals and detecting, identi-
fying and categorizing their features, characteristics and attributes.
Another type of deep-learning neural network featured prominently in the
news is Generative Adversarial Networks (GAN), where one neural network
trained to create imagery (GANT) is checked by another trained in the proper
boundaries of the imagery (GANC). This approach has been used successfully
in creating many anatomically correct, “deep fake” human facial images that
are highly accurate in their features yet do not represent any actual human
beings.20 Such networks can be used to supplement a variety of actual sensor
data for training purposes where the characteristics and constraints of real
sensor data are known but difficult to obtain in large volumes.
6.4 Training Dataset Development

Both machine learning and deep- learning AI require large amounts of
imagery and numerical data for effective training and solution development.
However, these numbers can differ by orders of magnitude, where thousands
of datasets are typically used for machine learning. In contrast, hundreds of

thousands or more are generally needed for deep-learning AI. Sources for
numerical data, imagery and training datasets used for machine learning
include the results of computer analysis of sensor signals in the frequency and
time domains, combined with spatial domain imagery generated by the sen-
sors and captured by the computer. However, associated risks include ampli-
fying biases against what may be excluded in the training set and favoritism
toward characteristics and conditions that are included.
This approach was followed in one project using sensor signal data, wave-
forms and imagery to generate the sufficiently large datasets needed to assess
the potential of deep-learning AI for detecting sensor degradation.21 The pro-
ject’s original estimate of 100,000+images rapidly expanded into millions
of images and sensor data representations useful for both machine learn-
ing and deep learning. Limited datasets were selected, where approximately
two-thirds were used for training and one-third for testing, validation, and
system evaluation. The training dataset consisted of two parts: a compre-
hensive dataset and a limited dataset, with the latter being a subset of the
former. The comprehensive dataset was shared for supervised learning in
the development and refinement of statistical processes and for unsupervised
learning in the development of neural network processes. It included com-
plete numerical data providing comprehensive imagery and a description of
imagery components that encompass scale, range, units and other factors.
The limited dataset was used for directed learning to assist in developing
machine learning techniques.
Factors to be considered when developing training datasets for machine
learning and deep-learning AI include ensuring the data contains all rele-
vant features and characteristics, and realizing that unsupervised learning
will probably require an order of magnitude greater volume than supervised
learning. Few, if any, errors should exist in the training dataset if at all avoid-
able, as this will likely confuse decision-making while in actual use. All fea-
tures and characteristics should also be normalized to the same scale. Bias
in machine learning and deep-learning training datasets where favoring of
one class or characteristic of the phenomenon being examined should be
minimized or eliminated altogether, when possible. To combat this type of
problem, it is necessary to ensure that the training dataset represents real-
world conditions in their entire functionality in terms of range, scope and
frequency of occurrence.
6.5 Smart Sensor System Verification, Validation and Test

Three aspects of determining sensor system design and operational correct-
ness must be considered throughout the development process, from the initial
design through final system acceptance. The first step involves verification
that the individual sensors and smart sensor systems fulfill their requirements
and have been correctly implemented in accordance with the overall system
design, and that interoperability of their individual components when inte-
grated into the entire system is demonstrated. Such a process will identify any
missing requirements that should be considered, unnecessary requirements,
and that the design process effectively implemented the requirements. This
topic becomes difficult with machine learning and deep-learning AI, where
unsupervised learning is used when developing the system to define its own
requirements as to what is and is not important based on the data presented
to it during training. This issue is somewhat less significant when dealing
with individual sensors and sensor systems, but becomes vitally important
when the concepts of sensor and smart sensor fusion, sensor degradation,
and cybersecurity are considered.
Validation of sensor system performance must also be accomplished to
determine that the entire system can correctly implement the requirements.
The validation process determines whether the requirements contained and
presented within training datasets were reasonable in terms of their ability to
reflect the actual functions the system must perform, and whether the design
was adequate to achieve the desired functions. Problems are likely to arise
when smart sensors act upon bias and insufficient scope of data is represented
within training datasets to achieve conclusions or solve problems in ways
that are correct, or in novel ways that are unanticipated and may or may not
be correct.
Testing of sensor system performance must also be performed to cover the
breadth and depth of system functionality to determine whether it is suffi-
ciently robust and capable of performing across all required conditions and
operational settings. This will include exercising all system functions with a
wide variety of test cases to demonstrate its capability to perform its required
functions and to ensure that, in the event of failure to perform properly, it
can do so in a manner that is not destructive and provides traceability of the
cause(s) of failure. The capability of the system to achieve interoperability
with other systems must also continue to be demonstrated throughout this
process.
A key element to testing smart sensors must consider the possibility of
achieving verification and validation of machine learning solutions using
statistical processes. However, deep-learning Al solutions are not amenable
to this approach due to their inherent lack of visibility into the functioning
of their internal processing elements. Merely implementing black box testing
techniques based upon system requirements and design is insufficient as deci-
sions are being made in real time, and it may not be possible to effectively
reproduce the same outcome by replicating the same exact stimulus recorded
by the sensors –a task that is essentially impossible to achieve in an oper-
ational environment. This is an area that is still in its infancy, and attempting
to achieve a definitive solution to the verification and validation of deep-
learning Al is beyond the scope of the present-day state of the art.
Lessons can be learned in the the verification, validation, and testing of

deep learning smart sensor applications, by what may be encountered in prac-
tice through examining other areas of machine learning and deep-learning AI
implementations that have been extensively evaluated and tested on a wide
scale by thousands or more users. The problems encountered are indicators
of what should be expected to occur using similar technologies.
For example, criticism of the most sophisticated of today’s chatbots reveals
significant limitations and errors where their results should not be trusted.
The OpenAI ChatGPT chatbot developed using deep-learning AI sometimes
fails to produce accurate answers, makes up information, uses real organ-
ization names to try to legitimize its claims, and even accuses real people of
crimes that they did not commit; all the while sounding extremely convin-
cing by producing eloquent prose and citing non-existent sources that seem
real even to the people it credits.22 Other mistakes include getting verifiable
facts wrong, detailing impossible processes it cites to create an answer, lying
to solve a CAPTCHA test and difficulties in solving simple math problems
related to age and distance.23 Google’s Bard chatbot produces errors and false
information with high levels of confidence, sometimes offers inaccurate infor-
mation or omits important details in its answers, and creates results that can
inadvertently mirror real-world biases and stereotypes contained in the train-
ing datasets.24
Based upon these observations and the present state of the technology,
AI-augmented sensors and smart sensors are extremely likely to not sense
the complete range of objects and conditions for which they are trained, may
falsely detect some objects and conditions that do not actually exist, may
provide incorrect justification for their findings and just sometimes get things
wrong. As such, the use of smart sensors and AI-based decision-making tools
should remain relegated to assistant and advisory roles, with trained mar-
iners making all final decisions. Exceptions should be considered in some
research and test situations when human safety and lives are not placed in
jeopardy, and when significant property damage and environmental risk are
not likely to occur.
6.6 Multi-S ensor Data Fusion

The topic of multi-sensor data fusion has prompted much literature over past
decades regarding the use of mixed and varied data from multiple sensors
to produce a greater awareness and understanding of conditions or events
as they occur. Examples can include examining discrete measurements indi-
cating low oil pressure, high oil temperature, low oil flow and fading power
to conclude a problem exists where the engine is not receiving a sufficient
supply of oil, possibly due to a malfunctioning oil pump. Likewise, a Radar
target two miles off the starboard bow at night that cannot be seen visually
may be detected by an infrared camera as having the shape of a small boat
approaching at fast speed without running lights. What has actually hap-
pened here is the fusion of heterogeneous data from different sensors to com-
pare, discriminate and comprehensively analyze the acquired information
with memory or experience to obtain a more accurate, complete, and more
reliable, consistent interpretation and description of the measured phenom-
enon than is possible using a single sensor.25
Sensor data fusion is a general term that can refer to any method of com-
bining data from multiple sensors. This can include traditional sensors such
as cameras, Radar, Sonar, and Lidar; and also smart sensors that can pro-
cess data locally and make decisions on their own. Much research in this
area has been performed using traditional sensors focusing on specific vessel
applications related to equipment test, diagnosis and prognosis;26,27 gaining
fuel efficiencies by examining the ship’s speed, fuel consumption and weather
parameters;28 and improving object recognition capabilities for autonomous
vessel navigation and collision avoidance.29 The results thus far have been
promising and are providing measurable improvements in sensor system
designs and gains in efficiency and situational awareness.
However, unlike traditional sensors, smart sensors can communicate
with each other and share data, metadata and analytical results in real time,
allowing them to collaborate and make better-informed decisions that can
aid mariners and directly influence autonomous agents operating unmanned
vehicles. Further gains are needed in the design of smart sensors and sensor
systems to achieve greater interoperability and for their use to become more
commonplace. The fusion of measurement analytical results from multiple
smart sensors, not just the measurements themselves, is likely to yield game-
changing improvements over traditional sensor fusion through improve-
ments in accuracy and reliability, greater efficiency, enhanced adaptability to
challenging situations, increased situational awareness, enhanced decision-
making, reduced costs and greater environmental compliance.
6.7 AI and Maritime Training

One bright light for AI use in the maritime industry beyond making sensors
better and more useful lies in the realm of training and education.30 The
development of personalized learning methods and tools that can screen for,
detect and target areas of knowledge deficiency in a student can measur-
ably improve testing and job performance. The value and realism afforded
to training situations and simulations cannot be underestimated when com-
binations of AI-generated and augmented virtual reality systems can create
circumstances where unexpected, real-time curveballs are thrown by an AI-
based automaton in adapting training to current situations.
In their recent study on seafarer training and skills needed to support
a decarbonized shipping industry, DNV noted a general trend towards
a “higher- skilled” seafarer with increased IT, digital, technical and
organizational competence being needed in the future and concluded specific

training and familiarization onboard are seen as an important part of future
training models to ensure seafarers are competent to make use of new tech-
nology.31 Likewise, attracting young people to a career in seafaring is challen-
ging, especially when competing with companies like Microsoft and Google
and the working environments they offer.32 What applies to decarbonization
technology and ergonomics in the engineering and deck departments also
applies to the understanding and operation of smart sensors onboard ships.
Concerted efforts must be made by shipping concerns and training organi-
zations to ensure new generations of smart sensors and sensor systems are
understood by crewmembers who may operate and maintain them with the
greatest effectiveness.
6.8 Effects of AI on the Maritime Workforce

Introducing AI onboard ships through smart sensors is just one example of
how these new technologies are permeating the maritime culture as a whole.
Much has been speculated on how AI may affect the jobs and occupations of
seafarers, whether automatons will work in harmony with humans to assist
them in performing their duties or if seafarers will be replaced. The maritime
industry is much like any other in our desire to gain efficiency in operations,
and doubtless, the management organizations of maritime firms have noticed
many of the impacts AI is having elsewhere. The things AI has demonstrated
as doing proficiently include well-structured tasks performed in occupations
such as bookkeepers, receptionists, couriers, proofreaders, customer support
and technical support.33 Things AI does not do well include managing people
and public relations, entrepreneur skills, human motivation, event planning
where many people and contacts are involved, and occupations that require
great experience involving many people and complex factors where instincts
groomed over many years of experience can be even more valuable than rules
and customs in solving problems and making things happen.
The last significant change in the rules for the game of chess occurred in
1849 with the introduction of the Staunton standard to the public, which is
still the international standard for tournament play.34 AI has demonstrated
extraordinary abilities to learn, innovate and dominate humans in playing
this game over the past several years. Human chess players are still exam-
ining and learning new methods and tactics for the opening moves and play
introduced by AI.
However, the maritime environment is not static. Indeed, it can change on
a minute-to-minute basis by constantly creating new challenges to operate
and maintain evolving equipment designs, specifications and performance
criteria; and having to cope with new and exciting underway scenarios posed
by the effects of weather and influences of the physical environment. Further
pressures are exerted on seafarers by having to keep rigid port schedules to
ensure timely loading, offloading and replenishment; digesting changes to

COLREGS, SOLAS, and other conventions; updates to navigation charts,
Notices to Mariners, Coast Pilots and other publications, plus licensing and
training requirements for their occupations –and the implications of all
these factors for the voyage they are presently conducting and for future
voyages. The risks to life, property and the environment are staggering.
Despite assertions by many to the contrary, AI cannot begin to cope with
so many constantly changing and complex factors. Nor should we expect it
to. Unfortunately, regulatory authorities and the automobile and trucking
industries have ignored these facts, permitting the introduction of self-driving
cars, taxis, trucks and other vehicles onto public highways that have resulted
in the slaughter of many innocents and great loss of property.35,36,37
Shipowners and maritime regulatory authorities (IMO and nation-states)
must not make this same mistake by allowing AI to displace licensed mar-
iners on the bridge and in engineering crews. The maritime workforce is
already in decline, and allowing such practices will inevitably accelerate fur-
ther reduction in properly trained and experienced seafarers. Tracey Mayhew
of the Seafarers International Union noted that increased automation will
see some jobs become obsolete and some new jobs will be created.38 The
introduction of AI should not exacerbate this problem. The consequences in
terms of potential loss of life, environmental damage and high costs incurred
through property loss and damage are too high and cannot be justified given
the present or near-future states of AI technology maturity. Smart assistants
are desirable, but should not be used as an excuse to reduce or replace watch-
standers at night or at any other time.39
At present, maritime law holds masters responsible for most everything
that occurs in or around the ship, including events that are outside their
control. However, this policy has its limits. The space shuttle commander
was not held liable for the defective voyage planning in 1986 resulting from
the near-freezing temperature launch that doomed Challenger. It is doubtful
that the software and AI industries are prepared to assume liability when
the courts determine the automaton “digital navigator” replacing the watch
officer makes a mistake that is regarded as negligent navigation which,
under admiralty law, does not justify an unseaworthiness claim against the
shipowner,40
Despite such warnings, there is still a great need for AI in maritime. The
primary training center for members of the International Organization of
Masters, Mates and Pilots (IOMM&P), the Maritime Institute of Technology
and Graduate Studies (MITTAGS), on their website acknowledges the reality
that AI is not a serious threat to seafarer jobs:41
As AI increases in popularity on ships and vessels, many crew members

and Deckhands are concerned they will lose their jobs as computers can
perform more and more. However, crew members shouldn’t worry. While
AI and machine learning can streamline and assist with operations, they
can only automate so much. Individuals offer unique skills and capabil-
ities that machines cannot replicate, making a human presence on ships
invaluable.
This is not to say that seafarers will be unaffected by AI, and this chapter
illustrates how AI can help to improve safety by enhancing situational aware-
ness across all seagoing departments. Improvements in seafarer education
methods, techniques and facilities to enhance realism and broaden the scope
of training can better prepare crew members for the future working with AI
to enhance their skills and employment opportunities.
References
1 Eric Niiler. The Robot Ships Are Coming … Eventually. Wired Magazine,
Business. 30 October 2020. www.wired.com/story/mayflower-autonom
ous-ships/
2 Andrew Tunnicliffe. Using AI to Navigate the Tricky Topic of Ship Navigation.
Ship Technology. 18 April 2019. www.ship-technology.com/features/ship-nav
igation-system/
3 Annie Brown. Utilizing AI and Big Data to Reduce Costs and Increase Profits
in Departments across an Organization. 13 April 2021. Forbes Magazine.
www.forbes.com/sites/sethmatlins/2023/06/22/the-forbes-worlds-most-influent
ial-cmos-list-2023/
4 Allianz. 2012. Safety and Shipping 1912– 2012: From Titanic to Costa
Concordia. Allianz Global Corporate & Specialty. 3 March 2012. www/agcs/
allianz.com/PDFs/Reports/
5 CBI. 2018. Massive Cargo Ships Are Going Autonomous. CBInsights. 28
August 2018. https://app.cbinsights.com/research/autonomous-shipping-
trends/
6 Staff. AI: The Worst-Case Scenario. The Week. https://theweek.com/artificial-
intelligence/1024341/ai-the-worst-case-scenario
7 Matt Egan. Exclusive: 42% of CEOs Say AI Could Destroy Humanity in Five
to Ten Years. CNN Business. 14 June 2023. www.cnn.com/2023/06/14/busin
ess/artificial-intelligence-ceos-warning/index.html
8 Yarden Gross. Why AI Is a First Responder for Overloaded Crews in Shipping’s
Toughest Era Yet in International Shipping News. 19 August 2021. www.helle
nicshippingnews.com/why-ai-is-a-first-responder-for-overloaded-crews-in-
shippings-toughest-era-yet/
9 Alexander Love. AI in Shipping: Areas to Watch in 2020. Ship Technology. 7
January 2020. www.ship-technology.com/features/ai-in-shipping/
10 Jessica Miley. 11 Times AI Beat Humans at Games, Art, Law and Everything
in Between. 12 March 2018. Interesting Engineering, Inc. https://interestingengi
neering.com/innovation/11-times-ai-beat-humans-at-games-art-law-and-eve
rything-in-between
11 Lakshmi Varanasi. AI Models Like ChatGPT and GPT-4 Are Acing Everything
from the Bar Exam to AP Biology. 25 June 2023. www.businessinsider.com/
list-here-are-the-exams-chatgpt-has-passed-so-far-2023-1?op=1
12 Brien Posey. What Is a Smart Sensor? Techtarget.com. 29 June 2023. www.tec
htarget.com/iotagenda/defi nition/smart-sensor
13 NCIS. Free Smartwatches for Sailors Might Be Covert Cyberattack. Maritime
Executive. 27 June 2023. www.maritime-executive.com/article/ncis-free-smart
watches-for-sailors-might-be-covert-cyberattack
14 FarSounder 2018. FarSounder Joins NOAA as a Trusted Node. FarSounder,
Inc. Press Release. 17 October 2018. www/farsounder.com/about/press_r
eleases
15 R. Glenn Wright. Scientific Data Acquisition Using Navigation Sonar. IEEE/
MTS Oceans Conference. Anchorage Alaska, USA. September 2017.
16 IBM. What Is Machine Learning? IBM, New York. www.ibm.com/topics/mach
ine-learning
17 Nic Gardner. Current and Future Uses of Artificial Intelligence in the Maritime
Industry. Thetius I.Q. 2021. https://thetius.com/current-and-future-uses-of-art
ificial-intelligence-in-the-maritime-industry/
18 John Nash. Checkmate Humanity: In Four Hours, a Robot Taught Itself Chess,
then Beat a Grandmaster with Moves Never Devised in the Game’s 1,500-year
History. The Daily Mail. 21 December 2017. www.dailymail.co.uk/sciencet
ech/article-5204513/Robot-taught-never-seen-chess-moves-hours.html
19 Coursera. 10 Machine Learning Algorithms to Know in 2023. Coursera,
Inc. Updated 16 June 2023. www.coursera.org/articles/machine-learning-alg
orithms
20 Tero Karras, Samuli Laine, and Timo Aila. A Style-Based Generator Architecture
for Generative Adversarial Networks. NVIDIA. arXiv:1812.04948v3 [cs.NE],
29 March 2019.
21 R. Glenn Wright. In-Stride Detection of Sensor Degradation. GMATEK, Inc.
Final Report. Revision: A, 28 April 2022. Contract: N6833520G2005. Naval
Sea System Command, Washington, DC.
22 Emily A. Vogels. A Majority of Americans Have Heard of ChatGPT, but Few
Have Tried It Themselves. Pew Research Center. 24 May 2023. www.pewresea
rch.org/short-reads/2023/05/24/a-majority-of-americans-have-heard-of-chat
gpt-but-few-have-tried-it-themselves/
23 ChatGPT Fails: 13 Common Errors and Mistakes You Need to Know. Search
Engine Land. Accessed 5 July 2023. https://searchengineland.com/chatgpt-fails-
errors-mistakes-400153
24 Google Bard AI Chatbot Advantages, Disadvantages, Review, Features & Bard
vs ChatGPT. Online Sciences. Accessed 5 July 2023. www.online-sciences.com/
technology/google-bard-ai-chatbot-advantages-disadvantages-review-features-
bard-vs-chatgpt/
25 Guo Chen, Zhigui Liu, Guang Yu, and Jianhong Liang. (2021) “A new view
of multisensor data fusion: research on generalized fusion,” Mathematical
Problems in Engineering, 2021, Article ID 5471242, p. 21. https://doi.org/
10.1155/2021/5471242
26 Wen Jiang, Weiwei Hu, and Chunhe Xie. (2017). A new engine fault diagnosis
method based on multi-sensor data fusion. Applied Sciences, 7(3), 280. https://
doi.org/10.3390/app7030280
27 Youyu Zeng and Qiang Xie. Real-Time Ship Fault Diagnosis Algorithm based
on the fusion of CNN and RNN. Proc. 2020 the 10th International Workshop
on Computer Science and Engineering (WCSE 2020). Shanghai 19–21 June
2020, pp. 70–75. doi: 10.18178/wcse.2020.06.012. ISBN 978-981-14-4787-7
28 Yuquan Du, Yanyu Chen, Xiaohe Li, Alessandro Schönborn, and Zhuo
Sun. (2022) Data Fusion and Machine Learning for Ship Fuel Efficiency
Modeling: Part III –Sensor Data and Meteorological Data. Communications
in Transportation Research, 2, p. 100072, ISSN 2772-4247, https://doi.org/
10.1016/j.commtr.2022.100072
29 DNV- Maritime. New Research Project to Investigate Sensor Fusion and
Collision Avoidance for Advanced Ships. DNV. Accessed 29 June 2023. www.
dnv.com/news/new-research-project-to-investigate-sensor-fusion-and-collision-
avoidance-for-advanced-ships-26127
30 P. Roberts. Effects of Artificial Intelligence on Maritime Training. 31 March
2023. Shoreside Maritime Training Magazine. https://shoresidetraining.co.uk/
news/the-effects-of-artificial-intelligence-on-maritime-training/
31 Raymond Antoni Kaspersen (PM), et al. Insights into Seafarer Training and
Skills Needed to Support a Decarbonized Shipping Industry. DNV. Report
no. 2022-0814, rev. 0. 2022-11-04. General Conclusions 3 and 4, p. 8.
32 Marcus Hand. Attracting Future Seafarers and Engine Room Ergonomics.
Seatrade Maritime News. 29 June 2023. www.seatrade-maritime.com/crewing/
attracting-future-seafarers-and-engine-room-ergonomics
33 Sophia Bernazzani, 10 Jobs Artificial Intelligence Will Replace (and 10 That
Are Safe). Blog Hotspot. Updated: 21 February 2020. https://blog.hubspot.
com/marketing/jobs-artificial-intelligence-will-replace
34 The Staunton Standard: Evolution of the Modern Chess Set. World Chess Hall
of Fame. St. Louis, MO, USA. https://worldchesshof.org/exhibit/staunton-stand
ard-evolution-modern-chess-set
35 Ariana Bindman. Incompetent Driverless Cars Are Wreaking Havoc on San
Francisco. SFGATE, 6 June 2023. www.sfgate.com/tech/article/cruise-waymo-
driverless-cars-san-francisco-18132953.php
36 Nearly 400 Car Crashes in 11 Months Involved Automated Tech, Companies
Tell Regulators. The Associated Press. 15 June 2022. www.npr.org/2022/06/15/
1105252793/nearly-400-car-crashes-in-11-months-involved-automated-tech-
companies-tell-regul
37 Ed Leefeldt and Amy Danise. Injured in a Car Accident? AI Is Your Legal
Adversary. Forbes Magazine. 9 June 2023. www.forbes.com/advisor/car-insura
nce/car-accident-injury-artificial-intelligence/
38 Tracey Mayhew. Uncertainty for Maritime Workforce of the Future. MARAD
Symposium, Achieving Critical MASS. 23 July 2019. www.maritime.dot.gov/
sites/marad.dot.gov/files/docs/about-us/foia/11696/seafarers-international-
union-uncertainty-maritime-workforce-future.pdf
39 Michael Grey. The Case for Fewer Eyes on the Bridge at Night. Seatrade
Maritime News. 20 June 2023. www.seatrade-maritime.com/crewing/case-
fewer-eyes-bridge-night
40 The Maritime Executive. Will Owners of Autonomous Ships Be Liable for
Autonomous Mistakes? 19 May 2023. https://maritime-executive.com/article/
will-owners-of-autonomous-ships-be-liable-for-autonomous-mistakes
41 MITAGS. How AI Is Changing the Maritime Industry. Maritime Institute of
Technology and Graduate Studies. Linthicum, MD, USA. Blog, News. 10 April
2023. www.mitags.org/ai-impact-maritime-industry/
Chapter 7
Shore and Intership

Communications
Electronic and optical technologies provide a means to extend the human

senses of vision and hearing and to project our ability to speak from the
vessels on which we sail to span the local vicinity and out to the entire
world in real time. It also enables our capacity to communicate data and
information generated by the electronic sensors installed on our vessels at
sea to other vessels, corporate offices, Remote Operating Centers (ROCs),
national authorities and other locations where such a need exists. This
includes access to telecommunications and broadband infrastructure for
crews to contact families and people ashore; and to fulfill many maritime-
specific needs to communicate with other ships of any size or nationality
related to urgent maritime safety information, distress alerts in an emer-
gency and contact rescue coordination centers anywhere in the world as
well as nearby ships.
This chapter describes means that exist to facilitate one-way and two-way
communication using modern technological tools that include wireless and
physical connections. Vessels must be appropriately equipped, and seafarers
must have a working knowledge of many modern tools under the SOLAS
Convention. Traditional communication methods involving sight and sound
also remain vital for all operations in the immediate vicinity of the vessel
through the use of lights, dayshapes, whistles, horns, bells, smoke, fire, and
gunfire.
While traditionally staffed and real-time, remotely operated vessels have
crewmembers knowledgeable and experienced in the art of communication,
unmanned vehicles must be taught not only these methods and techniques but
also the meaning and significance of the content of these communications.
7.1 Communications Infrastructure

The single most significant advance in communications to and from ships at
sea in the twenty-first century has been the advent of Wide Area Networks
(WAN) capable of spanning vast distances and providing many circuit paths
Doi: 10.1201/9781003377900-7
102 Shore and Intership Communications
to achieve global connectivity. Shipboard communications equipment util-

izing satellite and technologies make it possible to access the Internet using
public and dedicated private networks of various and diverse configurations
of fiber optic and copper cables, microwave links, ground stations and ter-
restrial networks. These same infrastructure form the backbone of VHF
and HF GMDSS messaging and other national emergency services world-
wide. Public Metropolitan Area Networks (MAN) provided by regional
governments, local cities and municipalities that serve limited geographic
areas such as ports and harbors achieve much the same connectivity using
Wi-Fi connections, as do private services associated with ports and indus-
tries within the ports limited to ships and other entities by their need to do
business.1
Detailed discussion on the topic of worldwide communications infrastruc-
ture is deferred to others who consider the broader aspects of information
technology (IT) as it applies to individuals, businesses and governments in
general, where communications by ships comprise only a tiny part of their
scope. However, it is mentioned here to provide a background that supports
the various means of communication by which ships interact with other ships
and shore entities. The following paragraphs discuss these various means of
communication by ships.
7.2 Satellite Communication

Satellites are commonly used for a wide variety of navigation-related mari-
time voice and data communications that include routine status updates
between ships and their home offices, tracking ship movements, and increas-
ingly, command and control of conventionally staffed vessels and unmanned
and autonomous vehicles. Also included are communications related to the
interruption of normal navigation by emergencies with regard to distress sig-
nals and critical messages related to accidents, loss of propulsion, medical
emergencies, piracy and significant matters requiring immediate attention.
This mode of communication is greatly relied upon while at sea for voice
and data communications due to its reliability and large bandwidth capacity.
Satellites in Low Earth Orbit (LEO) are located between 100 and 1,200
miles above the Earth’s surface, with an orbital period of between 90 and
130 minutes. The operator with the most LEO satellites is SpaceX which, as
of 2023, has approval to launch 7,500 of its planned 29,988-spacecraft con-
stellation.2 Other LEO satellite operators include Starlink, OneWeb, Iridium,
Inmarsat/Viasat3, SateliotX, Orbcomm, and Saab. One notable Medium
Earth Orbit (MEO) satellite provider for maritime is SES with their O3b
mPower broadband services. Located in Geosynchronous Earth Orbit (GEO)
Shore and Intership Communications 103
is Inmarsat/Viasat, which provides its Orchestra, Elera, and Global Xpress

networks.
A wide range of different bands of frequencies are used for satellite com-
munications depending upon service providers, the types of data transferred
and bandwidth requirements. These include the L-band (1–2 GHz), S-band
(2–4 GHz), C-band (4-8 GHz), X-band (8–12 GHz), Ku-band (12–18 GHz)
and the Ka-band (26.5–40 GHz) frequencies.
7.2.1 Very-S mall-A perture Terminal (VSAT) Satellite

Systems
Very-Small-Aperture Terminal (VSAT) has gained wide acceptance for mari-
time communications due to its small antenna footprint that makes installa-
tion possible on many smaller vessels at relatively lower cost when compared
to more traditional, high-bandwidth satellite systems. These systems pro-
vide Internet access as well as voice, video and data communications, fleet
management and other applications using satellites positioned in GEO that
operate in the C, Ku and Ka frequency bands. To gain additional bandwidth,
increase reliability, reduce latency and overcome the limitations associated
with any one satellite services technology or provider, many users are com-
bining VSAT with LEO systems such as Starlink and 5G services as a backup
to L-band services for guaranteed worldwide connectivity.4
7.2.2 Satellite-B ased Automatic Identification System

(S-A IS)
An early version of S-AIS commenced in 2008 with the launch of an Orbcomm
concept demonstration satellite equipped with AIS capability under the U.S.
Coast Guard Nationwide Automatic Identification System (NAIS) project
to test the feasibility and effectiveness of Automatic Identification System
(AIS) message reception and reporting from space for ship tracking and
other navigational activities.5 Data collected by the NAIS would support
maritime interests by promoting vessel and port safety through collision
avoidance and through detection, traffic identification and classification of
vessels out to 2,000 nautical miles from shore. Subsequently, S-AIS services
have expanded worldwide through other satellite systems and many com-
mercial services to support vessel management to ensure safer ship routing,
enable Vessel Traffic Monitoring and Information Systems (VTMIS), pro-
mote ship reporting, validate ship declarations, offer efficient Vessel Traffic
Services (VTS), manage arrivals and perform effective vessel traffic analysis.6
S-AIS is also a valuable tool to track and assist vessels in distress, monitor
piracy and prevent smuggling, and aid in environmental preservation and

fishery protection.
7.2.3 Remote Vessel Control via Satellite

Communication satellites also provide the primary means of remote control
navigation of vessels. First demonstrated in 2017, the 28 m tug Svitzer
Hermod safely conducted a number of maneuvers in Copenhagen harbor
while being controlled by an ROC.7 Wärtsilä also remotely controlled their
Highland Chieftain platform supply vessel in the North Sea using a standard
satellite link over 5,000 miles from its office in San Diego, California.8 In
2020, the fully autonomous Mayflower equipped with 6 smart cameras, 30
onboard sensors and 15 intelligent network edge devices capable of ana-
lyzing and using data in real time sent data and metadata from the voyage
back to the onshore team via Iridium Certus weather-resilient broadband
satellite connections.9,10 Such technology is steadily working its way into all
aspects of commercial shipping as a means for companies to gain further
insight into operational parameters while maintaining steady contact with
their vessels at sea and to assist in emergency situations should remote con-
trol of the vessel be needed.
7.2.4 Limitations of Satellite Communications

Providing a broad expansion of services to vessels at sea comes with limita-
tions associated with satellite communications that should be noted. These
can include frequency congestion, interference from weather and sunspots,
and propagation delay for GEO satellites. Coverage of specific geographic
areas may vary, especially with LEO and MEO satellites that are constantly
changing their relative position in the sky with respect to the ground and
when another satellite is not available to take the place of a satellite that
has gone out of view. Satellite communications are also more vulnerable to
security risks than terrestrial communications, as their data can be inter-
cepted by third parties.
7.3 Very High Frequency (VHF) Radio

Maritime VHF- FM radio provides two- way voice and data communica-
tions between vessels and shore stations. The useful distance of a maritime
VHF radio is generally within line of sight depending upon the height of
the antenna, weather conditions, interference from other users and natural
events. VHF radios with 25-watt output power can have a range of 20–60
nautical miles between antennas mounted on tall ships, towers and hills; and
5 nautical miles between antennas mounted on small boats at sea level. Small
handheld transceivers with 5-watt radiated output power cover considerably
less distances. Table 7.1 illustrates a list of VHF international transmitting

frequencies.11
Useful for distress and safety communications, VHF radio is one of the
primary means of communication in the event of a distress or safety situ-
ation at sea. Channel 16 is designated as the international distress, safety
and calling frequency that vessels are required to monitor at all times while
underway. Channel 13 is dedicated for bridge-to-bridge communications
in the local vicinity to assist in navigation by coordinating activities such
as passing or docking or sharing information about weather conditions or
hazards.
Broadcasts containing weather forecasts and other marine information
can also be received on this channel. Another use is to communicate with
shore stations such as marinas, harbormasters and the Coast Guard; and for
requesting assistance, reporting marine incidents, or making arrangements
for fuel, food, or other supplies.
There are many restrictions and operationally relevant conditions that are
specific and subject to national and regional requirements and regulations
not illustrated in this table. Examples include frequency use by land and air-
craft stations for search and rescue coordination, pollution and other safety-
related operations; radiated power restrictions; digital selective calling for
distress, safety and calling; AIS and mobile-satellite service for the reception
of AIS transmissions; simplex and duplex frequency transmissions; restric-
tions from harmful interference; and frequency use for digital systems and
new technologies.
7.3.1 Digital Select Calling (DSC)

Initiated as part of the Global Maritime Distress and Safety System
(GMDSS), a new generation of marine VHF (and HF) radios feature a
Digital Selective Calling (DSC) capability that enables mariners to send or
receive distress, urgency, safety and routine radiotelephone calls to or from
any similarly equipped vessel or shore station including automatic digitally
formatted distress alerts to coast guards or other rescue authorities any-
where in the world.12 These communications may be encrypted for security
purposes. DSC is a mandatory requirement under the SOLAS Convention
for all ships that are required to carry a VHF radio, and is recommended for
all other ships.
DSC utilizes the Maritime Mobile Service Identity (MMSI) and call sign
programed into the VHF radio upon initial installation of the vessel, and pos-
ition information acquired using a built-in GNSS capability present within
the radio itself or obtained from an onboard GNSS received connected to the
radio. The MMSI of the recipient vessel or shore station is entered, and the
message is initiated by pressing the “send” button on the radio. DSC is an
essential GMDSS component to improve safety at sea.
Table 7.1 VHF International Transmitting Frequencies
Channel Notes Transmitting Intership Port Operations and Ship

Designator Frequencies (MHz) Movement
Ship Coast Single Two Frequency

Stations Stations Frequency
60 m 156.025 160.625 x x
1 m 156.05 160.65 x x
1001 156.05 156.05 x
61 m 156.075 160.675 x x
2 m 156.1 160.700 x x
62 m 156.125 160.725 x x
3 m 156.15 160.750 x x
63 m 156.175 160.775 x x
1063 156.175 156.175 x
4 m 156.2 160.800 x x
64 m 156.225 160.825 x x
5 m 156.25 160.850 x x
1005 156.25 156.25 x
65 m 156.275 160.875 x x
1065 156.275 156.275 x
6 f 156.3 x
2006 r 160.900 160.900
66 m 156.325 160.925 x x
1066 156.325 156.325 x
7 m 156.35 160.950 x x
1007 156.35 156.35 x
67 h 156.375 156.375 x x
8 156.4 x
68 156.425 156.425 x
9 i 156.45 156.45 x x
69 156.475 156.475 x x
10 h,q 156.5 156.5 x x
70 f,j 156.525 156.525 Digital selective calling for distress
11 q 156.55 156.55 X
71 156.575 156.575 X
12 156.6 156.6 X
72 i 156.625 x
13 k 156.65 156.65 x x
73 h,i 156.675 156.675 x x
14 156.7 156.7 x
74 156.725 156.725 x
15 g 156.75 156.75 x x
75 n,s 156.775 156.775 x
16 f 156.8 156.8 Distress, safety and calling
76 n,s 156.825 156.825 x
17 g 156.85 156.85 x x
77 156.875 x
18 m 156.9 161.5 x x
1018 156.9 156.9 x
78 m 156.925 161.525 x x
1078 156.925 156.925 x
2078 mm 161.525 161.525 x
19 m 156.95 161.55 x x
1019 156.95 156.95 x
2019 mm 161.55 161.55 x
Source: US Coast Guard, 2023. VHF International Transmitting Frequencies. US Coast

Guard Navigation Center. Department of Homeland Security. www.navcen.uscg.gov/
international-vhf-marine-radio-channels-freq
Public Channel Notes Transmitting Intership Port Operations and

Correspondence Designator Frequencies (MHz) Ship Movement
Ship Coast Single Two

Stations Stations Frequency Frequency
x 79 m 156.975 161.575 X x
x 1079 156.975 156.975 X
2079 mm 161.575 161.575 X
x 20 m 157 161.6 X x
x 1020 157 157 x
x 2020 mm 161.6 161.6 x
x 80 157.025 161.625 x x
1080 157.025 157.025 x
x 21 y,wa 157.05 161.65 x x
1021 157.05 157.05 x
x 81 y,wa 157.075 161.675 x x
x 1081 157.075 157.075 x
x 22 y,wa 157.1 161.7 x x
1022 157.1 157.1 x
x 82 x,y,wa 157.125 161.725 x x
1082 157.125 157.125 x
23 x,y,wa 157.15 161.75 x x
1023 157.15 157.15 x
x 83 x,y,wa 157.175 161.775 x x
1083 157.175 157.175 x
x 24 w,wx,x, 157.2 161.8 x x
1024 w,wx,x, 157.2
2024 w,wx,x, 161.8 161.8 x (digi
xx
84 w,wx,x, 157.225 161.825 x x
1084 w,wx,x, 157.225 x (digi
xx
2084 w,wx,x, 161.825 161.825
25 w,wx,x, 157.25 161.85 x x
1025 w,wx,x, 157.25 x (digi
xx
2025 w,wx,x, 161.85 161.85
85 w,wx,x, 157.275 161.875 x x
1085 w,wx,x, 157.275
2085 w,wx,x, 161.875 161.875
26 w,ww, 157.3 161.9 x x
1026 w,ww, 157.3
2026 w,ww, 161.9
86 w,ww, 157.325 161.925 x x
1086 w,ww, 157.325
2086 w,ww, 161.925
27 z,zx 157.35 161.95 x
1027 z,zz 157.35 157.35
x ASM 2 z 161.95 161.95
87 z,zz 157.375 157.375 x
x 28 z,zx 157.4 162 x
1028 z,zz 157.35 157.35 x
ASM 2 z 162 162
x 88 z,zz 157.425 157.425 x
AIS 1 f, l, p 161.975 161.975
AIS 2 f, l, p 162.025 162.025
7.3.2 Weather Broadcasts, Notices to Mariners

Many countries broadcast maritime weather information on maritime VHF
radio. Some of the most common channels used for this purpose include:
• Channel 16 (156.8 MHz): The international distress, safety and call-

ing channel, it can also be used for safety-related weather information
broadcasts.
• Channel 6 (156.3 MHz): Weather broadcasts in the United States, Canada
and some other countries.
• Channel 22A (157.1 MHz): Weather broadcasts in the United Kingdom
and Ireland.
Broadcast Notice to Mariners (BNMs) provide mariners with information

about potential hazards such as dangerous weather, shoals, wrecks and other
vessels may also be transmitted on VHF channels 16 and 22A. Different VHF
frequencies over which national weather offices may send local and nearshore
coastal marine forecasts and weather observations and forecasts include
162.400 MHz, 162.500 MHz, 162.425 MHz, 162.525 MHz, 162.450 MHz,
162.550 MHz and 162.475 MHz. Coastal stations with ranges between 20
and 25 miles offshore can also broadcast predicted tides and real-time obser-
vations from buoys and coastal meteorological stations. Specific channels
and frequencies used for maritime weather broadcasts can vary from country
to country.
7.3.3 Emergency Services

Emergency services are available through maritime VHF radio under pro-
grams supported by the United States Coast Guard, the European Maritime
Safety Agency (EMSA) and other national entities. The Rescue 21 program
is the U.S. Coast Guard’s advanced command, control and direction-finding
communications system created to better locate mariners in distress and save
lives and property at sea and on navigable rivers.13 Rescue 21 helps identify
the location of mariners in distress using VHF radio via towers that generate
lines of bearing to the source of radio transmissions, significantly reducing
search time, meeting both the safety requirements needed for the expand-
ing U.S. Marine Transportation System and for International Convention for
the Safety of Life at Sea (SOLAS) standards. It extends coverage to a min-
imum of 20 nautical miles from the coastline; improves information sharing
and coordination with the U.S. Department of Homeland Security and other
federal, state and local first responders; and can also help watchstanders
recognize potential hoax calls by identifying discrepancies between a caller
reported and actual locations, thus conserving and reducing risk to valuable
response resources. Other features of the Rescue 21 system include improved
clarity, recording and playback capabilities for all communications, simul-

taneous monitoring and relaying of information over multiple radio frequen-
cies, portable tower communications equipment for use during emergencies
and natural disasters and automated transmission of urgent marine informa-
tion broadcasts.
The EMSA is an agency of the European Union that is responsible for
maritime safety and security. It operates several programs that provide fea-
tures similar to the U.S. Rescue 21 program, including the European Coastal
Monitoring System (ECMS) and the European Maritime Safety Information
System (EMSIS). EMSA programs acquire a comprehensive overview of activ-
ities at sea, which is a challenge for most authorities working in the maritime
domain through their Integrated Maritime Services.14 They aim to provide
coast guards and border guards, customs authorities, naval forces, pollution
response centers, rescue coordination centers and vessel traffic management
services all need detailed, reliable information about occurrences at sea in
real time to perform their duties effectively.
7.3.4 VHF Signal Propagation Effects

A significant exception to the stable and dependable characteristics of
VHF radio can be attributed to the phenomena of “atmospheric duct-
ing”, where occasionally the transmission of VHF electromagnetic waves
through a region near the tropospheric layer of the atmosphere can literally
extend hundreds of miles or more beyond the line of sight with little signal
attenuation.
Despite normal signal absorption from being reflected from the ionosphere
or passing over the surface of the earth, ducting can be attributed to radio
waves propagating from one end to the other end of the duct by undergoing
successive refraction from the troposphere.15 This duct is a region or layer
present in the lower atmosphere occurring from a temperature inversion that
behaves as a waveguide propagating radio waves along the curvature of the
earth and allows the spreading of wavefronts in a horizontal manner.
7.4 Medium Frequency (MF) and High Frequency (HF)

Radio
Maritime MF and HF radio also provide two-way voice and data commu-
nications between ships, boats and shore stations in the same manner as
VHF radio, but at much lower frequencies with drastically different radio
wave propagation characteristics. The useful distance of maritime MF and
HF radios can generally range from less than one hundred miles to thou-
sands of miles depending upon the frequencies involved and several other
factors. Tables 7.2 and 7.3 list the MF and HF Single Sideband Maritime
Radiotelephone Channels.
Table 7.2 MF Single Sideband Maritime Radiotelephone Channels
2 MHz Simplex 2.003 kHz 2.142 kHz 2.214 kHz 2.738 kHz
Channels 2.0825 kHz 2.182 kHz 2.635 kHz 2.782 kHz
2.086 kHz 2.203 kHz 2.638 kHz 2.830 kHz
2.093 kHz
MF radio waves in the 2 MHz band are characterized by a ground wave

that can propagate over the horizon, and skywave propagation that can
be reflected by the ionosphere. The ground wave is the dominant mode of
propagation for MF radio waves in coastal areas and near shore, whereas
the skywave is the dominant mode of propagation for MF radio waves in
open ocean areas. HF radio waves in the 4 MHz through 26 MHz bands are
characterized primarily by skywaves reflected by the ionosphere, which is the
dominant mode of propagation for HF radio waves in all areas, both land
and sea.
7.4.1 Digital Select Calling (DSC)

DSC capabilities available for maritime HF radios are essentially the same
as those described in the previous paragraph for maritime VHF radio. DSC
services, which may also be encrypted for security purposes, enable mariners
to send or receive distress, urgency, safety and routine radiotelephone calls to
or from any similarly equipped vessel or shore station, including automatic
digitally formatted distress alerts to coast guards and rescue authorities any-
where in the world.
Operating the DSC feature is accomplished where the Maritime Mobile
Service Identity (MMSI) and call sign programmed into the HF radio upon
initial installation on the vessel are encoded within a digital message along
with position information acquired using a built-in GNSS capability present
within the radio itself, or obtained from an onboard GNSS received con-
nected to the radio. The MMSI of the recipient vessel or shore station is
entered into the radio keyboard and the message is initiated by pressing the
“send” button.
7.4.2 Weather Broadcasts and Marine Safety Information

A variety of weather information and maritime safety information is avail-
able over MF and HF radio through NAVTEX, Radiofacsimile and radi-
oteletype. High seas forecasts and storm warnings are also available from
national sources worldwide from whom station locations and schedules may
be obtained. Transmission range is dependent upon operating frequency,
newgenrtpdf
Table 7.3 HF Single Sideband Maritime Radiotelephone Channels
ITU Channel No. Coast Transmit Ship Transmit Notes
4 MHz Duplex Channels 401–4 27 4.357–4 .435 kHz 4.068–4 .143 kHz 3 kHz spacing
4 MHz Duplex Channel 428 4.351 kHz Varies
4 MHz Duplex Channel 429 4.354 kHz Varies
4 MHz Simplex Channel 4.146–4 .149 kHz 3 kHz spacing
1 through 21
4.000–4 .060 kHz
4 MHz Shared Channels 3 kHz spacing
4 MHz (Calling) 421 4.147 kHz 4.125 kHz Distress and safety working
4.125 kHz simplex
4 MHz (USCG Calling) 424 4.435 kHz 4.134 kHz
Shore and Intership Communications

6 MHz (Calling) 606 6.215 kHz Distress and safety working
6.225 kHz simplex
Distress 833 8.291 kHz Distress and safety working
8.291 kHz simplex
8 MHz Duplex Channels 834–8 37 8.707–8 .716 kHz Varies
8 MHz (Calling) 821 8.779 kHz 8.255 kHz
12 MHz Duplex Channels 1201–1 241 13.077–1 3.197 kHz 12.230–1 2.350 kHz 3 kHz spacing
12 MHz Simplex Channels 12.353–1 2.365 kHz 3 kHz spacing
12 MHz (Calling) 1221 13.137 kHz 12.290 kHz Distress working 12.290
kHz simplex
111
(Continued)
newgenrtpdf
Table 7.3 (Continued)
112
ITU Channel No. Coast Transmit Ship Transmit Notes
Shore and Intership Communications

16 MHz (Calling) 1621 17.302 kHz 16.420 kHz Distress working 12.290
kHz simplex
18/1 9 MHz Duplex Channels 1801–1 815 19.755–1 9.797 kHz 18.780–1 8.822 kHz 3 kHz spacing
18/1 9 MHz Simplex Channels 18.825–1 8.843 kHz 3 kHz spacing
18/1 9 MHz (Calling) 1806 19.770 kHz 18.795 kHz
22 MHz (Calling) 2221 22.756 kHz 22.060 kHz
25/2 6 MHz Duplex Channels 2501–2 510 26.145–2 6.172 kHz 25.070–2 5.097 kHz 3 kHz spacing
25/2 6 MHz Simplex Channels 25.100–2 5.118 kHz 3 kHz spacing
25/2 6 MHz (Calling) 2510 26.172 kHz 25.097 kHz
time of day and atmospheric conditions and can vary from only short dis-
tances to several thousand miles.
The primary method of broadcasting Marine Safety Information (MSI)
is NAVTEX, which can provide coverage up to around 400 nautical miles
out to sea although this depends on many reception issues. All NAVTEX
broadcasts are made on 518 kHz, using narrow-band direct printing 7-unit
forward error correcting (FEC or Mode B) transmission. This service is com-
plemented by radio telephony with networks of remote radio sites around
coasts providing MF coverage of 150 miles.
HF radiofacsimile broadcasts originating from various worldwide loca-
tions are shown below.16 All emissions are amplitude modulation, single side-
band, suppressed carrier, analog (J3C) and frequencies are subject to change.
Australia: 2,628 kHz, 5,100 kHz, 5,755 kHz, 7,535 kHz, 10,555
kHz, 11,030 kHz, 13,920 kHz, 15,615 kHz, 20,469 kHz,
18,060 kHz.
Canada: 3,253.0 kHz, 4,271 kHz, 4,416 kHz, 4,292.0 kHz,
6,915.1 kHz, 6,496.4 kHz, 7,710.0 kHz, 8,456.0 kHz,
10,536 kHz, 13,510 kHz.
China: 4,199.75 kHz, 8,412.5 kHz, 12,629.25 kHz,
16,826.25 kHz.
Chile: 4,228.0 kHz, 4,322.0 kHz, 8,677.0 kHz, 8,696.0 kHz,
17,146.4 kHz.
Greece: 4,481 kHz, 8,105 kHz.
Germany: 3,855 kHz, 7,880 kHz, 13,882.5 kHz.
Japan: 148 kHz, 3,622.5 kHz, 7,795 kHz, 13,988.5 kHz.
Japan/Singapore: 4,316 kHz, 8,467.5 kHz, 12,745.5 kHz, 16,035 kHz,
16,971 kHz, 17,069.6 kHz, 17,430 kHz, 22,542 kHz.
Korea Republic: 3,585 kHz, 5,857.5 kHz, 7,433.5 kHz, 9,165 kHz,
13,570 kHz.
New Zealand: 3,247.4 kHz, 5,807 kHz, 9,459 kHz, 13,550.5 kHz,
16,340.1 kHz.
Russia: 4,481 kHz, 5,336 kHz, 6,446 kHz, 7,907 kHz, 7,908.8
kHz, 8,105 kHz, 8,444 kHz, 10,130 kHz.
Thailand: 7,395 kHz.
United Kingdom: 2,618.5 kHz, 4,610.0 kHz, 8,040.0 kHz, 11,086.5 kHz.
United States: 2,054 kHz, 4,298 kHz, 4,317.9 kHz, 4,346 kHz, 8,459
kHz, 8,503.9 kHz, 8,682 kHz, 12,412.5 kHz, 12,786 kHz,
12,789.9 kHz, 17,146.4 kHz, 17,151.2 kHz, 22,527 kHz.
7.4.3 Precision Time Broadcasts

Should GNSS time reference signals become unusable or are otherwise
unavailable, there exist precision time signals broadcast worldwide that can
be used for navigation purposes. A list of many of these signals and the fre-
quencies on which they are broadcast appears below:
• Canada, CHU (Ottawa), 3.33, 7.85 and 14.67 MHz.17

• China, BPM (Pucheng), 2.5 MHz, 5 MHz, 10 MHz, 15 MHz.18
• Germany, DCF77 (Mainflingen), 77.5 kHz.19
• Japan, JJY (Tokyo), 40 kHz, 60 kHz.20
• South Korea, HLA (Daejeon), 5 MHz.21
• Taiwan, BSF (Taichung), 77.5 kHz.22
• United Kingdom, MSF (Cumbria), 60 kHz.23
• United States, WWV (Colorado Springs, CO), 2.5, 5, 10, 15 and 20 MHz.24
• United States, WWVH (Hawaii), 2.5, 5, 10, 15 and 20 MHz.25
7.4.4 HF Signal Propagation Effects

One of the greatest factors affecting radio signal propagation in the lower
frequencies of the radio spectrum is the time of day and specifically, the effect
of the sun’s radiation on the ionosphere. High atmospheric ionization lev-
els during the day can make it more difficult for radio waves to propagate,
whereas at night the atmosphere is less ionized making it easier for radio
waves to propagate. One great exception to this rule is the effect of sunspots
combined with the usual daylight ionization effects that can inhibit normal
local, short-distance communications while simultaneously making it pos-
sible to communicate with stations thousands of miles away. This results in
“skip” conditions where radio waves repeatedly reflect between the surface
and the ionosphere over great distances.
Seasons can also affect propagation with winter ionization levels gener-
ally higher and more difficult for radio waves to propagate than with lower
summer levels. Weather can also affect propagation with high atmospheric
pressure causing radio waves to propagate farther than during periods of
low atmospheric pressure when radio waves, in general, propagate shorter
distances. Geographic features can also affect propagation where mountains
and other obstacles can block radio waves along with the curvature of the
Earth, which can also limit radio wave propagation distance.
7.5 Global Maritime Distress and Safety System

(GMDSS)
The GMDSS is an automated ship-to-shore and ship-to-ship distress and radio
communication safety system that uses satellites and/or terrestrial radio sys-
tems with digital selective calling (DSC) technology.26 Developed under the
International Maritime Organization (IMO) and mandated by the SOLAS
Convention for ships that are over 300 gross tons and also ships that carry
more than 12 passengers internationally, it provides safety of life information
and communication systems that inform vessels of navigation hazards and
weather conditions, and enable distress calls with pertinent location and iden-
tification information with the push of a button. GMDSS is comprised of safety
procedures, types of equipment and communication protocols used for safety
and rescue operations of distressed ships, boats and aircraft. It also allows ships
to be located quickly and efficiently in the event of a distress situation. A major
component is the COSPAS-SARSAT satellite search and rescue (SAR) system,
where instruments are flown on board LEO, MEO and GEO satellites provided
by the United States, Russian Federation, India and the European Union.27
These satellites are capable of detecting signals coming from the Earth’s sur-
face transmitted by an emergency Search and Rescue Transmitter (SART) that,
when activated, broadcasts a distress beacon on a frequency of 406 MHz to
send a signal that indicates the position of a distressed vessel.
All GMDSS ships must carry a 406 MHz Emergency Position Indicating
Radiobeacon (EPIRB), a VHF radio capable of transmitting and receiving
DSC and radiotelephony, a Navigational Telex (NAVTEX) receiver, a SART,
backup power systems and two-way VHF portable radios. In GMDSS Sea
Area 2, where radiotelephone coverage of at least one MF coast station in
which continuous DSC (2,187.5 kHz) alerting and radiotelephony services
are available, ships must also carry a DSC-equipped MF radiotelephone.28
For all voyages conducted in GMDSS Sea Areas 3 and 4, vessels must carry
either an Inmarsat F77, B or C ship earth station, or a DSC-equipped HF
radiotelephone/telex.28
Companies such as Inmarsat/Viasat using radionavigation transponders
can enable Satellite-Based Augmentation System (SBAS) services around the
world for coast guard and other organizations to enhance standard GPS/
Galileo accuracy of 5–10 meters to as little as 10 cm using satellite connect-
ivity, land-based infrastructure and software.30 Precise tracking such as this
could enable pinpoint safety navigation and help emergency services reach
vessels in distress more quickly.
7.6 Visual Communications

Several different means of visual communication are used to indicate ship
function and activities to other vessels, to identify aids to navigation and to
indicate distress situations at sea.
7.6.1 Lights and Dayshapes

A significant portion of the IMO Convention on the International Regulations
for Preventing Collisions at Sea (COLREGS) Part C is dedicated to commu-
nication using visual signals to be complied with in all weathers (Rule 20).
These signals have to do with the use of lights from sunset to sunrise, and
from sunrise to sunset in restricted visibility and may be exhibited in all other
circumstances when it is deemed necessary. Also specified is the use of day-
shapes by day.
Knowledge of the content, meaning and implications of each of these sig-

nals is required of watchstanders and autonomous assistants on the bridge,
and their use becomes secondhand during normal operations. An illustration
of some of these lights and dayshapes is provided in Figure 7.1. All lights
and dayshapes shown are international, and inland visual signals are likely
to be different in accordance with the customs and practices of individual
nations. Additional lights not shown but required when making way through
the water include a masthead light or lights, sidelights and a stern light. Also
not shown is an all-round flashing yellow light that can be found on an air-
cushion vessel when operating in the non-displacement mode, and any spe-
cial light or lights that may be used by a submarine on the surface.
a (Top) Pilot vessel –COLREGS Rule 29. (Middle) Sailing vessel under-
way, or single light with three quadrants: white, red and green –Rule
25. (Bottom) Vessel at Anchor –Rule 30, or power-driven vessel of less
than 7 meters in length whose maximum speed does not exceed 7 knots –
Rule 23.
b Vessel not under command or restricted in their ability to maneuver –
Rule 27.
c Vessel constrained by her draft –Rule 28. Vessel aground –Rule 30.
d Vessel engaged in mine clearance operations (<50 m) –Rule 27.
e Vessel proceeding under sail when also being propelled by machinery,
except that a vessel of less than 12 meters in length is not required to
exhibit the dayshape –Rule 25.
f (Top) Vessel engaged in trawling. (Bottom) Vessel engaged in fishing other
than trawling –Rule 26.
g Same as f. except additional daymark indicating gear extended beyond
150 meters horizontally from the vessel –Rule 26.
Figure 7.1 Masthead Lights and Dayshapes.

Notes: International Rules for the Prevention of Collisions at Sea, 1972 (COLREGS) as codi-
fied in the United States by Coast Guard Commandant Instruction M16672.2D Navigation
Rules –International and Inland. Part C, Lights and Shapes. 25 March 1995.
h Power-driven vessel towing astern –towing vessel less than 50 meters in

length; length of tow exceeds 200 meters –Rule 24.
i Vessel constrained by her draft –Rule 28.
j Vessel restricted in her ability to maneuver, or small vessel engaged in div-
ing operations (<50 m) –Rule 27.
k Vessel restricted in her ability to maneuver, at anchor (<50 m) –Rule 27.
l Vessel engaged in dredging or underwater operations when restricted in
ability to maneuver –not making way with an obstruction on the star-
board (or port) side –Rule 27.
These specific visual signals when combined with navigation lights while
making way that reflects various vessel configurations must be distinguished
from deck lighting and other background lights likely to be encountered.
This requires knowledge and ability that is relatively easily taught by educa-
tors and learned by mariners as part of their normal training and experience.
However, in addition to the basic steering and sailing rules of the COLREGS
Part B, these Lights and Shapes requirements of the COLREGS Part C must
be included in all training regimens for crewed vessels and unmanned and
autonomous vehicles using visible light cameras with the development pro-
cess properly verified and the results validated and tested. This includes train-
ing encompassing all ship viewing perspectives at sea through 360° in varying
visibility and weather conditions before such machines should be released
onto the high seas and inland waterways. Some research in this regard has
been published in the literature, but its scope is limited with further investi-
gation still needed.31
7.6.2 Coded Light Signals from Aids to Navigation

Lighted code signals differentiated by color and rhythm are used to identify
maritime aids to navigation, such as lighthouses, range lights, and buoys.32
Lighthouses and range lights may display distinctive light characteristics to
facilitate individual aid to navigation recognition that is verifiable using the
Light List and navigation charts, but no special significance should be attached
to the color or rhythm of such lights. Other lighted aids to navigation, such
as buoys, employ light characteristics to convey additional information. The
description below follows the International Association of Marine Aids to
Navigation and Lighthouse Authorities (IALA) B international aid to naviga-
tion marking system used in the Americas and other parts of the world, and
does not conform to the IALA A system used throughout Europe and many
parts of Asia. However, the concepts regarding light signals are still applic-
able to communication regardless of whether the IALA A or B convention
is used.
In much the same manner as for lights and shapes, lighted AtoN also
display various characteristics essential to ensure proper identification and
thereby position. These characteristics include green and red lights that mark
the sides of channels and locations of wrecks or obstructions that are to be
passed by keeping these lights on the port (left) hand of a vessel. Green and
red lights are also used on Preferred Channel Marks, where the topmost band
is green and red, respectively. Yellow or white lights having no lateral signifi-
cance can also be displayed on AtoN, whose purpose may be determined by
their shape, color, letters or numbers, and the light rhythm employed.
Light rhythms can vary greatly displaying both regularly flashing or regu-
larly occulting light rhythms. Some examples include when flashing lights
(frequency not exceeding 30 flashes per minute) may also be used, and for
situations where lights require a distinct cautionary significance such as at
sharp turns, sudden channel constrictions, wrecks, or obstructions, a quick
flashing light rhythm (60 flashes or more per minute) may be used. Composite
group flashing light rhythms consisting of groups of two flashes followed by
one flash may be found on some AtoN in addition to Safe Water Marks
that display a white Morse Code “A” rhythm (short-long flash) and Isolated
Danger Marks that display a white group flashing two flashes. Other lights
include those used on Special Marks that display yellow lights with fixed
or slow flashing rhythms, mooring buoys and information and regulatory
marks displaying white lights of various rhythms.
Lights containing sector guidance showing one color from most directions
and a different color or colors over definite arcs of the horizon as indicated
on the appropriate nautical chart are sometimes used to mark shoals or warn
mariners of other dangers. Approximate bearing information is provided
where the observer can note a change of color as the boundary between the
sectors is crossed.
Some research in the area of computer-based recognition of AtoN lights
and colors has appeared in the literature.33 However, there still exists much
work to be done before unmanned and autonomous vehicles will be capable
of dealing with visual AtoN recognition at night. Further information on the
IALA A and B system of buoyage and aids to navigation is available from
IALA.34
7.6.3 International Code of Signals (ISC)

The International Code of Signals (ISC) considers situations related to the
safety of navigation and persons, especially when language difficulties arise.35
Useful for all modes of communication, it contains alphanumeric characters
and visual flags representing specific information that may be combined with
several flags, or semaphores, to send various messages. Methods of visual sig-
naling messages may include the use of flags (see Figure 7.2),34 flashing lights
and sound signaling using Morse code symbols, voice over a loud hailer and
Morse signaling by hand flags or arms. In addition, the use of radiotelegraphy
Figure 7.2 International Code of Signals, Flags and Pennants.

Source: Athul, M., Yogitha, A., and Nandakumar, R. Flag semaphore detection using
Tensorflow and Opencv. International Journal of Recent Technology and Engineering
(IJRTE), 7(6), ISSN: 2277-3 878 (Online), March 2019.
and radiotelephony is appropriate for communicating characters and num-

bers to compose messages.
Signal flags are grouped according to alphabetic flags, numeric pennants,
substitutes, repeaters and special flags using six prominent colors that are
easily recognized including blue, yellow, red, black and white. The number of
flags used to create a message also conveys meaning with one flag for urgent
or common signals, two flags for distress and maneuvering, three flags for
compass points, relative bearings, standard times, verbs and punctuation;
four flags for geographical signals, names of ships, etc.; and five flags for time
and position.
Examples of international meanings for signal flags include:34
A (Alpha) Diver down.

B (Bravo) Carrying dangerous cargo.
C (Charlie) Yes (affirmative).
D (Delta)
Keep clear of me. I am maneuvering with difficulty.
E (Echo)
Altering course to starboard.
F (Foxtrot)
I am disabled, communicate with me.
G (Golf)
I want a pilot.
H (Hotel)A pilot on board.
I (India)
I am altering my course to port.
J (Juliet)
Vessel on fire keep clear.
K (Kilo)
I want to communicate with you.
L (Lima)
Stop your vessel instantly, I have something important to
communicate.
M (Mike) My vessel is stopped and making no way through the water.
N (November) No (negative).
O (Oscar) Man overboard.
P (Papa) In port: All personnel return to ship; vessel is about to sail;
At sea: It may be used by fishing vessels to mean: “My nets
have come fast upon an obstruction”.
Q (Quebec) I request free pratique.
R (Romeo) Reverse course.
S (Sierra) Engines are going astern.
T (Tango) Keep clear; engaged in trawling.
U (Uniform) You are heading into danger.
V (Victor) Require assistance.
W (Whiskey) Require medical assistance.
X (X-ray) Stop your intention.
Y (Yankee) Am dragging anchor.
Z (Zulu) I require a tug.
There are clearly hundreds of messages capable of being communicated

using signal flags, and many mariners are proficient in their usage and mean-
ing. Some research is also being performed to a limited extent regarding
machine recognition and understanding of flags.35,36 However, the scope of
this research does not extend to knowledge and understanding of multiple
flags communicating complex messages of the nature likely to be encountered
by vessels at sea as would be useful for unmanned and autonomous vehicles.
7.6.4 Visual and Audible Emergency and Distress Signals

It is the duty of the master or individual in charge of a vessel to render
assistance to any individual found at sea in danger of being lost, so far as
the master or individual in charge can do so without serious danger to the
master’s or individual’s vessel or to individuals on board. This implies that
watchstanders and other crewmembers are trained in methods to communi-
cate distress and in the understanding of their meanings.
Various means of attracting attention and securing assistance using visible

and audible means have been devised should the need arise. These include
a variety of signaling devices, both pyrotechnic and non-pyrotechnic, such
as hand-held red flares and parachute flares for use both day and night and
electric distress lights for the night. Day signals include hand-held and float-
ing orange smoke signals, a signal mirror, water dye markers, a flag showing
a square and ball displayed against an orange background, flying the
NOVEMBER flag directly over the CHARLIE flag, and flying a square black
flag with a black ball directly underneath. Additional visible signals can
include the flashing of a high-intensity white light flashing at regular intervals
from 50 to 70 times per minute, flames and/ or smoke emanating from
a vessel, a person on deck waiving their arms, the continuous sounding of a
fog horn, the use of five prolonged blasts from a horn, signaling the Morse
code “SOS” (● ● ● — — — ● ● ●) from a horn (possibly supplemented using
light signals in coordination with the horn), and the firing of a gun at regular
intervals of one minute. Figure 7.3 illustrates several signals and methods
that can be used to communicate distress situations at sea.
Figure 7.3 Visual Distress Signals.

Visual and audible signals within the line of sight of potential rescuers
may provide the most expeditious response, but they may also be subject to
misunderstanding by untrained observers and automated assistants that can
result in a significant lag in time before the significance of distress signals are
made clear, if at all. In addition to these local signals, distress signals should
be sent to coast guards and regional search and rescue authorities using the
more reliable and noticeable long-range emergency notification capabilities
provided through VHF and HF radio communications, Inmarsat, and other
satellite service providers.
Despite their effectiveness, the reliability of detection is reduced since
monitoring the horizon to detect visual and audible distress signals is mainly
accomplished by bridge watchstanders and observers who happen to gaze in
the right direction at the right time. Smart visible light and infrared cameras
with 360° coverage of the horizon and trained in flare, fire, and smoke detec-
tion using AI techniques can greatly assist in detecting such events through
continuous monitoring and the ability to immediately notify watchstanders
of such an occurrence. Camera coverage is becoming more widespread as
onboard security interests increase, and integration of visual signal compre-
hension technology can be accomplished relatively easily. Likewise, moni-
toring of audio to detect and locate gunfire is already possible and in use
throughout many of the larger cities of the world. Similar methods can be
used to train smart microphones and audio systems to detect ship horns emit-
ting warning, danger and distress signals.
7.7 Sound and Light Signals

In much the same manner as for visual and audible distress signals, Part D
of the COLREGS addresses sound and light signals and states that devices
installed on ships capable of producing loud sounds can include a horn
(whistle), bell, and gong. Such sounds, possibly in conjunction with light
signals, can coordinate passage between two vessels at close quarters, warn
of danger, or reveal own vessel’s presence and position under reduced visi-
bility conditions. Sounds must be produced using any sound signaling appli-
ance capable of producing the a “short blast” of about 1 second’s duration,
or a “prolonged blast” lasting from 4 to 6 seconds’ duration. In addition to
human sight and sound, high-resolution cameras and sensitive directional
microphones can supplement human capabilities in detecting and under-
standing these signals. Methods of using these sounds are described in the
following paragraphs.
7.7.1 Maneuvering and Warning Signals

When vessels are in sight of one another, a power-driven vessel underway
shall when maneuvering, indicate the nature of the maneuver by the following
signals (and, if equipped, may be supplemented by an all-round white visible

at a minimum range of 5 miles while the maneuver is being carried out):37
• One short blast (and light flash of at least one second’s duration, if
equipped) to mean, “I am altering my course to starboard”.
• Two short blasts (and light flashes, if equipped) to mean, “I am altering
my course to port”.
• Three short blasts (and light flashes, if equipped) to mean, “I am oper-
ating astern propulsion”.
A yellow light may be used in lieu of a white light in some world regions
depending on national regulations.
When vessels are in sight of one another in a narrow channel or fairway, a
vessel intending to overtake another shall indicate intentions by the follow-
ing signals:
• Two prolonged blasts followed by one short blast (and light flashes, if
equipped) to mean, “I intend to overtake you on your starboard side”.
• Two prolonged blasts followed by two short blasts (and light flashes, if
equipped) to mean, “I intend to overtake you on your port side”.
The vessel about to be overtaken shall indicate her agreement by the fol-
lowing signal on her whistle:
• One prolonged, one short, one prolonged and one short blast (and light
flashes, if equipped), in that order.
When vessels are in sight of one another and approaching each other
where, from any cause, either vessel fails to understand the intentions or
actions of the other, or is in doubt whether sufficient action is being taken
by the other to avoid collision, the vessel in doubt shall immediately indicate
such doubt by giving at least five short and rapid blasts (and light flashes, if
equipped).
Also, a vessel, when nearing a bend or an area of a channel or fairway
where other vessels may be obscured by an intervening obstruction shall
sound one prolonged blast. Such signal shall be answered with a prolonged
blast by any approaching vessel that may be within hearing around the bend
or behind the intervening obstruction.
These signals presuppose proper training of seafarers in the detection of
horns and lights that includes the ability to discern their presence in the midst
of background clutter noise. Bridge aids for watchstanders along with view-
ing and listening technology on unmanned and autonomous vehicles using
directional microphones can narrow the relative bearing of the signal, and
one or more cameras can be used to positively identify the signaling vessel.
Training of AI-based smart microphones using the specific horn blast pat-
terns and frequencies designated for use by horns, bells and gongs combined
with smart camera training using these same patterns represented by white
lights can today make possible the automated detection and identification of
these signals using existing technology.
7.7.2 Sound Signals in Restricted Visibility

Special sounds are designated for use by vessels in areas and during times
of restricted visibility during both day and night conditions.38 Under such
conditions in or near an area of restricted visibility, the following signals are
mandated, as appropriate:
• A power-driven vessel making way shall sound at intervals of not more

than 2 minutes one prolonged blast.
• A power-driven vessel underway but stopped and not making way shall
sound at intervals of not more than 2 minutes, two prolonged blasts in
succession with an interval of about 2 seconds between them.
• A vessel not under command, a vessel restricted in her ability to man-
euver, a vessel constrained by her draft, a sailing vessel, a vessel engaged
in fishing, and a vessel engaged in towing or pushing another vessel shall
sound at intervals of not more than 2 minutes three blasts in succession
consisting of one prolonged followed by two short blasts.
• A vessel engaged in fishing, when at anchor, and a vessel restricted in her
ability to maneuver when carrying out her work at anchor, shall instead
sound at intervals of not more than 2 minutes three blasts in succession
consisting of one prolonged followed by two short blasts.
• A vessel towed or, if more than one vessel is towed, the last vessel of the
tow, if manned, shall at intervals of not more than 2 minutes sound four
blasts in succession, namely one prolonged followed by three short blasts.
When practicable, this signal shall be made immediately after the signal
made by the towing vessel.
• When a pushing vessel and a vessel being pushed ahead are rigidly con-
nected in a composite unit they shall be regarded as a power-driven vessel
and shall give the signals appropriate for being underway, and making
way and not making way.
• Additional signals using bells and gongs are also identified for use by ves-
sels when at anchor, aground and pilot vessels performing pilotage duties.
Although in or near areas of restricted visibility the use of lights may be

of limited utility, audio monitoring using sensitive directional microphones
can greatly improve situational awareness that can complement Radar and
AIS sensor data. Training of smart microphones on the specific patterns and
frequencies mandated for horn, bell and gong use under these circumstances
can result in speedy detection and positive identification of nearby vessels.
7.7.3 Signals to Attract Attention

There are times when it may be necessary to attract the attention of another
vessel.39 This should be accomplished using light or sound signals that can-
not be mistaken for any authorized signal, or by directing a searchlight beam
in the direction of the danger, in such a way as not to embarrass any vessel.
Any such signal must not be mistaken for any aid to navigation, and the use
of high-intensity intermittent or revolving lights, such as strobe lights, shall
be avoided.
Automated computer-based detection and understanding of visual and
audible signals intended to attract attention would follow the same general
processes as for other light and sound signals. However, the range of impro-
visation and variability of such signals may make this task difficult.
7.8 Summary
This chapter has presented many different modes of communication as well
as visual and audible messaging and content that can take years for trained
mariners to master. Research has been performed but few case studies exist
where attempts to automate individual tasks have been overwhelmingly suc-
cessful, and such efforts continue. However, these technologies are at best
still experimental and as of yet are unproven.
Also notable is the fact that, with the introduction of unmanned and
autonomous technologies into the maritime mix, it appears that exceptions
are being made at the national levels to eliminate the need to possess full
knowledge of the requirements, capabilities, applications and means of use
for all communication methods expected of seagoing vessels. The full signifi-
cance in granting such exceptions to overall maritime safety and safety of
navigation is yet to be determined.
References
1 Gavin Wright. Metropolitan Area Network (MAN), definition. TechTarget.
February 2021. www.techtarget.com/searchnetworking/defi nition/metropoli
tan-area-network-MAN
2 Mike Wall, 2022. SpaceX Gets Permission to Deploy 7,500 Next-Generation
Starlink Satellites. Space.com. 2 December 2022. www.space.com/spacex-fcc-
approval-7500-starlink-satellites
3 Mike Schuler, 2023. Viasat Completes Acquisition of Inmarsat. gCaptain. 31
May 2023. https://gcaptain.com/viasat-completes-acquisition-of-inmarsat/
4 Marcus Hand. Are Hybrid VSAT and LEO Networks the Next Maritime
Communications Solution? Seatrade Maritime News. 02 June 2023. www.seatr
ade-maritime.com/technology/are-hybrid-vsat-and-leo-networks-next-marit
ime-communications-solution?
5 US Coast Guard, 2008. Coast Guard Announces Successful Launch of
Nationwide AIS Satellite. Coast Guard News. 23 January 2002, https://coa
stguardnews.com/coast-guard-announces-successful-launch-of-nationwide-ais-
satellite/
6 Daniel O’Donohue. How Cargo Ships and Marine Vessels Are Tracked.
Mapscaping. 9 September 2022. https://mapscaping.com/tracking-ships-and-
vessels/
7 Rolls-Royce, 2017. Rolls Royce Demonstrates Worlds First Remotely Operated
Commercial Vessel. Press Release. 20 June 2017. www.rolls-royce.com/media/
press-releases/2017/20-06-2017-rr-demonstrates-worlds-first-remotely-opera
ted-commercial-vessel.aspx
8 Wärtsilä, 2017. Successful Test of Remote Control Ship Operating Capability.
Wärtsilä Marine Solutions. Wärtsilä Electrical and Automation. September
2017. www.wartsila.com/docs/default-source/marine-documents/business-
white-papers/white-paper-o-ea-remote-control-ship.pdf?sfvrsn=16669e45_6
9 Mayflower 2022. It’s Time for the Mayflower Autonomous Ship. www.mas
400.com/technology
10 Iridium, 2021. The Mayflower Autonomous Ship Transforms Ocean Science
with the Help of Iridium. Iridium Communications Inc, Press release, 27 May
2021. https://investor.iridium.com/2021-05-27-The-Mayflower-Autonomous-
Ship-Transforms-Ocean-Science-with-the-Help-of-Iridium-R
11 US Coast Guard, 2023. VHF International Transmitting Frequencies. US Coast
Guard Navigation Center. Department of Homeland Security. www.navcen.
uscg.gov/international-vhf-marine-radio-channels-freq
12 DSC, 2023. United States Coast Guard, Navigation Center. U.S. Department of
Homeland Security. Digital Select Calling. https://navcen.uscg.gov/digital-select
ive-calling
13 Rescue 21. U.S. Coast Guard, Acquisitions Directorate. Homeland Security.
www.dcms.uscg.mil/Our-Organization/Assistant-Commandant-for-Acquisiti
ons-CG-9/Programs/C4ISR-Programs/Rescue-21/
14 EMSA, 2023. Enhanced Maritime Picture via Integrated Maritime Services.
European Maritime Safety Agency. www.emsa.europa.eu/we-do/digitalisation/
maritime-monitoring.html
15 Electronics Desk, 2023. Duct Propagation. https://electronicsdesk.com/duct-
propagation.html
16 Worldwide Radiofacsimile Broadcast Schedules, National Oceanographic and
Atmospheric Administration (NOAA), Bethesda, MD, USA. 21 October 2022.
17 CHU. National Research Council. https://nrc.canada.ca/en/certifications-eval
uations-standards/canadas-official-time/nrc-shortwave-station-broadcasts-chu
18 BPM. Chinese Academy of Sciences. www.sigidwiki.com/wiki/BPM
19 DCF. Physikalisch- Technische Bundesanstalt (PTB), www.ptb.de/cms/en/ptb/
fachabteilungen/abt4/fb-44/ag-442/dissemination-of-legal-time/dcf77/
20 JJY. National Institute of Information and Communications Technology.

https://jjy.nict.go.jp/news/index-e.html
21 HJA. Korea Research Institute of Standards and Science. https://en.wikipedia.
org/wiki/HLA_(radio_station)
22 BSF. National Time and Frequency Standards Laboratory. www.wikiwand.
com/en/BSF_(time_service)
23 MSF. National Physical Laboratory (NPL). www.npl.co.uk/msf-signal
24 WWV. National Institute of Standards and Technology (NIST). www.nist.gov/
pml/time-and-frequency-division/time-distribution/radio-station-wwv
25 WWVH. National Institute of Standards and Technology (NIST). www.nist.
gov/pml/time-and-frequency-division/time-distribution/radio-station-wwvh
26 GMDSS. Global Maritime Distress and Safety System (GMDSS). U.S. Federal
Communications Commission. www.fcc.gov/wireless/bureau-divisions/mobil
ity-division/maritime-mobile/ship-radio-stations/global-maritime
27 Cospas-Sarsat System Overview. National Oceanographic and Atmospheric
Administration. Bethesda, MD, USA. www.sarsat.noaa.gov/cospas-sarsat-sys
tem-overview/
28 US Coast Guard, 2023a. GMDSS Areas and Search and Rescue Areas. US
Coast Guard Navigation Center. Department of Homeland Security. https://
navcen.uscg.gov/gmdss-areas-and-search-and-rescue
29 gCaptain, 2023. Inmarsat to Launch I-8 Satellites to Power L-Band Network.
gCaptain Daily. 19 May 2023. https://gcaptain.com/inmarsat-to-launch-i-8-sat
ellites-to-power-l-band-network
30 Xin Hu, Yingjun Zhang, and Shaobo Wang. Ship Lights Recognition Using
Image Processing. 20th COTA International Conference of Transportation
Professionals. August 2020. DOI: 10.1061/9780784482933.039
31 USC Part 62. United States Code, Part 62, Subpart B. U.S. Aids to Navigation
System. Paragraph 62.45, Light Characteristics.
32 Xu Han, et. al. Multilabel Video Classification Model of Navigation Mark’s
Lights Based on Deep Learning Computational Intelligence and Neuroscience.
Volume 2021, Article ID 6794202, 13 pages. https://doi.org/10.1155/2021/
6794202
33 IALA Pubs. AtoN Design and Delivery. Stnd. 1020. www.iala-aism.org/cont
ent/uploads/2021/02/IALA-Technical-documents-Catalogue-Edition2-final-
compressed-more-min.pdf
34 Pub 102. International Code of Signals for Visual, Sound and Radio
Communications. United States Edition. Pub 102. 1969 Edition (revised 2020).
National Geospatial-Intelligence Agency.
35 Athul Motty, Yogitha A and R Nandakumar. Flag semaphore detection using
Tensorflow and Opencv. International Journal of Recent Technology and
Engineering (IJRTE), 7(6), ISSN: 2277-3878 (Online), March 2019.
36 Ming Gu, Kun Hao, and Zhiyi Qu. Flag detection with convolutional network.
CSAI’18: Proceedings of the 2018 2nd International Conference on Computer
Science and Artificial Intelligence, December 2018, pp. 258–262. https://doi.
org/10.1145/3297156.3297159
37 COLREGS, Rule 34. IMO Convention on the International Regulations for
Preventing Collisions at Sea, 1972. Rule 34. Maneuvering and Warning Signals.
38 COLREGS, Rule 35. IMO Convention on the International Regulations

for Preventing Collisions at Sea, 1972. Rule 35. Sound Signals in Restricted
Visibility.
39 COLREGS, Rule 36. IMO Convention on the International Regulations for
Preventing Collisions at Sea, 1972. Rule 36. Sound Signals to Attract Attention.
Chapter 8
Intraship and Internal

Communications
The previous chapter discussed how sensor and other data may be com-
municated beyond the confines of the vessel itself to other ships and shore
facilities using satellite, radio, visual and other methods. However, internal
communication of sensor data representing functions associated with engin-
eering, navigation, cargo, and human activity comprise a very large subset
of overall vessel communications. Security and crowd control technologies
found worldwide at military installations, financial institutions, government
facilities and even amusement parks have also been tailored adapted to these
same applications on ships. This is accomplished using a combination of
dedicated circuits and common data bus architectures utilizing both wired
and wireless transmission modes tailored to the specific needs of the sensors
communicating their measurements and observations.
The independent and self-contained environment found onboard surface
ships and vehicles provides a unique combination of data communication
requirements not found elsewhere. These requirements are supported using
protocols, including Ethernet, developed to support high-speed data com-
munications between computers and peripheral devices such as disk drives
and printers. These protocols are general in nature and have been adapted
to accommodate the needs of sensors and sensor systems. Others, such as
the Controller Area Network (CAN) and their derivatives that include the
National Marine Electronics Association (NMEA) 0183 and 2000 data bus
architectures, are tailored to support the maritime-specific needs of engin-
eering and navigation sensors as they evolved to accommodate expanding
requirements associated with vessel automation and digitalization.
Data bus technology developed for the operation and automation of indus-
trial control systems that include RS-485 form the basis for monitoring many
different types of bulk, liquid, gas and other types of cargo found onboard
vessels of all sizes, types and kinds. USB (Universal Serial Bus) and Bluetooth
wireless technology provide access to individual network system devices that
can be a gateway to the entire ship’s network infrastructure. Wireless Fidelity
(Wi-Fi) technology enables wireless access to the Internet and local network
resources by enabling devices to connect to a local area network (LAN).
Doi: 10.1201/9781003377900-8
130 Intraship and Internal Communications
This chapter hopes to bring together some of these concepts in a logical

way that will convey the unique requirements of ship sensors and how their
needs are considered in the performance of their everyday functions. This is
accomplished by describing some essential factors that must be considered
when configuring sensor communications using bus architectures, followed
by a high-level comparison between the different standards and protocols
used in shipboard and surface vehicle environments. Details are then pro-
vided describing the most prevalent data bus architectures used throughout
the maritime industry.
8.1 Integrated Bridge Systems

An integral part of any discussion on internal ship communications must
begin with the ultimate goal, which is the existence of an integrated bridge
system (IBS). This is defined as a combination of interconnected systems that
allow centralized access to sensor information or command/control from
workstations, with the aim of increasing safe and efficient ship management
by suitably qualified personnel.1 They are designed to perform two or more
operations associated with passage execution, communications, machinery
control; loading, discharging and cargo control; and safety and security.
There is a commonality in the navigation-related information needed for safe
passage along a route presented by an Integrated Navigation System (INS)
and an IBS, often on shared or common workstations. However, an IBS con-
siders an even broader range of engineering and navigation sensor data to
fulfill ship management needs by qualified personnel in the overall safe and
efficient operation of the vessel.
As a rule, sensor technology development and the range of their applica-
tions have always long outpaced instruments developed by IMO that address
their use. IBS is not an exception to this rule. SOLAS chapter V, adopted in
December 2000, cites in Regulation 19 carriage requirements that limit IBS
scope to shipborne navigational systems and equipment under the purview
of the officer in charge of the navigational watch.2 This chapter addresses
those engineering and navigation sensors directly integrated into IBS. In the
near term, passenger and cargo sensors are not likely to be considered to any
great extent in the design of these systems. This does not mitigate the fact
that many of these sensors report data on factors that can have significant
implications in terms of passage execution, safety and security that must not
be ignored.
8.2 Essential Factors in Sensor Data Communications

The needs of sensor data communications overlap and are shared across
many of the same systems, with their overall efficiency being highly dependent
upon several factors that include:
Intraship and Internal Communications 131
• Data Quality
• Data Rate
• Latency
• Bandwidth
• Security
• Flexibility
In addition, the physical characteristics of the data bus can also affect its
performance. For example, the length of the data bus and the distance data
must travel can affect latency, and the type of cable used can affect bandwidth
characteristics. Specific factors most important for a particular application
depend on the inherent requirements associated with the sensors attached
to the bus, the nature of sensor data communicated therein, the length of
the communications channel and other aspects of the application. As a rule,
a sensor data bus featuring a low error rate, high data rate, low latency,
high bandwidth and strong security will be more powerful than a data bus
with lesser performance characteristics. Note that these factors are associated
with the communication channel itself and are considered independent and
apart from sensor performance characteristics identified in Chapter 2 (2.3)
including linearity, range, accuracy, range, sensitivity, etc.
8.2.1 Data Quality

The quality of sensor data is based in part upon the frequency of occurrence
and the severity of errors that exist within the data. Error in terms of sensor
measurements can be defined as “the result of a measurement minus the true
value of the measurand”.3 Several types of errors are related to sensor data
quality, including outliers, missing data, bias, drift, noise, constant value and
uncertainty conditions.4
Outliers, or anomalies associated with measurement values or data points
that differ significantly from other observations, are one of the greatest errors
in terms of frequency of occurrence and significance. These can be caused by
failures in proper measurement by the instrument itself through improper
scaling. They can also be due to the effects of the environment on the data
bus used as the communication channel caused by temperature, humidity,
interference, poor connectivity or other influences. Another source of outliers
can include random errors caused by arbitrary fluctuations in the data bus.
For example, the measurement may be affected by vibrations or electrical
noise. Outliers may also be created by novel data or data points of a differ-
ent distribution than the rest of the data set due to a momentary interruption
of data or uncharacteristic data. In all cases, outliers can have a significant
impact on statistical data analytics that can lead to inaccurate results.
Missing data or data that is not otherwise available for an individual
observation or series of observations is likely to be attributed to sensor
malfunction, breakdown in the communication channel, or environmental

factors where the data collected may have been too extreme causing it to be
lost or corrupted. Missing data can have a significant impact on sensor data
analytics, where the absence of a large number of data points may cause
errors leading to inaccurate conclusions.
Measurement bias refers to systematic error whose outcome includes meas-
urement results that are consistently too high or too low. This can occur due
to errors within the sensor itself due to improper calibration and environ-
mental factors affecting the data bus that can skew measurements and affect
their accuracy. Multiple methods to acquire sensor data, as well as the use of
multiple sensors, can help to reduce measurement bias. Implementing more
effective means to minimize the effects of environmental factors can also help
in reducing measurement bias.
Drift that can cause a gradual and cumulative change in the output of a
sensor over time, thereby corrupting the signal, can be introduced into a sen-
sor data bus through noise and other errors in the measured signal. Sensor
data bus malfunction can also introduce errors due to electrical interference,
timing errors, or other factors causing the sensor to become unstable and can
lead to drift over time.
The sensor data bus can be a source of electrical noise that can interfere
with sensor measurements and corrupt sensor data. This noise can be caused
by the high-speed switching of transistors or other electrical interference by
improperly shielded data bus controller hardware.
Errors caused by electrical noise, timing problems, drift, noise and loss of
data bus stability can lead to reduced confidence in the degree to which the
measurement can be trusted to be accurate. Uncertainty can be quantified
in terms of a confidence level (in percent) wherein measurement values are
likely to contain the true value of the measurement. Uncertainty can also be
illustrated by the standard deviation in the spread of measurement values in
a data set, where a close spread would indicate greater certainty in the meas-
urement as opposed to a widespread that would indicate greater uncertainty.
8.2.2 Data Rate

Computer system data rate refers to the speed at which data is transferred
within the computer or between a peripheral device and the computer, and
is measured in different scales of bytes per second.5 For sensor applications,
data rate is the speed at which measurement data is sent across a commu-
nication channel. One of the key considerations includes the type of sensor
and the nature of the data, which can vary significantly between different
types of sensors. For example, a Global Navigation Satellite System (GNSS)
or Inertial Navigation System (INS) may produce new data at a millisecond
rate. At the same time, a temperature or level sensor may update its data
over periods measured in seconds. Also, imagery data produced by a camera
is often updated 20 or more times per second, with each update producing
large amounts of data that must be communicated at a high speed to main-
tain flow and continuity to the viewer or real-time image analysis processors
when capturing fast-changing phenomena. Each individual sensor has speci-
fied resolution and accuracy requirements that must be supported by avail-
able bandwidth and network capacity to avoid data congestion and potential
data loss. Another aspect of this issue is the capability of the network to store
and process the incoming data, where higher data rates may require larger
storage capacities and efficient data management strategies. These issues
must also be considered within the framework of existing regulatory require-
ments and standards that dictate specific data rates to ensure interoperability
exists between instruments of different manufacturers.
8.2.3 Bandwidth
Whereas data rate is the speed at which a sensor produces data to be sent
across a communication channel, bandwidth is the maximum amount of data
that can be transferred over the data bus at a given time. Bandwidth is gen-
erally measured in terms of kilobits per second (kbps), megabits per second
(Mbps) or gigabits per second (Gbps). A higher bandwidth allows more data
to be transferred simultaneously, which can improve system scalability. It is
critical that data bus bandwidth is adequate to support substantially more
than the maximum amount of data that may be produced by the sensors inte-
grated into the system to ensure the system does not get overwhelmed with
excessive data that will bog down sensor system performance.
8.2.4 Latency
Latency is the delay in time between when sensor data is generated and when
it is delivered to the end user, be it a human being or another electronic
system whose purpose is to process the data. Lower latency allows data to
be transferred more quickly, which can improve the responsiveness of the
system. Adequate bandwidth capacity to handle anticipated sensor data rate
requirements is necessary to ensure network latency is minimized.
8.2.5 Security
Data bus security is paramount to protect sensor data from unauthorized
access and to prevent sensor system hacking and hijacking. A secure data
bus uses encryption, authentication, access control, firewalls, intrusion detec-
tion and prevention systems, secure protocols, and other security measures
to protect the data and the sensors that generate the data. Robust security
measures require a comprehensive approach that considers all aspects of the
sensor data bus application, from physical security to network protocols and
access controls. Regular updates and improvements should be made to stay

ahead of evolving security threats.
8.2.6 Flexibility
The flexibility of a data bus is a measure of how it may be adapted to dif-
ferent applications and configured to meet the specific requirements of the
application. However, unique data bus architectures have been designed to
optimize specific types of sensor data. For example, the CAN bus formed the
basis for the creation of the NMEA-2000 bus, which is designed specifically
for maritime use by electronics associated with navigation and engineering
sensors. A similar adaptation known as J1939 was also created by the Society
of Automotive Engineers (SAE) for commercial vehicle use.6 CAN proved to
be quite flexible in its design but didn’t meet the specific needs of either the
maritime or automotive communities and was adapted to each application.
8.3 LAN (Local Area Network)

A LAN is a group of computers and other devices that are connected together
within a limited area, such as a ship, and allow devices to communicate with
each other. LANs are also used to share resources such as printers, files and
Internet access, as well as to connect sensors. Several different protocols can be
used for connecting devices on a LAN including Ethernet, Bluetooth, and Wi-Fi.
LANs can be used to connect sensors on a ship in a number of ways. One
way is to use a dedicated LAN that is specifically for sensors. This LAN would
be isolated from other networks on the ship, which would help to protect the
sensors from unauthorized access. Another way to connect sensors to a LAN
is to use a wireless network. This would allow the sensors to be placed any-
where on the ship without the need for cables. However, wireless networks
are more susceptible to interference than wired networks, so it is important
to choose a wireless protocol that is well-suited for the environment.
8.4 Overview of Sensor Data Bus Standards and

Protocols
There are several specific standards that apply to sensor connectivity on ships
to ensure reliable and safe communication between sensors and the ship’s
systems. The choice of sensor data bus designation depends on the specific
requirements of each application. It is important to note that sensor connect-
ivity on ships can vary depending on the type of vessel, its systems and the
specific sensors being used. Some of the key standards include:
• Ethernet
• CAN
• NMEA-0183
• NMEA-2000
• RS-485
• USB Port
• Bluetooth
• Wi-Fi
Table 8.1 provides a rough side-by-side comparison between the several

bus and protocol standards and their characteristics described in this chapter
in an attempt to highlight the contrast between them. However, comparing
standards with different capabilities can lead to misleading conclusions and
oversimplification. The intended goal is to highlight their use in the ship-
board maritime environment and not to paint a complete picture of each
standard’s overall strengths and weaknesses. Direct comparisons are difficult
as each standard serves different purposes and operates in diverse environ-
ments. Note also that this is only a representative sample of applicable stand-
ards, and other relevant standards have been left out of this comparison.
Ethernet is a widely used technology that defines the rules and standards
for transmitting data over a LAN. This includes wired and fiber optic net-
works that support a wide range of bandwidth and data rates.
The CAN network is a type of communication protocol and network
architecture commonly used in many applications including passenger and
commercial vehicles, and in the maritime and industrial sectors. It was devel-
oped to address the need for a robust and reliable communication system for
electronic control units (ECUs) in vehicles.
NMEA- 0183 is a maritime- specific standard communication protocol
developed by the NMEA for transmitting and receiving data between marine
electronic devices. NMEA 0183 has been widely adopted in the marine
industry. It is used for interconnecting navigation equipment, sensors and
other marine electronics on recreational boats, ships of all kinds and sizes,
and a wide variety of other watercraft.
NMEA-2000 evolved from the NMEA-0183 standard to provide signifi-
cant improvements in terms of data speed, capacity and network capabilities
to allow marine electronics, sensors and systems to communicate and share
data over a single network.
RS-485 is an industrial specification that defines the electrical interface
and physical layer for point-to-point communication of electrical devices and
industrial sensors, many of which are used on ships. It specifies the elec-
trical characteristics of a balanced differential serial communication inter-
face, which allows for multi-point communication and long-distance data
transmission.
The Universal Serial Bus (USB) standard is a widely used communication
protocol and physical interface for connecting various devices to a computer
newgenrtpdf
136
Intraship and Internal Communications
Table 8.1 Comparison between Shipboard Sensor Data Bus Standards
Characteristic Standard
Ethernet CAN NMEA NMEA 2000 RS-4 85 USB 3.1 Bluetooth 5.x Wi- F i
0183
100BaseTX 1000BaseT 1GBaseT
Data Rate 100 Mbps 1000 Mbps 10 Gbps 1 Mbps 9.8 kbps 250 kbps up to 10 10 Gbps 2 Mbps 11–3 000
Mbps Mbps
Data Format Binary Binary Binary Binary ASCII Binary Various Serial Packet Binary
packet
Protocol IEEE 802.3 IEEE 802.3 IEEE 802.3 Message Sentence Message Various Master/ Master/ IEEE 802.11
Slave Slave
Compatibility Widespread Widespread Widespread Moderate Moderate Widespread Widespread Widespread Widespread Widespread
Topology Several Several Several Bus Point-t o- Star Several Low Star Star
Point
Fault Moderate Moderate Moderate High Low Low High Moderate Moderate Moderate
Tolerance
Security High High High Moderate Low High None Low Moderate Moderate
or other host devices.7 On board ships it is generally used for updating navi-
gation charts and maintenance manuals, and for servicing many types of
electronic equipment.
Bluetooth is a wireless communication technology that allows electronic
devices to exchange data over short distances without the need for physical
connections. It is used in much the same manner as USB, except that it does
not have to be physically connected for access to be achieved to a Bluetooth
device.
Wireless Fidelity (Wi- Fi) is a wireless communication technology that
allows devices to connect to a LAN and the Internet without the need for
wired connections. It is based on the IEEE 802.11 family of standards, which
defines the specifications for wireless networking.
Many standards also fall under the purview of the International
Electrotechnical Commission (IEC), which develops and publishes inter-
national standards for electrical, electronic and related technologies,
including:
• IEC 60945: Standards for regulating all equipment used in marine areas.
• IEC 61158: A series of related standards that cover aspects of RS-485
that define the fieldbus standard for digital communication in industrial
automation.
• IEC 61162 (Maritime Navigation and Radiocommunication Equipment
and Systems): A series of standards that specify the data communication
protocols used for shipboard sensors and systems.
• IEC 61162- 1: General requirements for data acquisition systems for
marine applications covering the physical layer, the data link layer and
the application layer for NMEA-0183.
• IEC 61162-2: Requirements for data acquisition systems for marine appli-
cations that use the Ethernet protocol.
• IEC 61162- 3: Requirements for data acquisition systems for marine
applications that use the CAN bus protocol and the NMEA-2000 imple-
mentation of this standard.
Details on these individual standards are provided in the following

paragraphs.
8.5 Ethernet
Ethernet is a family of wired computer networking technology and protocols
commonly used on ships for connecting devices on a LAN. First introduced
in 1980 and standardized in 1983 as IEEE 802.3 and now also referred to
as ISO/IEC8802-3, it uses physical media that includes twisted-pair copper
wiring, coaxial cable and fiber optic cable. The data rate of Ethernet varies
depending on the type of physical media and the version of the standard.
Older versions of the standard that are no longer used in modern networks
include 10Base-T and 10Base-F, which are limited to speeds up to 10 Mbps.
Relevant contemporary standards include:8
• 100BaseTX and 100BaseT4 (Cat 3 or higher unshielded twisted-pair cop-

per wiring; 100 Mbps),
• 100BaseFX (fiber optic cable; 100 Mbps),
• 1000BaseT (Cat 5 or higher unshielded twisted-pair copper wiring; 1,000
Mbps) and
• 10GBaseT (Cat 5 or higher unshielded twisted-pair copper wiring; 10,000
Mbps or 10 Gbps).
Ethernet hubs were used to connect multiple Ethernet devices in a LAN, but
these have been largely replaced by Ethernet switches due to their superior
performance and features. Ethernet hubs operate at the physical layer (Layer
1) of seven layers described in the Open Systems Interconnection (OSI) model
computer systems use to communicate over a network.9 Ethernet switches
operate at the physical layer (Layer 1) and the data link layer (Layer 2) to pro-
vide intelligent data forwarding and dedicated bandwidth to each connected
device, making them far more efficient and scalable than older Ethernet hubs.
Also used are routers used to connect multiple networks and direct data
traffic between them. A router operates at the network layer (Layer 3) of the
OSI model and plays a critical role in facilitating communication between
devices and networks in complex network environments. Some Ethernet
switches offer Power over Ethernet (PoE) capabilities, providing power to
connected devices and sensors such as IP cameras and wireless access points,
eliminating the need for separate power supplies.
The OSI and Transmission Control Protocol/Internet Protocols (TCP/IP)
are two different but related networking models associated with the Ethernet
protocol. The OSI model is a conceptual framework that standardizes how
different networking protocols should interact and communicate with each
other, while TCP/IP is a suite of specific set of protocols used to interconnect
network devices on the Internet.10 While serving different purposes, the OSI
and TCP/IP models have not entirely replaced the other. TCP/IP has become
the dominant suite of protocols used for Internet communication, while the
OSI model remains a basic reference model for understanding networking
concepts and protocol interactions.
8.6 Controller Area Network (CAN) Bus

The CAN sensor data bus is a type of serial communication bus used to
transmit data between ECUs on a ship. The CAN data bus follows a serial
communication protocol to transmit data between ECUs in vehicles and
other control applications such as ships. CAN is a serial communication
protocol that supports distributed real-time control and multiplexing for use
within vehicles and other applications. The standard defines the data format
and protocol that are used to transmit data over the bus and specifies the
Classical CAN frame format and the newly introduced CAN Flexible Data
Rate Frame format. The Classical CAN frame format allows bit rates up
to 1 Mbit/s and payloads up to 8 bytes per frame. The Flexible Data Rate
frame format allows bit rates higher than 1 Mbit/s and payloads longer than
8 bytes per frame. Also described is the general architecture of CAN in terms
of hierarchical layers according to the ISO reference model for open systems
interconnection (OSI) according to ISO/IEC 7498-1.11
In the restructured ISO 11898 series most applicable to maritime
applications:
• Part 1 defines the data link layer, including the logical link control (LLC)
sub-layer and the medium access control (MAC) sub-layer, as well as the
physical signaling (PHS) sub-layer.
• Part 2 defines the high-speed physical medium attachment (HS-PMA).
• Part 3 defines the low-speed fault-tolerant PMA.
Other parts of CAN also include:
• Part 4 defines the time-triggered communication.

• Part 5 defines the power modes of the HS-PMA.
• Part 6 defines the selective wake-up functionality of the HS-PMA.
The ISO11898-1 standard specifies the characteristics of setting up an inter-

change of digital information between modules implementing the CAN data
link layer.12 ISO11898-2 specifies the HS-PMA of the CAN, a serial commu-
nication protocol that supports distributed real-time control and multiplex-
ing for use within road vehicles.13 This includes HS-PMAs without and with
low-power mode capability as well as with selective wake-up functionality.
ISO 11898-2 has been updated in parallel with ISO 11898-1 to combine
the functions described in ISO 11898-2, ISO 11898-5 and ISO 11898-6. The
future edition of ISO 11898-2 will cancel and replace the current ISO 11898-
2:2003, ISO 11898-5:2007 and ISO 11898-6:2013.
8.7 NMEA-0 183

The NMEA 0183© Interface Standard, copyrighted by the NMEA, is used
worldwide across many industry segments.14 The data bus protocol standard
for serial data communication defines electrical signal requirements, data
transmission protocol and time, and specific sentence formats for a 4800-
baud serial data bus. Each bus may have only one talker but many listen-
ers and is intended to support one-way serial data transmission. Data is
transmitted in a series of sentences, with each sentence containing a header, a

data field and a checksum. The header identifies the type of sentence, and the
data field contains the actual data. The checksum is used to verify the integ-
rity of the data. This data is in printable ASCII form and may include infor-
mation such as time, position in latitude and longitude, speed, water depth,
heading and distance traveled.
The IEC 61162-1 standard is a technical specification that defines the
physical and electrical characteristics of the NMEA-0183 data bus. It is still
widely used on ships, and it is supported by a wide range of devices including
GNSS receivers, depth sounders, autopilots, chart plotters, autopilots and
engine control systems.
8.8 NMEA-2 000

The NMEA 2000® standard, a registered Trademark of the NMEA, contains
the requirements of a serial data communications network to interconnect
marine electronic equipment on vessels.15 According to NMEA, the standard
describes a low- cost, moderate capacity bi-directional, multi- transmitter/
multi-receiver instrument network to interconnect marine electronic devices.
Based on the CAN, it is multi-master and self-configuring, and there is no
central network controller. Equipment designed to this standard can share
data, including commands and status, with other compatible equipment over
a single channel.
NMEA-0183 and NMEA-2000 are two different data buses that both are
used to communicate between marine electronics devices. NMEA-0183 is a
legacy standard that has been around for many years, while NMEA-2000 is a
relatively more recent standard offering many advantages over NMEA-0183.
These include higher data rates, a more efficient binary data format, a more
versatile bus rather than a star topology, greater fault tolerance and greater
security.
The IEC 61162-3 standard is a technical specification that defines the phys-
ical and electrical characteristics of the NMEA-2000 data bus. The standard
also defines the data format and protocol that are used to transmit data over
the bus.
8.9 RS-4 85
RS-485 is an industrial specification that defines the electrical interface
and physical layer for point-to-point communication of electrical devices
over long cabling distances in electrically noisy environments.16 It is used
onboard ships as a means to connect many different types of sensors for
cargo and ship infrastructure monitoring. A serial communication standard
that defines the electrical characteristics of a balanced differential voltage
interface, it can support up to 32 devices connected to the same ISO 8482
bus. The RS-485 data bus protocol is also known as Telecommunications

Industry Association (TIA)- 485, Electronic Industries Alliance (EIA)- 485
and International Standards Organization (ISO) 8482. Specifications include
electrical characteristics of RS-485, such as voltage levels of ±5 V, ±12 V, or
±24 V; current levels and the impedance of the bus.17 The standard specifies
timing requirements, including bit rate, start and stop bits, and parity bit. It
also describes the physical medium characteristics for twisted pair multipoint
interconnections in either two-wire or four-wire network topology, supports
binary and bi-directional signal transfer, and defines the electrical and mech-
anical design of the endpoint system branch cables and the common trunk
cable, which may be up to 1,200 m in length. Also specified are the compo-
nent measurements of the integrated type generators and receivers within the
endpoint system. The data signaling rate can go up to 12.5 Mbps. RS-485
can support topologies that include point-to-point, multi-drop and multi-
point configurations.
8.10 Universal Serial Bus (USB) Ports and Devices

USB ports and the devices that plug into these ports on ships connect many
peripherals to a ship’s computer system, including keyboards, mice, printers
and storage devices. They are also used to update the software on a ship’s
computer system and for data transfer between a ship’s computer system and
other devices such as laptops and external hard drives for backing up data
and sharing files. Many manufacturers of marine electronics are also incorp-
orating USB ports into their products.
USB devices are typically “plug-and-play”, where they can be connected
and disconnected without requiring a system restart. The USB interface pro-
vides power to connected devices, allowing many peripherals to draw power
directly from the USB connection without the need for additional power
sources. Note also that the use of USB thumb drives by crewmembers and
outside service and maintenance personnel poses significant exposure to
cyber criminals who can use flash drives to attack the ship’s computer sys-
tems and networks. The USB standard is developed and maintained by the
USB Implementers Forum (USB-IF), a nonprofit organization that promotes
and supports USB technology.
8.11 Bluetooth
Bluetooth is a wireless technology that allows devices to communicate with
each other over short distances.18 It performs many of the same functions
as a USB port and also provides the capability to connect sensors on ships.
It is also susceptible to the same risks as USB in providing access to attack
ship’s computer systems and networks through hacking and cyberattacks.
Operating in the unlicensed 2.4 GHz radio frequency band for industrial,
scientific and medical equipment, its low-power transmissions do not need

to be licensed. Bluetooth is managed by the Bluetooth Special Interest Group
(SIG), a consortium of various companies.
8.12 Wireless Fidelity (Wi-F i)

Wireless Fidelity, or Wi-Fi, represents a set of standards developed by the
Institute of Electrical and Electronics Engineers (IEEE), collectively known as
IEEE 802.11.19 These standards include IEEE 802.11a, IEEE 802.11b, IEEE
802.11g, IEEE 802.11n, IEEE 802.11ac and IEEE 802.11ax, which represent
operations on the 2.4 GHz and 5 GHz frequency bands with speeds ranging
between 2 Mbps and 1.3 Gbps. The 2.4 GHz transmits data at a slower speed
than 5 GHz but does have a longer range than 5 GHz. The 5 GHz trans-
mits data at a faster rate, but it has a shorter range because it has a higher
frequency.
References
1 MSC 67/22/Add.1, ANNEX 17. Resolution MSC.64(67), 4 December 1996.
Add 1. ANNEX 1. Recommendation on Performance Standards for Integrated
Bridge Systems (IBS).
2 SOLAS chapter V adopted in December 2000.
3 Joint Committee Guides Metrology: Evaluation of Measurement Data-Guide
to the Expression of Uncertainty in Measurement (GUM 2008), 2008.
4 H.Y. Teh, A.W. Kempa- Liehr, and K.IK. Wang. (2020) Sensor data
quality: A systematic review. Journal of Big Data 7, 11. https://doi.org/10.1186/
s40537-020-0285-1
5 Data Rate. PCMag Encyclopedia. PC Magazine. www.pcmag.com/encyclope
dia/term/data-rate
6 SAE J1939 –Recommended Practice for a Serial Control and Communications
Vehicle Networks. J1939 Digital Annex J1939DA_ 202305. Society of
Automotive Engineers (SAE). 10 May 2023. www.sae.org/standards/content/
j1939da_202305/
7 USB4® Specification V2.0. USB Implementers Forum, Inc. (including trade-
mark). 30 June 2023. www.usb.org/document-library/usb4r-specifi cation-v20
8 Ethernet Standards and Protocols Explained. Computer Networking Notes,
2023. www.computernetworkingnotes.com/networking-tutorials/ethernet-
standards-and-protocols-explained.html
9 Keith Shaw. The OSI Model Explained and How to Easily Remember Its 7
Layers. 14 March 2022, Network World. www.networkworld.com/article/
3239 6 77/ t he- o si- m odel- e xplai n ed- a nd- h ow- t o- e as i ly- r emem b er- i ts- 7 - l ay
ers.html
10 Mary E. Shacklett. What Is TCP/IP? TechTarget, Networking. July 2021. www.
techtarget.com/ searchnetworking/definition/TCP-IP
11 ISO/ IEC 7498- 1:1994. Information Technology –Open Systems

Interconnection –Basic Reference Model: The Basic Model. https://webstore.
ansi.org/standards/iso/isoiec74981994
12 ISO 11898- 1:2015(en). Road Vehicles, Controller Area Network
(CAN) –Part 1: Data Link Layer and Physical Signalling. International Standards
Organization (ISO), 2015. www.iso.org/obp/ui/#iso:std:iso:11898:-1:ed-2:v1:en
13 ISO 11898- 2:2016(en). Road Vehicles, Controller Area Network (CAN) –
Part 2: High-Speed Medium Access Unit. International Standards Organization
(ISO), 2016. www.iso.org/obp/ui/#iso:std:iso:11898:-2:ed-2:v1:en
14 NMEA 0183- Version 4.11. Serial Data Networking. National Marine
Electronics Association (NMEA). November 2018. www.nmea.org/nmea-
0183.html
15 NMEA 2000® Version 3.000. The Modern Marine Interface Standard. CAN
Based Serial Data Communications. Vessel Systems Integration. April 2022.
www.nmea.org/nmea-2000.html
16 Jason Kelly. RS-485 Serial Interface Explained. CUI Insights. CUI Devices.
www.cuidevices.com/blog/rs-485-serial-interface-explained
17 ISO/IEC 8482:1993. Telecommunications and Information Exchange between
Systems, Twisted Pair Multipoint Interconnections. International Standards
Organization (ISO). www.iso.org/standard/20954.html
18 Bluetooth Specifications and Documents. The Building Blocks of All Bluetooth
Devices. Bluetooth Special Interest Group (SIG). www.bluetooth.com/specifi
cations/specs/
19 IEEE 802.11- 2020. IEEE Standard for Information Technology
Telecommunications and Information Exchange between Systems. Local and
Metropolitan Area Networks: Specific Requirements. Part 11: Wireless LAN
Medium Access Control (MAC) and Physical Layer (PHY) Specifications.
Institute of Electrical and Electronics Engineers, 2020. https://standards.ieee.
org/ieee/802.11/7028/
Chapter 9
Cybersecurity and Ship Sensors
During a conference in 2020, Robert Rizika of the NavalDome cyber defense

company reported that cyberattacks on maritime industry operational tech-
nology (OT) systems increased by 900% from 2017 to 2020, with many
more attacks going unreported and the trend continuing to increase.1
Reported cyberattacks among maritime industry stakeholders have increased
from 43% in 2018 to 74% in 2022, according to Ford Wogan of the Jones
Walker’s Maritime Practice Group.2 The International Association of Ports
and Harbors (IAPH) has been quoted as stating the maritime industry glo-
bally suffered a fourfold increase in cyberattacks between February 2020 and
May 2020 alone.3 This is a significant issue that is ever-increasing in import-
ance and impact on maritime operations.
Many intrinsic cyberattack vulnerabilities must be overcome in data, net-
work hardware, software and sensor-intensive critical ship infrastructure
utilizing communication networks that service transportation systems. The
specific focus of this chapter is on ships and by extension, the ports they call
upon where cyberattacks were identified by the U.S. Secretary of Homeland
Security as “the most significant threat to port infrastructure”.4 Disruptions
related to cyberattacks can affect tens of thousands of ships and thousands of
seaports, many of which are vital to energy infrastructure. Cyberattacks can,
and often do, propagate throughout the worldwide logistical supply chain.
This problem will continue to increase in importance as port digitalization
expands through greater integration of smart devices and Internet of Things
(IoT) technology that often includes wireless communications.
The hundreds to thousands or more sensors onboard a typical ship are
of diverse vintages and technologies and are designed and manufactured by
many different companies. Sensors are maintained by even more companies
and operated by a third group of companies; all following different business
models and security practices. Smart and IoT sensors further complicate the
problem by promoting widespread connectivity that often takes shortcuts
across multiple security layers to perform their jobs. Yet the unique char-
acteristics of numerous and diverse sensors and the wide range of functions
and applications they serve are often ignored and relegated to a level of lesser
Doi: 10.1201/9781003377900-9
Cybersecurity and Ship Sensors 145
importance than cyber systems in general. Indeed, sensors are designed to

share data, but sensor communications are often unencrypted and fairly easy
to intercept. This is a mistake.
Maintaining safe and secure critical infrastructure across the sensor envir-
onment is essential to ensure efficient, uninterrupted and uncompromised
vessel operations. This issue encompasses onboard sensors and extends
throughout the entire sensor ecosystem, including off-vessel users, company
offices and remote control centers. This also extends to government agencies
with which ships interact to fulfill regulatory requirements and the port facil-
ities that comprise the land-ship interface where routine and essential ship
operations take place.
The challenges of ensuring sensors remain operational, viable and free
from interference and disruption embrace many of the same principles
devised for cybersecurity in general. The complexity of cyber-physical-data
critical sensor infrastructure can also hide vulnerabilities to small and seem-
ingly unimportant changes and variations within their makeup that can make
them susceptible to catastrophic and destabilizing behavior such as cascading
failures and open back doors to infiltration. The ability to detect such sensi-
tivities and secure against attacks to trigger them is a key factor when imple-
menting effective sensor cyber security.
Cyberattacks on large information technology infrastructure often make
headlines when the identities and personal data of millions of people and
organizations are stolen. Likewise, the immediate needs of communities,
businesses and households are adversely affected when critical operational
technology, such as electricity generation and energy distribution infrastruc-
ture, is compromised. However, the effects of sensor compromise may often
be subtle, and long-term and manifest themselves indirectly as a result of
breakdowns and similar events that are hard to trace to their source. A prime
example occurred in 2010 when thousands of centrifuges at the Natanz fuel
enrichment plant in Iran were destroyed. This was accomplished through the
infiltration of computer systems by the Stuxnet computer virus that changed
sensor signals to make it appear operations were nominal yet the centrifuges
were actually spinning far above their normal speed and tearing themselves
apart.5
Similar analogies may be made to shipboard sensor systems. Some examples
of the possible consequences of cyberattacks include the ability to:
• Incapacitate onboard computers to make sensors inoperable or blind.

• Spoof GNSS signals on cue to ground a vessel in the middle of a critical
waterway.
• Corrupt electronic navigation charts with false data and soundings to
obscure hazards to navigation.
• Hijack ship and vehicle controls.
146 Cybersecurity and Ship Sensors
• Implant descriptions of unauthorized people within smart cameras

using facial recognition to provide crewmember and passenger access to
restricted areas.
• Disable engine pollution monitors and controllers, resulting in spills and
discharges of hazardous chemicals and pollutants with great environ-
mental impact.
• Degrade the resolution and effectiveness of cameras and other sensors.
• Render object and event recognition capabilities of sensors ineffective or
useless.
• Weaponize civilian ships for nefarious purposes.
Areas of commonality between information technology (IT) and OT sys-

tems and cybersecurity, in general, are described to provide context and to
draw attention to topics unique to ship sensor functions and installations.
The nature of cyberattacks is also covered providing examples of how the
attacks are committed and the potential consequences of these attacks. Also
discussed is the muddy grey cyber line that distinguishes where the ship ends
and the port begins, especially in current times when constant contact is rou-
tinely maintained between ships and shoreside organizations through satel-
lite and terrestrial 5G communication links. Details are provided regarding
critical ship and associated port infrastructure that require protection and the
means through which this may be accomplished. The chapter concludes by
citing resources describing precautions that may be taken to help minimize
the potential for cyberattacks on sensor systems and to limit their severity
when they occur.
Please note that while this discussion specifically addresses sensor systems
onboard ships, many of these same principles apply directly to a wide variety
of sensor applications and related networks that span the maritime industry
throughout the ports themselves and to transportation in general. These
concepts can also be readily adapted to meet sensor cybersecurity needs of
telecommunications, energy production and distribution, pharmaceutical,
medical and many other industries.
9.1 Aspects of Cybersecurity Unique to Sensor Systems

Cyberattacks against shipborne sensor systems often take a fundamentally
different form than many types of attacks routinely committed against com-
puter networks in general in terms of their form and function, and the conse-
quences of their occurrence. Sensor networks, by their very nature, are likely
to have relatively few active human users and real-time interactions with
other Information Technology (IT) and Operational Technology (OT) sys-
tems, and they should operate through one or more firewalls that can detect
the legitimacy of access requests. They should also be well-segmented and
isolated by virtue of access being physically or electronically limited to only
a relatively few people for maintenance and system upgrades. Vulnerabilities

encountered through outbound communications from a sensor network
are also likely to be greatly minimized as the destinations for sensor data
and analytics products should be well-defined and restricted. These charac-
teristics significantly hamper several of the more prominent types of direct
cybersecurity attacks, such as phishing for data through malicious emails,
attacks resulting from visiting legitimate websites with weak security that are
infected, and clone websites pretending to be legitimate that are also infected
with executable scripts. Brute force attacks attempting to guess login cre-
dentials are also not likely to succeed since multiple attempts to gain access
can easily be detected and cause the user to be locked out. Also unlikely
are denial-of-service (DoS) attacks originating from external IT sources that
attempt to flood and overwhelm a sensor network server with traffic until it
can no longer respond or crashes, thereby preventing access to authorized
users. However, vulnerabilities to DoS attacks still exist via the sensors, the
people who maintain them, and the crews and passengers that are present on
the ship.
Much more likely to occur are insider attacks resulting from the introduc-
tion of malware through software and firmware updates accomplished using
USB devices as well as “trusted” users whose end-point systems are them-
selves compromised. Such circumstances allow viruses and other malicious
code to be passed along to computer server(s) and/or sensor controller(s) that
interact with the server network and directly with the sensors. The malware
may act immediately upon insertion or may lie dormant for long periods
of time, awaiting the occurrence of predefined circumstances before acting.
These circumstances may include a combination of unique operating condi-
tions exhibited by an engine or other ship’s machinery, a specific date and
time, a vessel’s position defined by latitude and longitude, and recognition of
known unique visual or digital signature(s) introduced externally to the ship
through an onboard camera.
Opportunities for breaching security by monitoring data communicated
between sensors and their controllers and between the controllers and the
server(s) are also made available by middleware that functions as a hid-
den layer between an operating system and sensor software applications.
Middleware increases the possibility to insert ransomware that targets known
vulnerabilities, or malicious software that models the sensor environment to
devise methods on its own to detect and attack vulnerabilities within the net-
work itself and/or the individual sensors operating within the network.
Another problem in maintaining cybersecurity is the need to ensure con-
formance and interoperability of methods and processes to provide users
with resilient and high-performance network capabilities that include quality
of service, prioritization and the avoidance of service pre-emption from both
accidental and nefarious sources. Cybersecurity technologies are relatively
new and many products offered by the various vendors and different service
providers often exhibit poor communication and interoperability at both the

endpoint applications and at the servers that manage these applications. Such
conditions provide multiple vulnerabilities and opportunities to infiltrate,
disrupt and otherwise compromise sensor operations.
9.2 Cyberattacks and Sensor Systems

Several examples of cyberattacks on sensors used onboard ships are described
in this paragraph. It must be noted that other industries more advanced in
using smart sensors and automation technology are also experiencing this
same problem. Examples are cited from which lessons can be learned and
ships can be protected in future practice. An attempt is made to provide
coverage across a range of sensors and sensor types most commonly found
on ships and maritime vehicles. These examples are provided to initiate dis-
cussion through which further action may be taken to help ensure incidents
such as these do not occur or are minimized in their effect.
9.2.1 Engineering Sensors

Various engineering functions and measurable parameters discussed in
Chapter 2 are rapidly evolving with the development in the commercial
marketplace of vessel control systems for local and remote monitoring and
control of ship infrastructure such as engines and propulsion, generators,
batteries, steering, environmental monitoring, alarms and many other sys-
tems throughout the vessel. These systems form the core OT to operate the
vessel itself, and any attempts to interrupt their normal operations may place
the crews, passengers, cargo and the ship itself in jeopardy. The following
incidents and events help to illustrate some essential points on this topic:
• A plan was identified to interfere with normal seawater ballast operations

necessary to maintain ship proper trim, stability and structural integrity.6
This included a scheme to take control of the ballast pumps remotely by
sending commands via a satellite link to cause problems that could result
in the sinking of the ship or other significant and irreparable damage to
the vessel.
• Vulnerabilities in a ship gas turbine power control system were revealed
by several methods through which it was possible to gain access to and
tamper with various control parameters.7 The scenario identified an
onboard distributed control system connected to a remote center via a
satellite network. This single communications link was connected to a
complex network structure with a large data flow that was extremely vul-
nerable to several cyberattack methods. These included viruses and DoS
attacks that could result in abnormal temperature and pressure changes
in the gas turbine and may lead to damage to system components and the
complete shutdown of the turbine. This study ultimately resulted in the

development of a method for risk assessment that can detect the vulner-
abilities, identify potential attack mechanisms and take the appropriate
defensive measures.
• A comprehensive analysis of vulnerabilities to systems responsible for ship
power generation and propulsion equipment identified numerous threats
of cyberattacks, many of which can result in a high impact on ship opera-
tions that were also highly likely to occur.8 These include falsifying data
regarding engine lubrication, tampering with engine commands, inter-
ruption of engine automation system operations, data integrity violations
that can cause malfunctions in critical operations such as rudder control,
alteration of fuel consumption data, falsification of equipment mainten-
ance data and compromise of handling systems resulting in damage or
loss of cargo and interruption of emergency handling processes.
• Several types of vessels, including container ships, tankers, and cruise
ships, were examined to illustrate the ease with which taking over con-
trol of various ship’s systems could be accomplished. The participants in
this experiment were able to disable entre systems as well as compromise
engines, pumps, machinery, steering gear controls, overwrite signals com-
ing from ballast sensors, pumps and fuel level sensors.9
9.2.2 Navigation Sensors

The complete range of navigation sensors described in Chapter 3 are also
valuable targets for cyberattacks due to their essential nature in ensuring the
safety of navigation. The literature is filled with examples of how potential
and actual attacks on sensors can at best create confusion that may require
immediate corrective action on the part of watchstanders; and at worst cost
lives, loose cargo and sink ships with dreadful environmental and supply
chain consequences. The following are just a few examples describing cyber-
attack attempts and vulnerabilities with navigation sensors.
• A comprehensive cyber risk assessment was performed to determine

the vulnerability of the ship’s Electronic Chart Display and Information
System (ECDIS).10 An outdated Windows operating system was discov-
ered on the computer where ECDIS was installed, providing little protec-
tion against cyberattacks.
• A report of efforts to compromise ship systems described their ability to
attack a ship's navigation system and radar and take over the steering of
the ship.11 System penetration was achieved by sending an email to the
captain’s computer that enabled access to ECDIS. An attack was simu-
lated at night through a narrow canal altering the ship’s course while con-
tinuing to provide normal instrument display indications to the Officer of
the Watch, effectively running the ship aground.
• Additional information was obtained regarding the above experiment

that targeted the ship’s radar system using the Local Area Network (LAN)
that is also connected to ECDIS, the bridge alert system and the voyage
data recorder (VDR).12 The results included being able to eliminate radar
targets from the screen while the system appeared to be working perfectly.
• Numerous attacks on Global Navigation Satellite System (GNSS) ser-
vices have been reported worldwide, particularly in and around Russia,
Korea and in the Mediterranean, Red and Baltic Seas. These efforts have
resulted in DoS and spoofing incidents that render satellite positioning
indications useless or worse, deceptive so that a ship could be lead astray
for nefarious purposes.
• A spoofing attack was made on a 213-foot yacht using the U.S. Global
Positioning System (GPS) version of GNSS in the Mediterranean.13 The
experiment took place approximately 30 miles off the coast of Italy while
en route from Monaco to Rhodes, Greece, as the yacht sailed in inter-
national waters. False GPS signals that were not distinguishable from
authentic signals covertly overpowered the ship's navigation system until
control of the ship’s navigation system was achieved. Subtle maneuvers
were then made to position the yacht a few degrees off its original course
to follow a parallel track hundreds of meters from its intended one.
Throughout this time, the chart display showed a straight-line path being
followed while the ship’s wake formed a pronounced curve, showing that
the ship had turned.
• Communications with a U.S. RQ-170 drone aircraft were jammed by
Iran, where they were also able to reprogram their GPS coordinates to
land in Iran and be captured rather than land at their intended location
in Afghanistan.14 This same exact scenario could also be used to hijack
a ship and divert it into the territorial waters of a hostile country or the
hands of pirates.
• Automatic Identification Systems (AIS) have traditionally been easy to
hack and are a primary sensor targeted when attempting to disguise ship
locations and information. AIS broadcast data includes vessel type, name,
position, course and speed, along with other information useful to nearby
vessels, vessel traffic services (VTS), and port authorities. It is communi-
cated using unencrypted Very High Frequency (VHF) radio signals lack-
ing internal authentication and integrity checking of any data.15 Pirates
often use AIS tracking to target certain vessels and carry out malicious
activities, including shutting the system off while underway to hide their
presence. There are many incidents where the physical locations of ships
did not correspond with their positions as indicated on AIS, and were
off sometimes by hundreds or even thousands of miles. The problem is
especially prevalent with naval vessels, where between August 2020 and
July 2021, almost 100 naval vessels with suspected false AIS tracks were
detected.16 This is also a major problem with commercial vessels.
• VDRs are targets for attack as they record data from multiple sensors,
including speed, direction, position and radar images in addition to audio
recordings and other data that are indispensable to help identify the cause
of maritime incidents. One incident was reported in India where VDR
data from a cargo ship that hit a fishing vessel was erased after an infected
USB stick was inserted to retrieve the data.17 An analysis of another VDR
cited in the same source also detected weak encryption, an insecure mech-
anism for user authentication and various service vulnerabilities.18 This
could lead to modifying data to, for example, delete audio conversations
from the bridge, delete radar images, or alter speed or position readings,
and even spy on a vessel’s crew as VDRs are directly connected to micro-
phones located on the bridge.19
• Video cameras are appearing on many ships as an essential means to
monitor engine rooms, cargo holds, closed spaces and other locations that
are generally unoccupied and sensitive areas where security is an issue.
Many of these cameras are connected to networks and are IoT devices
containing vulnerabilities that could let attackers watch live camera
feeds, create botnets, or use hacked devices as a stepping stone to further
attacks.20 Another problem is that many cameras used on tankers, grain
carriers and other potentially hazardous surroundings do not meet safety
requirements for use in explosive environments.21 Such characteristics can
be exploited to cause abnormal operations that may overheat the camera
and cause an explosion.
• Satellite communications have traditionally been a high-value target for
attack. Weak encryption and old IT equipment pose key vulnerabilities.22
Also, if an individual IoT device is not encrypted or every stage of data
communication is not properly protected, an entire network of connected
devices can be manipulated. Several incidents identified in this paragraph
originated through the interception and manipulation of satellite commu-
nications. Key components to satellite security are the ground networks
that move data and links that transmit and receive satellite data through
which back door assaults can take place.23 Satellites are not useful if their
access and performance are not assured. This has been a significant tac-
tical target for both sides fighting the war in Ukraine.
• Maritime Very Small Aperture Terminals (VSATs) used for two-way real-
time data communication via satellites present cyber threats to vessels due
to the value of the data they transmit and their role as attack vectors by
providing access to vessel IT and OT infrastructure.24 Cyberattacks on
vessels can be executed by discovering and exploiting vulnerabilities in a
sequential manner by first penetrating telecom equipment and taking con-
trol of the IT segment, finding problems in segmentation and accessing the
OT layer and controlling OT equipment and arbitrarily intercepting and
modifying Transmission Control Protocol (TCP) sessions enabling man-
in-the-middle and denial of service attacks against vessels at sea.
• Sensor data busses such as those conforming to NMEA 0183/2000, CAN

and other standards are susceptible to both physical attacks by insiders
as well as software attacks that can disable them and cause the sensors
and displays to freeze and/or go black.25 A problem with the bus itself
may cause every instrument dependent on the bus to malfunction. Attacks
may be accomplished by physically connecting to available ports (e.g.,
USB, PS2, serial, etc.) to insert malware, using stolen credentials, scanning
the network topology, physically tampering with a device and various
other means.26 Spoofing wireless communications using Wi-Fi, sniffing
network traffic between sensors, deleting files from a compromised sensor
or controller and sending improper commands are several methods used
to achieve digital attacks on instrument buses.
• Environmental sensors are being integrated into many current unmanned
and autonomous maritime surface vehicles and future manned vessels
that are presently in use and being tested on autonomous trucks, busses
and automobiles. These include Light Imaging Detection and Ranging
(LIDAR), millimeter radar (mmRadar), acoustic (e.g., ultrasound) and
other types of sensors. Numerous attempts to attack such systems have
been detected and countermeasures developed that will prove useful in
future ship cyber defenses.27
• Weather sensors generally do not garner a great deal of press coverage as
targets for cyberattacks, but their significance onboard ships cannot be
overstated. This is especially true when automated ship control systems
use wind velocity, direction and other factors to help maintain a stable
keel while underway. Weather sensors now incorporate many new types
of technologies without moving parts to perform their function when
compared to past-generation devices and are often IoT-capable. Such sen-
sors are especially susceptible to ultrasound and laser interference, and
they can be exploited as a target to create an entry point to infiltrate
onboard networks.
9.2.3 Crew, Passenger and Cargo Sensors

The list of shipboard sensors can be extended to those used to monitor crew
and passenger access and locations, and cargo as described in Chapters 4 and
5. The principles and tactics associated with cyberattacks on these sensors
are similar to those used against engineering and navigation sensors.
• Facial recognition cameras, fingerprint readers, iris scanners, and other

biometric security methods are increasingly being used to limit crew
member and passenger access to secure places onboard ships. These
devices have been shown to be vulnerable to several modes of attack.28
These include attacks via USB and other active ports to update device
firmware and insert malicious software, and accessing a network that
leads to the IP camera device to monitor network traffic between the

device and the server. Also found possible was the ability to impersonate
a connected access control device and send the server data it would rec-
ognize and validate.
• Discrete and IoT temperature, smoke, explosive vapor, and other sensors
placed within inhabited, cargo, and sealed spaces are often interconnected
via Bluetooth, LAN, and other networks that suffer many of the same
vulnerabilities as all networks in general and can be used as gateways to
infiltration. However, additional precautions need to be taken to ensure
physical obstructions are not placed in close proximity to sensors so as to
inhibit their normal operation by isolating them from the properties being
sensed and interrupting network connections by reducing signal strength.
The placement of sensors beyond their useful range has also resulted in
service interruptions and disconnections from the network causing unre-
liable temperature measurements.29
• Cargo container monitoring criteria have changed significantly since 2019
with the adoption of updated sea carrier Minimum Security Criteria that
require extensive use of security cameras, vessel monitoring and other
requirements and best practices.30 Many of these strategies are imple-
mented through the use of smart sensors and IoT devices subject to the
same vulnerabilities previously cited. This can help facilitate attacks that
can eliminate individual containers from being tracked by corrupting
databases, disguising their movements by corrupting GNSS signals and
placing them in false locations, and changing their contents on manifests.
• The use of infrared cameras and perimeter Sonar may help to detect and
track adversaries who attempt to overcome security precautions and
evade ship defenses. The heat signature of a human in the water can be
easily detected using infrared cameras, although the intensity of the sig-
nature may be reduced somewhat as a result of the swimmer wearing
an exposure suite. Sonar can detect a swimmer on the surface by day or
night, someone swimming just below the water’s surface using snorkel
gear, a diver approaching underwater from depth using open or closed-
circuit breathing apparatus, as well as diver delivery vehicles. Several dif-
ferent manufacturers offer sonar products of this nature that generally
have a range of up to 1,000 meters at depths of 50 meters.
9.3 Where the World Ends and the Ship Begins

As ships become more automated and the number of onboard sensors
increases, their vulnerability to cyberattacks dramatically escalates while in
port and exists as part of the port through the establishment of direct physical
and electronic connections with port infrastructure. This is due primarily to
continuing port automation that also increases network vulnerability as more
equipment and sensors are added and with greater use of remote-controlled
port machinery such as gantries, cranes and bridges. Another major con-
tributor to the problem is access to the ship by technicians, maintainers and
port employees. Nefarious interference with port operations, systems, appar-
atus and devices comprising port infrastructure can result in death or injury
to port employees, ships’ crews and passengers; destruction of property and
cargo, loss or compromise of sensitive data, extended cargo delivery sched-
ules and/or environmental hazards. This is aggravated as a result of attack
strategies that can evade the present generation, less than real-time detection
capabilities of disruption to landside and shipboard facilities. The problem
extends to port state control as implemented through port control systems
(PCS), cargo control systems (CCS), and related means to connect multiple
systems across different government, public and private organizations for the
secure and intelligent exchange of information.
Cybersecurity must not only focus on ship infrastructure, but also on port
of origin and destination segments of the entire shipping process. Figure 9.1
illustrates these segments and the additional shoreside components and pro-
cesses that can be vulnerable to cyberattacks through physical and electronic
methods.
This cybersecurity environment is very complex and consists of mul-
tiple segments related to logistics, engineering, security and management
elements representative of cargo transported, compliance, safety and other
functions. This also includes physical port and vessel facilities in terms of
cyberinfrastructure and sensor-based systems through which maintenance is
performed, security is sustained along with many other operations, and upon
which ships rely to safely and efficiently transport cargo between ports. By
focusing specifically on only the ports and ships themselves, rather than the
entire shipper-to-consignee process, including handling, haulage and ware-
housing; the problem is reduced to a more manageable size and complexity.
The physical interfaces between the port and the ship include wire, fiber-
optic cabling and hoses to provide electrical power, communications and
utility services in addition to gangways and ladders across which are trans-
ported physical sensors, new and replacement parts and systems, and various
media to update network and sensor systems by technicians and maintainers.
Figure 9.1 Port and Ship Infrastructure Segments of the Entire Shipping Process.
This also includes any devices and equipment brought onboard by crewmem-
bers and passengers, and devices that may be embedded into cargo and/or
their containers.
Digital interfaces while the ship is in and close to the port include Wide
Area Network (WAN) and wireless communications provided by the port
itself and its immediate vicinity by the local community. In addition, all ship-
board capabilities in terms of satellite, radio and terrestrial cellular commu-
nications continue to be available. Significant changes occur in both the port
and ship cybersecurity environment when shoreside connections are slipped
and the ship leaves port, and when a ship arrives in port and shoreside con-
nections are again established. The only remaining shipboard capabilities
that should exist can be limited to satellite, radio and terrestrial cellular com-
munications, as well as cellular devices in the personal possession of crew-
members and passengers. However, there are exceptions.
While onboard sensors are operational, they can be subject to a wide
variety of external stimuli that provide unconventional and unintended,
two-way access to the ship’s digital infrastructure. These sensors include
cameras, weather instruments and many other devices accessible via light,
laser, acoustic, microwave, radio frequency (RF) sources and other means
of stimulus that may contain data, information and intelligence embedded
within their transmissions. All such means of access must be considered as
part of the infrastructure.
9.4 The Attack Surface, or the Protect Surface?

Much of the literature discussing cybersecurity has traditionally discussed the
attack surface against which cyberattacks are made, and the need to reduce
this to as small a footprint as possible. One definition describes it as the total
number of all possible entry points for unauthorized access into any system,
including vulnerabilities and endpoints that can be exploited.31 This also rep-
resents the entire area of an organization or system that is susceptible to
hacking. The smaller the attack surface, the easier it is to protect. However,
most maritime infrastructure is already massive and as new technologies are
introduced the attack surface continues to expand. Also, with the increasing
use of IoT devices and sensors the attack surface has expanded exponentially.
The U.S. Department of Defense (DoD) has recently released its Zero Trust
Strategy and Roadmap, which anticipates current and future cyber threats
and attacks that go beyond the traditional perimeter defense approach.32
Rather than looking at the attack surface from a high level, in Zero Trust,
the exact nature of what is needed to be protected is defined as the protect
surface, which is the smallest possible reduction of the attack surface.33 It is
defined based on the data to be protected, application usage of sensitive data,
asset vulnerability and services that can be exploited to disrupt operations.
The protect surface is orders of magnitude smaller than the overall attack
surface and it is always knowable. Firewalls and other controls are moved as
close as possible to the protect surface rather than the perimeter at the attack
surface, where it is decidedly further away from what needs to be protected.
This is much like identifying the protect surface as a person’s lungs to place a
“firewall” to battle lung cancer rather than defining the attack surface as the
skin at the body’s perimeter to keep the cancer from entering the body itself
(probably a bad analogy, but I am not a medical doctor so I will use this as
my excuse!). In this way, it is possible to determine what traffic moves in and
out by a very small number of users or resources that actually need access to
sensitive data or assets. As with the attack surface, organizations must con-
stantly monitor their protect surface to identify and block potential threats
as quickly as possible. However, in theory, the smaller footprint makes this
process more manageable.
Actual methods to model the protect surface are still being developed.
With implementation of distinct DoD Zero Trust capabilities and activities
anticipated by 2027, there remains little time to determine how this will actu-
ally be accomplished.
9.5 Ship and Port Cyber Infrastructure

It is necessary to create a model to identify, assess and specify attributes and
characteristics comprising port and ship infrastructure to be able to assess
what actually must be protected. The focus should be on the design and
implementation of the overall architecture through which networks char-
acterize cargos, sense and perceive the physical port and maritime environ-
ments as well as internal capabilities to fulfill network requirements. This
includes how network data and the physical and environmental characteris-
tics they represent are understood, and the fusion of these data to achieve a
goal of comprehensive, continuous, near-real-time (in fractions of a second)
assessment of potential threats through reduced-order and requirements-
based variability modeling. Considering the contrasts between the concepts
of attack surface and protect surface modeling described in the previous
paragraph, it is easy to see the differences in the magnitude of effort required
to perform these tasks.
The ship and port physical and digital infrastructure that interfaces with
the ship must be considered part of one entity while the ship is in port.
However, the scope of infrastructure decreases dramatically when the ship
gets underway and the number of entry points to the ship are greatly reduced.
9.5.1 Ship Infrastructure

Computer networks onboard ships are generally comprised of the same types
of equipment and software found in ports. They are tailored to the specific
requirements needed to perform their missions in addition to moving their
passengers and cargo from port to port wholly within the confines of the hull,
deck and superstructure. However, unique to the ship architecture are com-
puter systems to support sensor activity that should be segmented to isolate
these functions from general IT operations.
Activities are undertaken through which data, information, algorithms
and software updates pass to and from company IT offices, ports, govern-
ment offices, regulatory agencies and the ship. These activities include freight
and customs, maintenance data and diagnostics, navigation information,
ship sensor data, software and software patches, middleware, and firmware.
9.5.1.1 Ship Physical Infrastructure

The physical ship infrastructure is comprised of the computer networks and
all the devices and interfaces through which access to ship network sys-
tems may be achieved. There is great commonality between ship and port
IT systems functions including servers, computers, laptops, mobile devices
and tablets, modems, routers, switches, controllers, peripherals including
disk drives, memory storage devices and monitors; printers, speakers, sen-
sors and cameras. Segmentation of shipboard networks can often be found
by function based upon cargo loading/offloading and management, engin-
eering, propulsion and maintenance; bridge communications and navigation,
administrative and crew welfare, access control, passenger services and man-
agement, and private and public networks. Included are physical interfaces
that are used by maintenance and IT teams to download maintenance data,
voyage data, sensor data and other information to laptops that are also used
to update middleware and firmware.
Significant differences exist between port and ship OT encompassing sen-
sors and equipment used in the engineering department. Engineering systems
cover many functions with unique modalities and characteristics associated
with engine, propulsion, steering, electrical, heating, ventilation and air con-
ditioning (HVAC); distillers, condensers, evaporators and associated pumps,
valves and piping. Control and health monitoring of engineering systems dis-
tributed throughout the vessel is accomplished through the use of networks
of actuators and sensors to monitor and generate alarms resulting from out-
of-tolerance and abnormal conditions. Sensor networks are integral to the
monitoring of engines and drive trains, bearings and other components that
consider temperature, wear and vibration, water infiltration into oil, as well
as monitoring exhaust gas and cylinder liners. New sensor technologies also
exist for level sensing, water ingress, pressure monitoring, deck winch torque
and load monitoring, battery management systems and many other engin-
eering sensing applications.
Deck department systems forming part of the physical infrastructure
include the devices, monitors and peripherals associated with ECDIS, depth
and navigation sonars, GPS/ GNSS, AIS and satellite AIS, digital radar/
Automatic Radar Plotting Aid (ARPA), broadband satellite communications,

high frequency (HF) and VHF communications, digital select calling (DSC),
weather sensors, navigation and security visible and infrared cameras, navi-
gational text messaging (NAVTEX) and autopilots.
The physical infrastructure has greatly expanded as passenger ships
integrate into voyager cabins fire and smoke detectors supplemented with
floor, wall and shower sensors to detect occupancy and movement. Facial
recognition devices are used to open cabin doors, and wearable devices
containing passenger-and crew-unique identity information are read by
thousands of sensors both on board the ship and in port to enable hospi-
tality and victualling services as well as speed the boarding and debark-
ation processes.
Also forming part of the physical infrastructure are shipping containers
outfitted with sensors to help track movements when repositioned on ships,
within ports and during transit from their place of origin to the destination.
Sensors also help detect and predict conditions at sea conducive to container
loss overboard and monitor the environment and detect excessive heat. Heat
and smoke detectors are widely employed on many Roll-On/Roll-Off (Ro-Ro)
vessels and ferries to detect and isolate smoldering and combustion among
vehicles being carried. Tanker ships and gas carriers utilize special sensors to
monitor levels, pressures, gas buildup and to detect foreign materials within
their cargo. Bulk carriers feature sensors to detect and possibly help prevent
cargo liquefaction by monitoring temperature and humidity levels both in
their holds as well as within the cargo contents.
Underlying many onboard networks are digital communication buses for
data transfer between sensors and onboard computers. Internal ship commu-
nication is accomplished using one or more data buses that include local area
network (LAN), RS232/RS485, NMEA-0183/IEC 11623, NMEA-2000/IEC
61162-3, controller area network (CAN) and Bluetooth; with communica-
tions beyond the confines of the ship accomplished through 5G terrestrial
networks, wide area network (WAN) and satellites. Limited data transfer
is also performed using VHF radio for AIS data between ships and vessel
traffic control centers. It should also be noted that digital communications
protocols are also now being used on a limited basis for data, telemetry and
image exchange using HF radio primarily as a backup for long-distance sat-
ellite communications.
9.5.1.2 Ship Digital Infrastructure

Digital ship infrastructure is comprised of all of the software, middleware,
firmware, and related mechanisms reachable by the outside world through
which access to ship network systems may be achieved. The digital elem-
ents of ship IT infrastructure can be summarized as mirroring those of port
infrastructure counterparts as tailored to the specific requirements of the
ship and its mission(s). This is discussed in the paragraph for Port Digital
Infrastructure.
However, ship OT digital infrastructure differs greatly from those of ports
in terms of functionality and data content unique to support the needs of
physical components serving engineering, deck, passenger, cargo and other
relevant operations. This varies widely depending on the type of ship involved
and its function.
9.5.2 Port Infrastructure

Ports of origin as well as destination ports serve as primary focal points
where the vast majority of physical and electronic interactions with ships
occur. Port activities are represented in the physical and digital infrastruc-
ture necessary to accomplish all functional aspects of ships’ voyages. This
includes the loading and offloading of cargos, passengers and crews and their
corresponding records and documentation; transmission of bills of lading,
logistics, customs and compliance data and the gathering of operational
and maintenance data from preceding voyage(s) and legs of a voyage. Also
included is the repair and installation of physical equipment at the port itself,
shipboard deck and engineering equipment as well as computers, sensors and
operational equipment, software, middleware, and firmware; provisioning
and victualing supplies. Our approach examines port physical and digital
infrastructure as separate and unique from ship infrastructure yet comprising
a hybrid ship/port infrastructure through which the vast majority of interac-
tions with the ship take place. This comprises a complex and varying infra-
structure that also makes the ship highly vulnerable while in port.
9.5.2.1 Port Physical Infrastructure

The physical port infrastructure is comprised of all of the devices and inter-
faces visible to and accessible by the outside world through which access to
port and ship network systems may be achieved. This includes a broad range
of port equipment and devices related to IT and OT regardless of whether it is
physically located at the port itself or remote from the port yet connected via
electronic means. The breadth of the physical infrastructure spans multiple
levels from the largest of cranes and gantries used to load and unload ships’
cargos; to the buildings that house the servers, computers, personal comput-
ers (PCs), firewalls modems, routers, switches, memory devices, clients, IoT
endpoints, computer and other electronic IT ports, cables and wireless con-
nection paths that process data and information related to port operations;
as well as the entry points, security devices, cameras and sensors associated
with these buildings and surroundings. Also included are Internet web portals
that allow shippers and consignees to track cargo, check pickup availability,
review ship schedules, file bills of lading and many more functions. Physical
infrastructure also extends to the cargo containers outfitted with smart sen-
sors monitoring their contents and movement of the containers themselves,
and to smart sensors placed within and monitoring the conditions of bulk
cargos, including programmable logic controllers (PLCs) and arrays (PLAs)
integral to remote control and smart sensor designs. Also considered are the
systems used to process the cargo, passengers and crews, maintenance, data
acquisition, processing and analytics across various commercial organiza-
tions and governmental agencies performing both direct and supporting roles
within the port and onboard ships.
Their configurations may be characterized in terms of physical assets, the
functions they perform, how they are interconnected, the means by which
they are accessed and other pertinent details. This includes noting specific
details by equipment types and models, capabilities, accessible ports and the
means used for access and update, including human access using physical
(laptop, mobile devices, USB stick) and electronic (LAN, WAN, Bluetooth,
etc.) methods.
9.5.2.2 Port Digital Infrastructure

Digital port infrastructure is comprised of all of the software and related
mechanisms reachable by the outside world through which access to port
and ship network systems may be achieved. Digital elements include oper-
ating systems, application software, middleware, firmware, utilities, device
drivers, databases, enterprise software, application suites and communica-
tions software. Additional elements include IoT endpoint software, mobile
and web applications, and websites such as iDockworks, Navis N4, TOS
Web Portal and VoyagerTrack;34 remote desktop protocol (RDP) endpoints,
email inboxes, virtual private networks (VPNs), cloud services as well as
supply chain infrastructure and services. Represented within these elements
are a wide variety of data and data types, communication protocols, imple-
mentation software and the communications capabilities needed to perform
routine and extraordinary company operations in terms of services and the
transportation of passengers, goods and materials on a day-to-day basis.
Also included are capabilities for handling security, regulatory, safety, main-
tenance, logistics, passengers and crews, voyage planning, voyage and cargo
tracking, victualing and other human needs provisioning under nominal and
abnormal conditions and circumstances. Significant attention must be given
to the use of cloud services that produce a very large digital infrastructure
and are managed very differently than on-premise storage, change dynamic-
ally and can easily compromise security. Also needed to be considered is the
effect on port digital infrastructure resulting from digital twin technology
used to duplicate all port software functions in a virtual representation of
operational systems.
9.6 Present-D ay Precautions Against Cyberattacks

Many excellent resources are available to assist in establishing cyber security
onboard ships, in ports, and in IT and OT infrastructure in general. A good
starting point is the International Maritime Organization (IMO) itself, which
specifically addresses and provides guidance for maritime cyber risk and risk
management. Additional guidance includes publications and other resources
available from the Baltic and International Maritime Council (BIMCO),
American Bureau of Shipping (ABS), Digital Container Shipping Association
(DCSA), International Association of Classification Societies (IACS),
International Association of Ports and Harbors (IAPH) and the United States
Coast Guard. Much of the guidance and many of the recommendations
that have been developed for maritime use are guided by the U.S. National
Institute of Standards and Technology (NIST) National Cybersecurity Center
of Excellence in the development of their NIST Cybersecurity Framework. The
contributions of each of these organizations are summarized in the following
paragraphs.
International Maritime Organization (IMO) –The IMO has issued guidance

and adopted a resolution to address maritime cyber risk, which may result in
shipping-related operational, safety, or security failures as a consequence of
information or systems being corrupted, lost, or compromised. Guidelines on
Maritime Cyber Risk Management (MSC-FAL.1/Circ.3) provide high-level
recommendations to help safeguard shipping from cyber threats and vulner-
abilities that thwart effective cyber risk management.35
The IMO Maritime Safety Committee in June 2017 also adopted Resolution
MSC.428(98), Maritime Cyber Risk Management in Safety Management
Systems, to encourage national administrations to ensure cyber risks are
appropriately addressed in existing safety management systems no later than
the first annual verification of the company’s Document of Compliance after
1 January 2021.36
Baltic and International Maritime Council (BIMCO) –In 2018, BIMCO,

in cooperation with other maritime industry organizations, released their
Guidelines on Cyber Security onboard Ships based on high-level principles.37
The update to this document issued in 2020 with its 4th edition seeks to
establish safety, security and commercial risk awareness that results from a
lack of cyber security measures. Topics considered include the protection of
shipboard IT infrastructure and connected equipment, a system for authen-
tication and authorization of users, protection of data used in the ship envir-
onment, management of IT users, management of communication between
the ship and the shoreside, and development and implementation a cyber
incident response plan based on a risk assessment.
American Bureau of Shipping (ABS) –ABS in 2016 released The Application

of Cybersafety Principles to Marine and Offshore Operations, which provides
best practices for cybersecurity across the marine and offshore communities to
provide insights for operations, maintenance and support of cyber-enabled sys-
tems.38 They have been updated to align with the ABS Guide for Cybersecurity
Implementation for the Marine and Offshore Operations, which provides
safety and security requirements and recommendations on the assessment
of cybersecurity systems.39 It also provides guidance for vessel readiness for
preventing and managing cyber events that may compromise the safety and
security of the data, systems, and vessels of a company or organization.
Digital Container Shipping Association (DCSA) –DCSA has published the

DCSA Implementation Guide for Cyber Security on Vessels for the global
container shipping industry.40 The guide outlines best practices to assist
company cyber security personnel who are responsible for fleet-wide cyber
security onboard vessels. It provides guidance on cyber-related aspects of IT
such as computers, electronic manuals and networks and applications OT
that includes engine control, ECDIS, onboard measurement and control sys-
tems, PLCs and remote support for engines.
International Association of Classification Societies (IACS) –IACS has

released its Recommendation on Cyber Resilience to help achieve flexibility
in cyber security practices.41 Areas of focus include the design, integration
and/or maintenance of computer-based systems to support secure operations.
Also discussed is protection against unauthorized access, misuse, modifica-
tion, destruction, or improper disclosure of information created, archived, or
used in onboard computer-based systems or transported through networks
connecting such systems.
International Association of Ports and Harbors (IAPH) –In 2021, IAPH pub-
lished its Cybersecurity Guidelines for Ports and Port Facilities to provide the
international port industry with a set of cybersecurity guidelines based on
successes achieved by ports and port facilities from around the world.42 They
are designed to assist executives in the port industry to foster greater collab-
oration within their organizations, as well as more broadly with their local,
regional, national and international partners and stakeholders.
United States Coast Guard –The U.S. Coast Guard’s (CGCYBER) Maritime
Cyber Readiness Branch (MCRB) supports the cybersecurity mission in
the commercial maritime transportation community.43 They provide many
resources that can assist in developing, maintaining and updating strategic
planning and cyber policy; assessing threats, vulnerabilities and impact of
loss to the Maritime Transportation System (MTS), sustaining strong part-
nerships with key MTS stakeholders to develop insight into future MTS
threat vectors; and oversee cybersecurity support to MTS operations through

threat assessments, cyberinfrastructure posture improvements, info sharing
and cyber domain incident awareness.
National Institute of Standards and Technology (NIST) –The NIST released

its Framework for Improving Critical Infrastructure Cybersecurity, focusing
on using business drivers to guide cybersecurity activities and considering
cybersecurity risks as part of the organization’s risk management processes.44
Released in 2018, the Framework consists of three parts: the Framework
Core, the Implementation Tiers and the Framework Profiles. The Framework
Core is a set of cybersecurity activities, outcomes and informative references
that are common across sectors and critical infrastructure. Elements of the
Core provide detailed guidance for developing individual organizational
Profiles. Through the use of Profiles, the Framework will help an organ-
ization to align and prioritize its cybersecurity activities with its business/
mission requirements, risk tolerances and resources. The Tiers provide a
mechanism for organizations to view and understand the characteristics of
their approach to managing cybersecurity risk, which will help in prioritizing
and achieving cybersecurity objectives.
9.7 Future Directions to Enhance Cyber Security

Opportunities exist to create multiple solutions that enable rapid develop-
ment and dynamic updates of accurate and comprehensive reduced-order
models of the protect surface. This is especially important when dealing
with complex critical infrastructure to detect critical system elements that
would make them susceptible to model-based attacks. Novel artificial intelli-
gence (AI)-based and other methods are needed to autonomously build and
update these models to represent classes of network system functionalities
that span multiple, different ship and port infrastructure configurations and
defend against a multitude of diverse threat vectors. Such methods must
minimize human-guided domain knowledge and training requirements and
feature dynamic development, adaptation and reconfiguration to overcome
vulnerabilities exposed by destabilization attacks. These must also result in
an order-of-magnitude reduction in development time over existing methods,
enhanced resilience to broad-based attacks and reduced brittleness to previ-
ously unknown cyber tactics.
Strategies and solutions to cyber security issues and problems in the United
States are investigated by several agencies under the Executive Office of the
President, several of which have already been identified in this chapter. The
National Science and Technology Council (NSTC) is the principal means by
which the Executive Branch coordinates science and technology policy across
the diverse entities that make up the Federal research and development enter-
prise. Its primary objective is to ensure that science and technology policy
decisions and programs are consistent with the President’s stated goals. The
NSTC prepares research and development strategies that are coordinated
across Federal agencies aimed at accomplishing multiple national goals.
The 2019 Federal Cybersecurity Research and Development Strategic Plan
aims to coordinate and guide U.S. Federally funded R&D in cybersecurity,
including the development of consensus-based standards and best practices.45
The Plan identifies four interrelated defensive capabilities (deter, protect,
detect and respond) and six priority areas for cybersecurity R&D (artificial
intelligence, quantum information science, trustworthy distributed digital
infrastructure, privacy, secure hardware and software, and education and
workforce development) as the focusing structure for Federal cybersecurity
R&D activities and investments to benefit the Nation.
Signed into law in December 2022, the Quantum Computing Cybersecurity
Preparedness Act represents major legislation in the area of cybersecurity that
encourages the U.S. Federal government to adopt technology that is protected
from decryption by quantum computing.46 This act addresses the migration
of executive agency information technology systems to post-quantum crypt-
ography, where encryption is strong enough to resist attacks from quantum
computers developed in the future. After the National Institutes of Standards
and Technology (NIST) has issued post-quantum cryptography standards,
guidance will be issued requiring each executive agency to develop a plan to
migrate the information technology of the agency to post-quantum cryptog-
raphy. Also included will be a strategy to address the risk posed by IT vulner-
abilities to weakened encryption due to the potential and possible capability
of a quantum computer to breach such encryption and the development of
standards for post-quantum cryptography. Further information on quantum
sensing is described in greater detail in Chapter 11, Next Generation Sensing.
References
1 Vanguard. Maritime Cyber Attacks Increase by 900% in Three Years. 29 July
2020. Vanguard Media Limited, Nigeria. www.vanguardngr.com/2020/07/
maritime-cyber-attacks-increase-by-900-in-three-years/
2 MarineLink. Cyber Attacks on the Rise at US Ports and Terminals. 5 October
2022. www.marinelink.com/news/cyber-attacks-rise-us-ports-terminals-499964
3 Jones Walker. 2022 Ports and Terminals Cybersecurity Survey. Jones Walker
LLP. www.joneswalker.com
4 Alejandro Mayorkas, Secretary of the U.S. Department of Homeland Security.
Testimony before the U.S. Senate Committee Hearing Channels on Threats to
the Homeland. 17 November 2022.
5 David Kushner. The real story of stuxnet. IEEE Spectrum. 26 February 2013.
https://spectrum.ieee.org/the-real-story-of-stuxnet
6 Deborah Haynes. Iran’s Secret Cyber Files. Sky News. 27 July 2021. https://
news.sky.com/story/irans-secret-cyber-files-on-how-cargo-ships-and-petrol-
stations-could-be-attacked-12364871
7 Kuo-Hui Yeh, Wenli Shang, Tianyu Gong, Chunyu Chen, Jing Hou, and Peng
Zeng. (2019). Information security risk assessment method for ship control
system based on fuzzy sets and attack trees. Security and Communication
Networks, Article ID 3574675. https://doi.org/10.1155/2019/3574675
8 G. Kavallieratos, S. Katsikas, and V. Gkioulos. (2019). Cyber-attacks against
the autonomous ship. In: Katsikas, S., et al. Computer Security. SECPRE
CyberICPS 2018. Lecture Notes in Computer Science, vol. 11387. Springer,
Cham. https://doi.org/10.1007/978-3-030-12786-2_2
9 World Maritime News. Nightmare Scenario: Ship Critical Systems Easy Target
for Hackers, December 2017. https://worldmaritimenews.com/archives/238
869/nightmare-scenario-ship-critical-systems-easy-target-for-hackers
10 B. Svilicic, J. Kamahara, M. Rooks, and T. Yano. Maritime cyber risk man-
agement: An experimental ship assessment. Journal of Navigation, 72(5),
1108–1120.
11 Offshore Energy. Nightmare Scenario: Ship Critical Systems Easy Target for
Hackers. 21 December 2017. www.offshore-energy.biz/nightmare-scenario-
ship-critical-systems-easy-target-for-hackers/
12 Bartlomiej Hyra. M.Sc. Thesis, Master of Science in Engineering. DTU Compute
Department of Applied Mathematics and Computer Science, Technical
University of Denmark, Kongens Lyngby (2019). https://backend.orbit.dtu.dk/
ws/portalfiles/portal/174011206/190401_Analyzing_the_Attack_Surface_of_
Ships.pdf
13 University of Texas. UT Austin Researchers Successfully Spoof an $80 million
Yacht at Sea. 29 July 2013. https://news.utexas.edu/2013/07/29/ut-austin-rese
archers-successfully-spoof-an-80-million-yacht-at-sea/
14 Defense Aviation. Iran Displays Captured US RQ-170 Sentinel, 2011. www.
defenceaviation.com/iran-displays-captured-us-rq-170-sentinel/
15 Corey D. Ranslem. Secure at Sea: AIS Fraught with Vulnerabilities. 10 January
2020. www.the-triton.com/2020/01/secure-at-sea-ais-fraught-with-vulnerab
ilities/
16 BBC News. Warship Positions Faked Including UK Aircraft Carrier. 2 August
2021. www.bbc.com/news/ technology-58027363
17 Eduard Kovacs. Ship Data Recorders Vulnerable to Hacker Attacks. Security
Week. 11 December 2015. www.securityweek.com/ship-data-recorders-vulnera
ble-hacker-attacks/
18 Sean Gallagher. Hacked at Sea: Researchers Find Ships’ Data Recorders
Vulnerable to Attack. ARS Technica, 10 December 2015. https://arstechnica.
com/information-technology/2015/12/hacked-at-sea-researchers-find-ships-
data-recorders-vulnerable-to-attack/
19 Ruben Santamarta. Maritime Security: Hacking into a Voyage Data Recorder
(VDR). IOActive. 9 December 2015. https://ioactive.com/maritime-security-
hacking-into-a-voyage-data-recorder-vdr/
20 Danny Palmer. Critical IoT Security Camera Vulnerability Allows Attackers to
Remotely Watch Live Video –and Gain Access to Networks. 17 August 2021.
www.zdnet.com/article/critical-iot-security-camera-vulnerability-allows-attack
ers-to-remotely-watch-live-video-and-gain-access-to-networks/
21 Abhishek Bhanawat . What Are Intrinsically Safe Cameras Used on Ships?

Marine Insight. 18 March 2019. www.marineinsight.com/marine-safety/what-
are-intrinsically-safe-cameras-used-on-ships/
22 Thorsten Stremlau. The Vulnerability of Satellite Communications. Security
Magazine. 19 April 2021. www.securitymagazine.com/articles/94689-the-vul-
nerability-of-satellite-communications
23 Jennifer Hlad. To Protect Satellites, Secure Your Networks, Chief of Space Ops
Says. Defense One. 31 January 2023. www.defenseone.com/threats/2023/01/
protect-satellites-secure-your-networks-chief-space-ops-says/382424/
24 Jessie Hamill- Stewart and Andrew Sallay. VSAT Connectivity Comes with
Cybersecurity Threats to Vessels. Maritime Executive. 23 May 2023. www.
maritime-executive.com/editorials/vsat-connectivity-comes-with-cybersecurity-
threats-to-vessels
25 John Harries. A Fundamental Danger of NMEA 2000 Networks. 26 May 2013
and updated 11 July 2021. www.morganscloud.com/2013/05/26/nmea-2000
missing-the-obvious/
26 Trend Micro, Incorporated. Cyberattacks against Intelligent Transportation
Systems: Assessing Future Threats to ITS, 2017, pp. 27–28. https://documents.
trendmicro.com/assets/white_papers/wp-cyberattacks-against-intelligent-tra
nsportation-systems.pdf
27 Zeinab El-Rewini, Karthikeyan Sadatsharan, Niroop Sugunaraj, Daisy Flora
Selvaraj, Siby Jose Plathottam, and Prakash Ranganathan. 22 June 2020. IEEE
Sensors Journal, 20(22), ISSN 1530-437X.
28 Vincenzo Ciancaglini, Joey Costoya, Philippe Lin, and Roel Reyes. Exploitable
Flaws Found in Facial Recognition Devices. Trend Micro. 22 September 2020.
www.trendmicro.com/en_us/research/20/i/exploitable-flaws-found-in-facial-
recognition-devices.html
29 Safetytech Accelerator. Startup Deploys Wireless Sensors across Ship’s Cargo
Hold to Predict Fire. 3 March 2020. https://safetytechaccelerator.org/case-stud
ies/safetytech-startup-deploys-wireless-sensors-across-ships-cargo-hold-to-pred
ict-fire/
30 Jasper Helder and Lars-Erik Hjelm. Focus on shipping in fight against drug
smuggling. Lloyds List, 1 May 2020. https://lloydslist.maritimeintelligence.info
rma.com/LL1132202/Focus-on-shipping-in-fight-against-drug-smuggling
31 Katie Terrell Hanna. Attack Surface Definition. WhatIs.com. September 2021.
www.techtarget.com/whatis/defi nition/attack-surface
32 U.S. Department of Defense. Department of Defense Releases Zero Trust
Strategy and Roadmap, 22 November 2022. www.defense.gov/News/Releases/
Release/Article/3225919/department-of-defense-releases-zero-trust-strategy-
and-roadmap/
33 John Kindervag. Define a Protect Surface to Massively Reduce Your Attack
Surface. Palo Alto Networks, 4 September 2018. www.paloaltonetworks.com/
blog/2018/09/defi ne-protect-surface-massively-reduce-attack-surface/
34 Some of this information was obtained from: www.portsamerica.com/locati
ons/portofbaltimore-maryland
35 IMO MSC-FAL.1. Guidelines on Maritime Cyber Risk. MSC-FAL.1/Circ.3/

Rev.1, 14 June 2021. wwwcdn.imo.org/localresources/en/OurWork/Facilitat
ion/Facilitation/MSC-FAL.1-Circ.3-Rev.1.pdf
36 IMO MSC.428(98). Resolution MSC.428(98). Maritime Cyber Risk
Management in Safety Management Systems, 16 June 2017. wwwcdn.imo.org/
localresources/en/KnowledgeCentre/IndexofIMOResolutions/MSCResoluti
ons/MSC.428(98).pdf
37 BIMCO, et al. The Guidelines on Cyber Security Onboard Ships. Baltic and
International Maritime Council. Version 4, 2020. www.bimco.org/about-us-
and-our-members/publications/the-guidelines-on-cyber-security-onboard-ships
38 ABS Cybersafety Vol. 1. Guidance Notes on The Application of Cybersafety
Principles to Marine and Offshore Operations. September 2016. Vol. 1.
American Bureau of Shipping. Spring, TX, USA. ww2.eagle.org/content/dam/
eagle/ r ules- a nd- u ides/ c urrent/ o ther/ 2 50_ c ybersafetyV1/ C yberSafety_ V 1_
Cybersecurity_GN_e.pdf
39 ABS Cybersafety Vol. 2. ABS Guide for Cybersecurity Implementation for the
Marine and Offshore Operations. February 2021. Vol. 2. American Bureau
of Shipping. Spring, TX, USA. ww2.eagle.org/content/dam/eagle/rules-and-
guides/ c urrent/o ther/ 2 51_ cybersafety_ 2 021/ c ybersafety- v2- c ybersecurity-
guide-feb21.pdf
40 DCSA. DCSA Implementation Guide for Cyber Security on Vessels v1.0. Digital
Container Shipping Association. 3 October 2020. https://dcsa.org/wp-content/
uploads/2020/03/DCSA-Implementation-Guideline-for-BIMCO-Compliant-
Cyber-Security-on-Vessels-v1.0.pdf
41 IACS. Recommendation on Cyber Resilience. International Association of
Classification Societies. No. 166. IACS Rec. 2020/Corr.2. 2022. www.iacs.org.
uk/news/iacs-launches-single-standalone-recommendation-on-cyber-resilience/
42 IAPH Cybersecurity Guidelines. IAPH Cybersecurity Guidelines for Ports and
Port Facilities. International Association of Ports and Harbors, Tokyo, Japan.
Version 1.0. 2 July 2021. https://sustainableworldports.org/wp-content/uplo
ads/IAPH-Cybersecurity-Guidelines-version-1_0.pdf
43 USCG MCRB. U.S. Coast Guard Maritime Cyber Readiness Branch (MCRB),
Alexandria, VA. www.dco.uscg.mil/Our-Organization/CGCYBER/Maritime-
Cyber-Readiness-Branch/
44 NIST. Framework for Improving Critical Infrastructure Cybersecurity. Version
1.1 National Institute of Standards and Technology, 16 April 2018. https://nvlp
ubs.nist.gov/nistpubs/CSWP/NIST.CSWP.04162018.pdf
45 NSTC. Federal Cybersecurity Research and Development Strategic Plan.
National Science and Technology Council. December 2019. www.nitrd.gov/
pubs/Federal-Cybersecurity-RD-Strategic-Plan-2019.pdf
46 Quantum Computing Cybersecurity Preparedness Act. Public Law117-260.
www.congress.gov/bill/117th-congress/house-bill/7535
Chapter 10
Vessel Sensor Degradation
The most reliable means for detecting sensor degradation has traditionally
been well-trained mariners with years of experience who can readily dis-
tinguish between normal operation and anomalies in sensor performance
throughout a wide range of working scenarios. Unfortunately, there is a
shortage of well-trained mariners that is not expected to be resolved soon
making this problem more significant.1 As ship automation continues to
increase and sensor systems gain added complexity, the detection of sen-
sor degradation becomes even more difficult. Further complicating matters is
the introduction of unmanned and autonomous vehicles and their need for
enhanced and extended sensor suites and sensor data fusion technology that
allows them to operate without seafarers onboard. Still, the problem of sen-
sor degradation detection remains unresolved.
Unmanned and autonomous vehicles function as passenger, cargo and
other types of vessels included in the partial list cited in the first chapter of
this book. These same technologies continue to be integrated on the bridges
of conventionally staffed vessels to assist seafarers in the performance of
their duties. The volume of imagery, signal and digital data generated by
sensors in recent years has increased by orders of magnitude to advance the
development of groundbreaking hardware and software leading to new and
improved processes that make better use of these resources. The promises of
innovations in system design and function have only begun to be tapped and
include providing new perspectives from which enhanced situational aware-
ness and safety of navigation may be achieved.
Much research has been performed, and entire industries established to
account for and deal with the failures of sensors, the various failure modes to
which they are susceptible and the many methods and techniques that can be
applied to prevent, detect and even predict failure. However, little has been
said and even less investigated regarding the insidious degradation of sensors
especially over extended periods of time that can cause misleading and decep-
tive indications and shield anomalies. Such occurrences can stymie reliable
and dependable ship operations and reduce overall safety.
Doi: 10.1201/9781003377900-10
Vessel Sensor Degradation 169
Degradation implies one or more aspects of sensor data corruption that

can reduce the quality of the service they provide even though the sensor itself
may be fully operational and pass all functional test parameters. Obvious
examples of sensor degradation include:
• Mud or other contaminant on the lens of a camera that partially blocks

its view.
• A close-by obstruction in the partial field of view of a Radar antenna that
causes a dead zone without returns on a Radar display.
• Excessive barnacle and other growth on a Sonar transducer that reduces
effective range.
• Interference from other systems causing strange behavior in a display.
• Corrosion in a transmission line resulting in excessive impedance that can
produce unexpected results.
Sensor degradation caused by human agencies with nefarious intent

includes physical and cyberattacks that are also widespread and problem-
atic. Many such events can often be diagnosed by a mariner who has gained
a high level of expertise using such systems. However, watchstanders with
learned knowledge of sensor theory but little practical experience in sensor
operation will not necessarily be capable of reliably discerning degradation.
More importantly, unmanned vessels and autonomous vehicles have even
less ability to do so.
Less obvious examples of sensor degradation include the gradual loss of
pixels in a camera over time that steadily reduces resolution and accuracy,
the effects of precipitation and fog on light propagation and infrared night
vision performance, and inadequately documented performance anomalies in
new types of sensors recently introduced onto the bridge. Similar effects can
be identified in engineering applications such as optical devices that become
contaminated and grow cloudy, proximity sensors that shift from optimal
alignment and resistive transducers that age prematurely and exceed specifi-
cations due to environmental and other factors.
This chapter highlights much of the original research performed by the
author and others aimed at detecting and overcoming sensor degradation.
Applications for this technology span many industries, much of which has
focused on the automotive and trucking industries. However, comprehensive
cases are made for sensor degradation detection technology use in the mari-
time sector to enable unmanned surface and undersea vehicles to operate
much in the same fashion as a conventionally staffed vessel, and to assist the
navigation and engineering watches on conventional ships in the perform-
ance of their duties. What follows are illustrations of the fruits of these tech-
nologies when applied to solve real-life problems and application scenarios
aboard ships and maritime vehicles.
170 Vessel Sensor Degradation
10.1 Background and Investigative Approach

The ability to detect and recognize both rapid and subtle degradation in sen-
sor performance over time and, if possible, compensate for their effects is
essential to continue and maintain the nominal performance of ship opera-
tions and vehicle systems in a safe manner. This represents a highly significant
limitation that stymies commercial and naval vessels as well as unmanned
maritime vehicle operations. Also significant is the means to communicate
sensor status and degradation characteristics to one or more operations cent-
ers, vessel traffic services and Remote Control Centers (RCCs); and other
ships and maritime vehicles via broadband and low-bandwidth data links to
collaborate and to provide periodic system health status. Such capabilities
enhance ship and vehicle resilience to achieve voyage objectives and provide
a means to continuously learn new tactics to defeat or minimize such threats.
Efforts by several researchers have resulted in various methods to detect sen-
sor degradation as well as internal sensor failures. They have also resulted
in new ways to make systematic adjustments necessary to mitigate and com-
pensate for reduced and degraded, corrupted and perplexing sensor inputs to
increase the probability of a successful voyage.
One approach involves the performance of sensor signal analytics on a
variety of navigation instruments through sensor data correlation from
three different perspectives.2 These include the distribution of signal char-
acteristics across the frequency spectrum at any point in time (Frequency
Domain), changes that take place in the signal over time (Time Domain) and,
where applicable, processed sensor signals as represented in imagery (spatial
domain) created from Frequency and Time Domain signal data.3 Machine
learning and deep-learning artificial intelligence (AI) are used to detect char-
acteristic signatures and trends in sensor data indicative of rapid as well as
gradual sensor degradation from various natural and manmade phenomena.
Detection of degradation events consider failure modes and effects due to
external causes such as grounding and allision, extreme sea states and phys-
ical damage; and internal causes resulting from failure in vehicle propulsion,
steering, electrical, hydraulic, communications, navigation and health moni-
toring systems. This three-domain approach facilitates the classification of a
broad range of sensor types and signal characteristics and provides a basis
for the detection of sensor degradation, the prognosis of pending and actual
sensor and system failures, and the implementation of alternative mitigation
and recovery methods tailored to the unique attributes of each cause.
A second approach explores the effects of ultrasound noise on the per-
formance of navigation instruments utilizing a gyroscope, accelerometer, and
magnetometer.4 Considered in this approach is sensor data fusion from a
number of sensors to produce more accurate information than if data were
taken from individual sensors and used separately. The primary goal is to
enhance performance and reliability by detecting inconsistencies in sensor

measurements when under the influence of acoustic and magnetic emissions
external to the system.
A third approach focuses on the influence of electromagnetic interference
on Ethernet networks on critical sensor systems.5 Specifically considered is
their effect on Time-Triggered Ethernet (TTE) in terms of precise signal tim-
ing, redundant pathways and careful switching to ensure that information
arrives precisely when needed. TTE is already in use in flight-critical aircraft
systems, aboard spacecraft and power-generating wind turbines and its use is
becoming more prevalent on ships and maritime vehicles.
This chapter discusses a group of sensors most applicable and common
to ships and maritime vehicles selected and weighted by priority for voyage
requirements as well as safety of navigation. These sensors include Sonar,
Radar, visual and infrared cameras, Lidar, and other systems through which
vehicle situational awareness related to external factors, events and threats
is maintained. Also considered were internal vessel sensors, such as exist for
a wide variety of engineering system sensor applications, as well as inertial
navigation (INS) and other systems that require periodic updates of position,
navigation and timing (PNT) inputs to maintain accuracy by minimizing
their susceptibility to increasing progressive error over time.
Decision-making capabilities were also developed using machine learning
and deep-learning AI techniques to determine potential sources (sensors and
sensor inputs), types (rapid or gradual occurrence) and symptoms (noise,
distortion, etc.) of sensor degradation. Considered is whether sensor degrad-
ation is very likely to occur soon or is currently in evidence and the various
alternatives to mitigate and compensate for the effects of degradation. These
alternatives include solutions targeted to individual sensor systems as well
as approaches that consider the fusion of alternative and secondary sen-
sor data sources to derive an acceptable solution. Theoretical assumptions
regarding both the causes of sensor degradation and their effects were vali-
dated through experimentation using a research vessel at sea in an oper-
ational environment for evaluating new technologies for civilian ships and
unmanned, remotely operated and autonomous surface vehicles. Inherent to
this solution is the ability to perform operations in communications-denied
environments without assistance from remote controllers or observers or
with limited communications where small data packets containing relevant
sensor information can be sent to provide status and diagnostic information
to a remote location. The approach followed examines a constant stream of
external and internal vehicle sensor data, continually assesses data charac-
teristics for evidence of potential and actual degradation and, upon detec-
tion, potentially recommends and implements alternative courses of action
to maintain readiness, preserve the safety of navigation, and achieve voyage
objectives.
10.2 Sensor Systems Examined

A comprehensive suite of navigation and engineering sensors and sensor
types with a variety of inputs, outputs and signal characteristics were tar-
geted in this research. External causes of sensor degradation originating from
outside the ship or vehicle were considered, along with causes originating
from within the vessel due to internal system performance outside of nominal
limits (meaning degraded) and failure that may cause direct and unintended
indirect consequences. This includes both onboard and offboard sensors and
related systems that provide situational awareness of relevant phenomena
from the perspectives of space, at the water’s surface and below the waterline.
More than 24 systems were considered, including Sonar, Radar, Lidar, visible
light and infrared imaging, GNSS, meteorological, oceanographic and other
types of sensors. Also considered were internal vehicle system sensors associ-
ated with propulsion, steering, electrical, communications, and navigation,
where both rapid and gradual sensor degradation can occur. This includes
INS, where the absence of external references for PNT data results in pro-
gressively increasing errors in velocity that are compounded into even larger
errors in position resulting from integration drift. Prognostics of imminent
failure and diagnosis of degraded performance in vehicle engineering systems
can also be detected using noncontact sensors. Examples include electronic
signatures of individual motors, pumps and other individual equipment indi-
cative of worn relays, burned windings, corrosion, broken insulation and an
increasing lack of efficiency for specific system components.6
A subset of these was selected for experimentation, including five types of
Sonar (echosounder, multibeam, side-scan, chirp and navigation) and two
types of imaging (visible light and infrared) sensors; with additional consid-
eration being given to Radar and Lidar. Internal vehicle sensors include INS
and sensors associated with alternators, tanks and other onboard systems.
These selections were based in part on applicability across a broad range of
ship and vehicle types, diversity of data and signal types, and access to inter-
mediate data, signal elements and endpoint imagery representative of com-
mon sensor network inputs. Analysis of signal characteristics is performed
as they appear in two or three domains, as applicable, during nominal oper-
ation and while experiencing both rapid and gradual performance degrad-
ation over longer periods of time. Causalities of sensor degradation were
considered including naturally occurring environmental conditions, interfer-
ence from internal and external sources, and damage from hazards to navi-
gation, adversarial action and internal sensor failure modes.
10.3 Sources of Sensor Degradation

External causes of sensor degradation considered included marine growth
on hull exteriors and sensor transducers; interference, electrical and physical
damage to sensors caused by grounding and allision, as well as the effects of
nefarious acts by adversaries. Internal sensor system failure modes were con-
sidered where their effects can be ambiguous and hard to distinguish yet are
vital to the development of an effective mitigation and compensation strategy.
Sensor degradation is considered on the basis of the symptoms of its occur-
rence and their effects as may be attributed to changes in signal characteristics
from both external causes and causes internal to the vehicle itself. This is
accomplished through system failure and is based upon findings in the rele-
vant literature as well as experimental results related to failure prognostics,
fault diagnosis and sensor analytics. However, practical aspects of software
design and development, training data set creation and the development of
cases for verification, validation, and testing require the acquisition and use of
very large volumes of data and imagery. These datasets must be obtained in a
relevant operational environment to train machine learning and deep-learning
AI solutions to adequately perform these tasks. Different types of experiments
were performed that provided direct evidence of sensor degradation over
time. Training data sets were created consisting of millions of instances of
discrete signal measurements, characteristics, imagery and metadata for nom-
inal and aberrant sensor performance at increasing degrees of hull fouling,
interference, damage and internal failures. Experiment goals and methods are
specific to each undertaking and described in the paragraphs that follow.
10.3.1 Biofouling
Gradual sensor degradation over time by decreasing sensor efficiency is caused
by biofouling, which is a result of the growth of marine plant and animal life
on underwater transducers. This growth can attenuate Sonar and other sen-
sor signals transmitted from active sensors as well as reflected and passive
signals received by onboard sensors. In addition, hull roughness attributed
to biofouling greatly reduces hull efficiency resulting in reduced speed given
the same engine thrust, greater wear and tear on propulsion and steering
systems, and differences in vehicle handling and control surface response that
can alter variables used to predict INS progressive error. This is particularly
critical for unmanned vehicles expected to be deployed at sea for months at a
time without manned intervention. During such long deployments significant
fouling would be expected to occur. It is necessary to ascertain the effects of
fouling on sensor signals through direct measurement in the Frequency and
Time Domains at the sensor and transducer interface and, where applicable,
determine the symptoms of fouling on Spatial Domain representations of
processed transducer signals.
10.3.2 Interference
Numerous different experiments took place over a span of many months to
simulate and effect actual rapid and gradual sensor degradation in a rele-
vant operational environment due to interference from natural and manmade
sources both innate to the natural environment and resulting from adver-
sarial action. This includes spoofing and denial of service (DoS) attacks on
GPS/GNSS systems themselves, and sensor systems that depend on continu-
ously updated PNT data. Also included in these experiments was the use of
multiple extraneous Radar, Sonar and Lidar signals directed toward normally
operating systems with the intention of jamming, confusing and degrading
these systems outside of normal operating specifications.
Degradation of visual and infrared sensors was also accomplished using
actual environmental conditions known to facilitate degraded performance
that includes various precipitation intensities (light, moderate and heavy)
and types (rain, snow, sleet, hail, etc.). Experiments were performed under
varying visibility from clear and unlimited conditions to zero visibility due
to fog and other causes. The effects of icing, turbulence, condensation, and
other factors, such as a buildup of salt deposits from sea spray that degrade
sensor performance were also considered through natural occurrence and/or
simulation of these conditions. A simulation of GPS/GNSS interference with
INS was performed to eliminate any potential for interference with the navi-
gation of nearby aircraft and vessels.
10.3.3 Damage
Many ways exist to damage sensors in a manner that falls short of complete
failure and results in degraded capability. For external sensors, these include
misalignment and partial blockage of view from being hit with debris as
well as from large wave strikes, groundings and allision that causes phys-
ical damage to sensors and transducers. Also considered is the buildup of
contaminants from the combustion of nearby vessel components, intense
heat from fire or internal overheating caused by lack of ventilation, extreme
cold from severe operational conditions, glancing strikes by laser weapons,
and even the application of chemicals designed to render sensors blind or
otherwise inoperable. For internal sensors, performance degradation can ori-
ginate from failures within the various propulsion, electrical, hydraulic, com-
munication, navigation, and other systems within the vehicle. A subset of
vehicle temperature, pressure, voltage, current, power and other sensors and
measurement characteristics commonly associated with these systems were
selected for demonstration.
10.3.4 Degradation Modes

Experiments associated with sensor degradation differ from the sensor dam-
age experiments described in the previous paragraph as the causes of both
rapid and gradual degradation can be due to anomalies integral to the tech-
nologies used and the design of the system components themselves, and
not necessarily caused by external factors. Examples of visual and infrared
cameras include the presence of many horizontal lines across an image as

well as splashes of color that represent a bad sensor element rather than
external damage. The causes of such failures can be difficult to discern and
range from electrostatic discharge to laser strikes, excessive heat, humidity,
loose electromechanical connections and long-term exposure to sunlight.
Sonar degradation modes that can cause data corruption include loss of
transducer watertight integrity, electrical shorts, electrical arcing on the
transducer, low insulation resistance and thermal runaway. Additional
failure modes exist with respect to the connectors themselves. These anom-
alies can manifest as noise within sensor signal measurements, spots and pix-
ilation across images, lack of dynamic range, partial or full image freezing,
and many more symptoms peculiar to the types of sensors used.
These and other degradation modes have been identified for each of the sen-
sor types considered. Failure modes were simulated or physically introduced
into a number of these systems to measure sensor performance and signal
characteristics across the two or three domains, as applicable. Measurement
datasets representative of sensor degradation modes were also created for
use in developing and training machine learning and deep-learning AI sensor
capabilities.
10.3.5 Symptoms and Characteristics

Signatures of sensor degradation detected within electrical signal character-
istics across the frequency spectrum can include reduced or excessive signal
strength in transmitted and received signals; the presence of ringing, inappro-
priate signals or undesirable harmonics, the absence of required signals and
incorrect signal characteristics. Change in signal characteristics over time
can provide a direct correlation to causative events and insight into rates
of sensor degradation, provide context regarding the nature and severity of
the anomaly, and possible alternative courses of corrective action. For more
complex sensors such as Sonar, Radar, Lidar, and visual/infrared devices
where signal processing of transducer data results in the creation of a graphic
representation, degradation can be manifest as noise, distortion, pixelation,
incompleteness, loss of resolution, and other as Spatial (or pixel) Domain
anomalies within the imagery. For INS, we examine the fusion of sensor
inputs used to update PNT and consider the potential effects of GPS/GNSS
signal degradation and unavailability, spoofing, and DoS attacks.
10.4 Sensor Degradation Detection

Methods used and processes followed are described for the development of
capabilities to detect, identify and mitigate sensor degradation resulting from
various causalities. This includes, were possible, prognosis and diagnosis of
sensor system failures to distinguish them from external causes of sensor
degradation. Details of the characteristics of system components are pro-

vided relevant to sensor degradation detection, the development processes
used and the methods and procedures followed in evaluating the results.
10.4.1 Design and Implementation

The design of system hardware components, software architecture, machine
learning, and deep-learning AI to implement the prognostic and diagnostic
elements of sensor degradation was accomplished based on the specific
requirements associated with the selected sensor types. Also included were
capabilities for mitigation and compensation for sensor degradation.
System hardware is comprised of commercial-of-the-shelf (COTS) sensor
components providing digital, analog to digital conversion; NMEA, Ethernet
and other data bus capabilities, and a COTS computer. A software-defined
radio (SDR) performs oscilloscope and spectrum analyzer functions needed
to accomplish sensor data analytics for Frequency and Time Domain signal
measurements and characteristics. Special data conversions are also per-
formed as may be needed to support the unique requirements of vehicle sen-
sors, such as resistance, voltage, current and other measurements.
The computer controlled SDR operations and received SDR sensor data
analytics products for further analysis and correlation between the Frequency
and Time Domains. It also performed Spatial Domain analytics using imagery
products created by more complex sensors such as Sonar, Radar and Lidar.
Embedded within the computer are sensor data reasoning capabilities in the
form of machine learning and deep-learning AI solutions responsible for ana-
lyzing three domain data. This determined characteristics needed to deter-
mine whether sensor degradation exists now or may exist in the immediate
future, and identified specific details regarding sensor degradation. In add-
ition to external sensors, the computer also acts as the primary interface to
the vehicle’s internal propulsion, steering, electrical, hydraulic, communica-
tions, navigation, and health monitoring systems.
10.4.2 Rapid Prototyping

Rapid prototyping helped to quickly create a working software architec-
ture and functional implementation to implement effective reasoning cap-
abilities for one type or class of sensors and sensor signal characteristics.
Subsequently, the system’s scope was widened to encompass additional sen-
sor systems and characteristics. Each sensor type represents one iteration of
an incremental development model where design, development and testing
were accomplished for the set of new functions introduced during each cycle.
In addition, retesting of the results of previous cycles was performed to ensure
these capabilities still function properly and have not been adversely affected
by newly developed capabilities. This process continued until an engineering
model of the system was created and made available for testing over a period
of several months.
10.4.3 Verification, Validation and Test

Determining correctness throughout our development processes included
final engineering model development prior to testing. Verification involved
determining that sensor requirements were fulfilled and that they had been
correctly implemented. This process identified any missing requirements
that should be considered, unnecessary requirements, and that the design
process effectively implemented the requirements. Performance validation
was accomplished to ascertain that the system is capable of correctly imple-
menting the requirements. The validation process determined whether the
requirements were reasonable in terms of their ability to reflect the actual
functions that the product must perform, and whether the design was
adequate to achieve the desired functions. Performance testing covered the
breadth and depth of system functionality to determine whether it is robust
and capable of performing across all required conditions and operational
settings. This included exercising all system functions with a wide variety of
test cases to demonstrate its capability to perform required functions and to
ensure that, in the event of failure to perform properly, it is able to do so in
a manner that is not destructive and provides traceability as to the cause(s)
of failure.
While it is possible to achieve verification and validation of machine learn-
ing solutions using statistical processes, deep-learning AI solutions are not
amenable to this approach due to the inherent lack of visibility into the func-
tioning of their internal processing elements. Merely implementing black box
testing techniques based on system requirements and design is insufficient as
decisions are being made in real time. It may not be possible to effectively
reproduce the same outcome without replicating the same exact stimulus
recorded by sensors expressed within Frequency, Timeand Spatial Domain
representations. This is presently an area of active research being actively pur-
sued through many academic and industry initiatives, and by the U.S. DoD
under DARPA’s Assured Autonomy project.7 Results to date have focused
primarily on visualizing neural network activity by using activation maxi-
mization that attempts to determine what weights of individual node of a
neural network provide the greatest amount of activation. Another considers
determining how often specific paths through a neural network are activated
using a “heat map” approach to visualizing neural network activity. Other
efforts are also presently underway that may or may not provide useful solu-
tions to this problem. Our approach examined implementing deep-learning
AI solutions to the best of our ability and tracking the progress of others in
this area. However, attempting to achieve a definitive solution to the verifica-
tion and validation of deep-learning AI was beyond the scope of this effort.
Table 10.1 Sensor Suite Matrix
Sensor Manufacturer Model
Visual Camera GoPro/S CW Hero8/G ladiator 8,

Detective 4
Infrared Camera FLIR MD625
Radar Furuno/G armin 1953C/Fantome 24
Sonar five manufacturers various models
Lidar Ouster OS2-1 28
GPS/G NSS Furuno/G armin GP-3 7/8 616xsw
Inertial Navigation Advanced Navigation GNSS/I NS
Heading Sensor Garmin GAX-1 141720
Automatic Identification ICOM/G armin MA-5 00TR/A IS-8 00
System
Engineering Maretron DSM-5 70
Weather Maretron WSO100
Satellite SiriusXM Marine
Microphone SCW Detective 4
Electronic Navigation Chart Garmin BlueChartG3
10.4.4 Physical Research Environment

Development and testing were accomplished using many different engineering
and navigation sensors, a partial list of which is shown below in Table 10.1.
These sensors were located onboard the 10-meter research vessel outfitted
with sensors, as illustrated in Figure 10.1. Located approximately 24 feet
above the waterline, the Lidar sensor is installed atop the navigation light
array with clear visibility through 360° free of all physical obstructions. Atop
the Radar mast is a Furuno 12 kw, 72 nm (analog) Radar with a 48-inch
open array antenna and a Garmin 40 w, 48 nm digital Radar with a 24-inch
radome. Various sensors are located on a platform below the Radar mast,
including GPS and GNSS antennae, satellite receiver antenna and weather
sensor.
Beneath this platform are mounted six low-light video cameras, each with
a 60° field of view for a total 360° coverage with internal microphones. Also
mounted below this platform are six additional cameras, each also with a
60° field of view for a total 360° coverage to supplement the other cameras
in providing flexibility with capture areas and rates along with options for
image stability. Mounted forward and above the pilothouse is a very high-
resolution bullet camera aimed directly ahead of the vessel with a motor-
ized varifocal lens for optical zoom providing a range of fields of view, with
the FLIR infrared camera mounted directly alongside. The heading sensor is
located below the pilothouse floor at the approximate center of the vessel.
Four Garmin Sonar systems are installed on the vessel providing echo-
sounder, side scan and forward- looking navigation capabilities. The
Figure 10.1 Ten-M eter Vessel Autonomy Testbed.
FarSounder forward-looking Sonar transducer is affixed to an over-the-side

mount shown aft on the starboard side of the vessel behind the pilothouse.
A pair of EchoPilot forward- looking Sonar transducers are mounted
athwartships approximately one-third of the vessel’s length from the stern.
Another echosounder transducer is located on the starboard side and the
side-scan transducer is mounted on the stern of the vessel, with one add-
itional echosounder transducer (not shown) mounted on the starboard side
on the bottom near the stern. All Sonar transducers have been mounted at
locations on the hull that ensure minimal or no interference will occur with
other sensors.
Vessel sensors are routed internally via direct connection and Ethernet,
NMEA-2000 and -0183 data bus to the helm console displays and controls.
These signals are also routed to the instrument rack mounted in the forward
cabin of the vessel that contains the computer system, sensor data recorders,
sensor-specific computers, multi-display video monitor and uninterruptable
power supply.
10.5 Detection Results

Results obtained across the areas studied are provided in the appropriate
headings of this paragraph.
Please Note –Examples of nominal and degraded performance are illus-

trated using fully functional sensors that pass parametric and self-test criteria.
Yet the data and display information they provide is tainted at best, and at
worst misleading and unusable to where vessel operations and safety of navi-
gation may be jeopardized. Such knowledge can be instrumental in reconfig-
uring onboard sensor systems to compensate for the effects of degradation,
and to determine whether voyage objectives may still be accomplished.
Benchmark biofouling measurements of nominal signal characteristics
were made on a clean hull using two different types of Sonar systems: a
single-beam echosounder and a Forward-Looking Navigation Sonar (FLS).
Signal characteristics were also measured while in the presence of various
degrees of hull fouling, ranging from soft fouling with bacterial and diatom-
aceous slime and algae, calcareous macrofouling and hard fouling that may
also include barnacles, tubeworms, mussels and oysters; and composite foul-
ing including elements of both soft and hard fouling.
A testbed for this purpose was established using an echosounder and fish
finder Sonar, and a 3D Forward-Looking Navigation Sonar. The testbed
transducers and hull were continuously immersed in salt water with over
two years of fouling in place with a fouling rating (FR) of 100 over the
entire hull’s wetted surface.8 The EchoPilot navigation Sonar was fully func-
tional but partially inoperative due to the extensive degree of fouling, and
the Lowrance echosounder and fish finder was operational while exhibiting
sensor degradation in terms of loss of range and resolution. Several iterations
of cleaning of the hull took place to progressively reduce Fouling Rating from
100 to 0, with signal measurements taken at each degree of fouling. Although
not a perfect replication of natural progression of fouling, this experiment
provided us with rapid results over the course of several days that approxi-
mate naturally progressing fouling conditions. Experiment results were docu-
mented with signal characteristics and measurement values taken at each
stage of fouling denoting electrical performance across the three domains for
both testbeds, with Figure 10.7 (bottom) illustrating the effects.
The following figures provide examples of sensor degradation due to inter-
ference, damage and anomalies for the various sensors. Unless otherwise
noted, each of these sensor degradation events were effectively and consist-
ently detected.
10.5.1 Visible Light Camera

Camera images shown in Figure 10.2 depict various types of natural and
manmade degradation that include fog, heavy precipitation, physical dam-
age, chemical obscurants and physical obstructions. These are just a few of
the many images obtained from the three types of onboard visual cameras
listed in Table 10.1 under different weather conditions at different times of
day that were used in testing image degradation detection capabilities. All
Figure 10.2 Degraded Visible Spectrum Camera Images.
illustrated states of degradation were detected in addition to many more that

are not illustrated. The drone aircraft that operates off the research vessel
also provides live video and data feeds amenable to the use of sensor degrad-
ation detection technology.
10.5.2 Infrared Camera

Figure 10.3 illustrates nominal (left) and degraded (right) imagery obtained
from the infrared camera. Sensor degradation was accomplished as a result
of heavy rain showers. The degree of degradation encountered was generally
proportional to shower intensity.
Additional degradation was accomplished using many of the same tech-
niques performed for visual camera degradation shown in Figure 10.2
with similar results. This includes the use of chemical obscurants, physical
obstructions and weather conditions such as fog. All such degraded states
were detected.
Figure 10.3 Nominal and Degraded Infrared Camera Images.
Figure 10.4 Nominal and Degraded Radar Images.
10.5.3 Radar
Figure 10.4 illustrates nominal (left) and degraded (center, right) images rep-
resentative of different types of degradation that takes advantage of various
combinations of events, timing and signal characteristics to facilitate detec-
tion. All images provided below are from the Garmin 40 w, 48 nm digital
Radar system. The degraded image in the center illustrates the effect of
an obstruction introduced into the signal propagation path. Small, partial
obstructions through complete blockage of Radar signals was employed for
these experiments with varying time durations.
The degraded image to the right depicts the result of electromagnetic inter-
ference with the heading sensor upon which the Radar system depends for
proper geospatial orientation. The primary effect of this experiment was to
cause the Radar display to gyrate wildly resulting in small to extreme distor-
tion of the displayed imagery. Similar results were obtained with the Furuno
12 kw, 72 nm (analog) Radar using the same sensor degradation techniques.
Both degraded states were detected in addition to several more that are not
illustrated.
10.5.4 Sonar
Much of our effort focused on exploring the different types of degradation
that may be experienced by the various types of Sonar sensors likely to be
found on both conventional and unmanned vessels. These sensors included

side-scan, traditional echosounders, forward-looking navigation and chirp
Sonars produced by manufacturers that include Furuno, Garmin (shown
here), Lowrance, FarSounder, and EchoPilot.
10.5.4.1 Side-S can Sonar

Figure 10.5 illustrates side-scan Sonar images in both their nominal (top)
and degraded (middle and bottom) states. Degraded states were accom-
plished through the use of physical obscuration over a portion of the sensor
field of view, introduction of bubble screens and other natural phenomena
into the water column and the physical removal of the transducer from the
water as a result of performing high-speed turns with the vessel. All illus-
trated degraded states were detected in addition to several more that are not
illustrated.
Figure 10.5 Nominal and Degraded Side-S can Sonar Sensor Images.

10.5.4.2 Fish Finder Echosounder

Degradation experiments for echosounders with visual displays were per-
formed where the image on the left of Figure 10.6 (top) obtained from the
Garmin instrument depicts nominal echosounder performance while the
center and right images show degraded performance. Note that some ves-
sels are likely to be equipped with older and more primitive echosounders
that merely provide a numerical indication of depth below the hull. Such
devices are generally no-longer used in the commercial marketplace and the
research vessel is not so equipped. However, our underwater remotely oper-
ated vehicle (ROV) has high-resolution depth/pressure and temperature sen-
sors that are also used for such experiments.
10.5.4.3 Chirp Sonar

The scanning Sonar illustrated in Figure 10.6 (bottom) provides a nearly
photographic, wide image below the vehicle where structure, submerged
objects and fish can be discerned in nominal conditions. The nominal image
is provided on the left, with degraded images detected shown in the center
and right.
10.5.4.4 Forward-Looking Navigation Sonar

Several different forward-looking navigation Sonar systems were examined,
and their performance was degraded as part of our evaluation. Figure 10.7
(top) illustrates the Garmin ForVϋ navigation Sonar in both the nominal
and degraded states, while Figure 10.7 (middle) illustrates the nominal and
degraded states for their LiveVϋ navigation Sonar. The EchoPilot forward-
looking navigation Sonar imagery shown in Figure 10.7 (bottom) illustrates
Figure 10.6 N ominal and Degraded Fish Finder (top) and Chirp Sonar (bottom)
Images.
Figure 10.7 Nominal and Degraded ForVϋ, LiveVϋ and EchoPilot Navigation Sonar
Sensor Images.
sensor degradation where the starboard of two sensors mounted athwart-

ships on the bottom has become completely disabled due to bottom fouling.
The port sensor was also fouled extensively but resulted only in slightly
degraded performance.
Figure 10.8 provides an example of degradation using the FarSounder
Argos 350 forward-looking navigation Sonar resulting from an obstruction
in the field of view of the transducer. Additional cases of degradation were
attempted using bubble screens and other techniques in the field of view.
However, rather than resulting in degraded operation, the Sonar accurately
depicted the phenomena with little observable degradation. Further examin-
ation of resilience to degradation effects on this sensor is warranted.
10.5.5 Lidar
The Oster OS2-128 long-range Lidar system produces an image spanning
360° with the vessel centered in the middle. The top image in Figure 10.9
illustrates a nominal operation where there is a natural blind spot on the
navigation light array (outlined in the blue circle, left) directly below the field
Figure 10.8 Nominal and Degraded FarSounder Navigation Sonar Sensor Images.
Figure 10.9 Nominal and Degraded Lidar Images.
of view of the sensor. The imagery in the red circle (right) illustrates the area
that has been degraded in the bottom image as a result of an obstruction
placed in the laser propagation path that causes a shadow effect where detail
is lost.
This phenomenon was observed using several different types of obstruc-

tions of differing sizes, ranging from the smallest at 3 × 7 inches, which
resulted in the degradation shown in Figure 10.9, to 24 × 36 inches, which
caused massive shadows. All episodes of degradation within these obstruc-
tion sizes were routinely detected, while small sizes were marginally detect-
able depending on the background content.
The images shown in Figure 10.10 (top) represent the effects of interfer-
ence caused by the reflection of laser energy from nearby structures back
into the sensor from where it originated. The actual degradation mani-
fested itself as a rapidly blinking display from 0.5 to 2 times per second
where a large portion of the image was missing. This type of sensor deg-
radation was consistently detected without any difficulty. Degradation also
occurred, as shown in Figure 10.10 (bottom, left), resulting from the effects
of heavy rain very similar to that exhibited by the FLIR sensor shown in
Figure 10.3. Figure 10.10 (bottom, right) illustrates the effect of an insect
(wasp) landing on the sensor resulting in horizontal lines within the red
circle.
Additional degradation was attempted by inflicting strikes on the sensor
with a 30 mW blue laser (450 nM), 25 mW green laser (532 nM) and 3 mW
infrared lasers (780 nM and 905 nM). The green laser caused the Lidar
system to freeze the image and had to be restarted. The blue laser seemed to
cast shadows from the direction of origin, which were detected. The infrared
lasers caused interference lines as well as occasional blinking of the display,
which were also detected.
Figure 10.10
L idar Degradation due to Laser Interference (Top), Heavy Rain
(Bottom, Left) and Insect Landing on Sensor (Bottom, Right).
10.5.6 Global Positioning System (GPS)/G NSS

Various efforts were undertaken to achieve DoS attacks on the Garmin GPS/
GNSS, without success. The problem stems from the ubiquitous nature of
modern displays where this capability is built into each display, and each is
connected to a common NMEA-2000 bus. Our research vessel testbed has
four Garmin 8616/8612xsw displays, as shown in Figure 10.1 (top, center)
that can function without an external antenna. Each time one or more units
were disabled through physical obstruction and electrical isolation of satellite
signals there was a seamless transition to other active devices. Attempts to
simultaneously isolate all receivers proved to be impractical. Consideration
was given to attempting to electrically interfere with the satellite signals using
our onboard SDR transceiver generating low power (<100 mW) complex
signals. However, our geographic proximity to several large metropolitan air-
ports and veritable nonstop activity of nearby commercial, general aviation
and military aircraft within the line of sight was determined to be too risky
and this approach was not attempted. However, the effect of such transmis-
sions on various different types of GNSS receivers in close proximity would
certainly prove interesting.
An alternative approach used our legacy Furuno GP-37 GPS receiver that
was not connected to the NMEA-2000 bus, but rather to a fully independent
NMEA-0183 bus shared by the Advanced Navigation INS. In this case we
were successful in isolating this receiver by insulating the antenna from satel-
lite signals using metal screening. This resulted in the activation of the WAAS/
DGPS ERROR! Alarm, the alarm icon appearing on the user interface screen
and reporting the condition over the NMEA-0183 bus, which was detected.
Degradation was also detected using bottom-depth contour tracking in con-
junction with INS experiments.9,10
10.5.7 Inertial Navigation (INS)

The Advanced Navigation GNSS Compass/INS provides an accurate dual
antenna GNSS-based heading that is not subject to magnetic interference
and can maintain accurate heading during GPS outages. Positioning data
is provided across the NMEA-0183 bus. During a loss of heading fix the
GNSS Compass will continue to provide accurate heading for up to 10 min-
utes using its gyroscopes. A heading error will slowly accumulate during this
period at a rate of approximately 15 degrees/hour. During a loss of position
fix the GNSS Compass will continue to provide accurate position and velocity
for up to 2 minutes using its inertial sensors.11 This receiver was successfully
isolated by insulating the antenna from satellite signals using metal screen-
ing. Real-time tracking of actual GNSS position was accomplished with the
Garmin GPS using the NMEA-2000 bus, while tracking of INS positioning
was accomplished over the NMEA-0183 bus.
Figure 10.11 Inertial System Degradation over Time without GPS/G NSS Inputs.
Figure 10.11 illustrates the progression of INS error over time along an
approximately 12-mile track, with INS positioning depicted by the red line
while vessel positioning via GPS is shown by the black dashed line with the
boat icon. The black dashed line track to the right of the vessel track is an
artifact remaining from a previous voyage and should be ignored.
GPS/GNSS signals were disrupted upon voyage commencement. The top-
left image depicts the beginning stage of the voyage at a point approximately
10 minutes from leaving the pier where the INS was still providing accurate
positioning and heading information. A series of turns during this period
appeared to hasten the onset of positioning error while heading informa-
tion remained relatively accurate. The top-right image progressing into the
bottom-left image depicts a steady positioning error that continued relatively
unchanged while navigating a relatively straight line. At a point close to voy-
age termination after approximately 45 minutes, the track shown in the bot-
tom right image shows the INS course has degraded significantly.
Detection of INS degradation was accomplished as a GNSS reception outage
was reported by the system over the NMEA-0183 bus. Secondly, positioning
error was detected through bottom depth and contour tracking that determined
local depths at the vessel position did not match those reported by the INS but
indeed corresponded to a position hundreds of feet away from the vessel.8,9
10.5.8 Heading Sensor

Degradation to the heading sensor was introduced through the application
of electromagnetic fields in close proximity to the sensor. This resulted in
significant errors in instruments such as Radar and ENC dependent upon the
heading sensor and required recalibration of the ENC display. Events stem-
ming from heading sensor degradation were detected and are discussed under
the corresponding instruments.
10.5.9 Automatic Identification System (AIS)

The primary mode of AIS degradation we attempted to detect was the effects
of many transmitters that tend to overload AIS receivers, resulting in large
blank spots being displayed in the AIS field of view. Such conditions are
experienced primarily in areas of Asia where AIS is used to mark fishing nets
and buoys in addition to the normal use of this technology. Other examples
of degradation can be experienced through DoS and spoofing attacks. All
such examples require the transmission of significant (>100 mW) VHF radio
signals on AIS frequencies, which we determined to be potentially dangerous
and outside the scope of our investigation. However, we were successful in
achieving a denial-of-service attack using low-power (<100 mW) VHF trans-
missions of complex waveforms that resulted in a blank AIS display, which
was consistently detected.
10.5.10 Engineering Sensors

There are a wide variety of engineering sensors and data types that are
interconnected via the NMEA-2000 bus network.12 Similar information is
available on the NMEA-0183 and other bus networks. However, their data
communication protocols and formats can vary widely but the concepts
involved are applicable to many of the prevailing instrument and sensor bus
technologies found in commercial maritime, naval applications such as the
U.S. Navy Unmanned Maritime Autonomy Architecture (UMAA),13 as well
as for aviation, automotive and railroad use.
The primary focus was centered primarily on data that was available from
the various sensors utilized in our experiments, including:
• AC Power
• DC Power
• Depth
• Engine
• Environment
• Fuel Flow
• GPS
• Heading
• Navigation
• Pressure
• Rudder
Figure 10.12 Degraded Engineering Sensor Detection via NMEA-2 000 Data Bus.
• Speed
• Tanks
• Temperature
• Time
• Vessel
• Wind
Sensors from the above list were evaluated in the detection of sensor deg-
radation. Several of these are shown in Figure 10.12. Moving from top left
to bottom right, these images denote sensor degradation to various degrees
reportable by the sensors themselves over the NMEA-2000 bus and include
instances where Radar service was lost, low DC voltage levels were detected,
the connection with the Heading Sensor was lost, and satellite weather ser-
vice was no longer available. This means of discerning sensor degradation
and anomalies over the NMEA-2000 bus complements the three-domain
approach (Time, Frequency and Spatial) and can provide greater clarity and
specificity that would enable not only the detection of sensor degradation
but also the identification of what types of degradation exist and possibly the
means to overcome the degradation.
10.5.11 Weather Sensors

Degradation of onboard weather sensors was performed by introducing
external stimuli and situations that would interfere with normal operation.
Figure 10.13 Nominal and Degraded Weather Sensor Images.
Figure 10.14 Nominal and Degraded Satellite Sensor Images.
This includes placing a box over the weather sensor, spraying the sensor with
streams of water, applying a heat gun to artificially increase ambient tempera-
ture and subjecting the sensor to high winds from a local source that included
a fan. Figure 10.13 illustrates the effects of a stream of water that resulted in
maximum readings in wind speed and relative humidity, abnormal tempera-
ture readings and other measurements that were uncharacteristic and not
representative of the physical environment. Anomalies were readily detected
related to water, temperature and wind while isolating the sensor from the
environment. However, wind anomalies were not readily discernable from a
calm day without wind and with steady temperatures.
10.5.12 Satellite Sensors

Degradation of reception of satellite transmissions for weather and other
services was performed by isolating the receiving antenna through the use
of metal shielding. The loss of signal resulted in the degradation shown in
Figure 10.14 which was easily detected.
Note that in the absence of signal the weather display became blank.
However, the AIS overlay still continued without interruption.
Figure 10.15 Microphone Placement and Audio Waveforms.
10.5.13 Microphone
Audio sensors are vital to ensure compliance with COLREGs Rule 5 require-
ments to maintain a proper look-out by (sight and) hearing.14 Degradation
may result in a lessening or complete inability to hear sounds such as horns,
whistles, sirens and bells emanating from other vessels as well as aids to
navigation that include buoys and lighthouses. This experiment utilized
microphones built into the six onboard cameras oriented at 60-degree angles
surrounding the vessel. The Camera 1 microphone illustrated in Figure 10.15
was degraded using a covering resulting in lower sensitivity as illustrated by
the reduced and distorted sound recorded on the corresponding waveform.
The degradation illustrated for Camera 1 and similar events directed at the
other microphones were consistently detected with directionality determined.
10.5.14 Electronic Navigation Chart/E lectronic Chart

Display System (ENC/E CDIS)
Sensor degradation as may apply to ENC and ECDIS was investigated and
determined that the effects would very likely appear as errors and incon-
sistencies in the chart display. However, such events are most likely to ori-
ginate from the navigation, AIS and other sensors that feed data to the ENC.
Therefore, we determined the best means to detect degradation would be
directly from these sensors themselves rather than through the ENC which
is secondary to the sensor. Opportunities to further investigate ENC/ECDIS
degradation detection may prove to be useful.
10.6 Image Snippets for Remote Analysis

In each case where sensor degradation detection was achieved a small snippet
of degraded sensor imagery is captured that could be sent to a Remote
Figure 10.16 Image Snippets to Return to Remote Control Center for Analysis.
Figure 10.17 Transmission of Snippets via Email Using High Frequency (HF) Radio.
Control Center (RCC) via low bandwidth communication channel for human
analysis at a RCC as well as machine analysis. Snippets, samples of which
are shown in Figure 10.16, are 160 × 145 pixels in size with each requiring
memory from 62k and 118k of memory, without compression. Compressed
snippets would result in even smaller memory requirements.
This is an example of just one way to present this information in a manner
useful to staff at a remote-control center. Figure 10.17 illustrates an example of
how the communication of snippets was accomplished using high-frequency
(HF) radio communications whereby imagery from three sensors was for-
warded from the research vessel directly to our offices for use by a remote
operator. HF radio was selected to demonstrate this capability over a low-
bandwidth communications channel that can span long distances assuming
that satellite communications were unavailable. Other methods to achieve
the same result using broadband communications are readily available.
10.7 Remote User Interface

A user interface, illustrated in Figure 10.18, was created for the display of
sensor status and sensor image snippets at an RCC.
Sensors that are currently not configured into the active vessel architecture
are grayed out while active sensors are colored in red, orange, yellow and
Figure 10.18 User Interface for Remote Display of Sensor Status.
green to illustrate general sensor degradation status. A number from 0 to 100

is also shown to indicate degree of sensor degradation.
10.8 Sensor Degradation

Detection of the existence of sensor degradation is based upon determin-
ation that the characteristics of the examined Frequency, Time and/or Spatial
Domain sensor signal representations differ from the characteristics associ-
ated with nominal sensor operation as embodied within the training data sets
used during decision-making software development. This process is tailored
to the data characteristics of specific sensor types and models with respect to
degrees of variation from nominal, and nothing else. All examples of degrad-
ation presented in this chapter were detected in this manner.
The training requirements for detection presume a very large volume of
data across the three domains over long periods of time, preferably with con-
tinuous learning so that the training data will continue to improve throughout
the useful life of the sensor(s). This data must contain representative samples
of nominal operating conditions for all sensors involved. Examples include
day, night and twilight; calm to rough seas, fair to stormy weather, empty
screens to ones filled with targets and shallow to deep waters. This is an area
where deep-learning AI takes a lead role to ensure that robust and compre-
hensive training is achieved with minimal false indications of degradation.
10.8.1 Identification
The capability to identify specific causes of sensor degradation requires a sig-
nificant investment in terms of further training using the specific characteris-
tics of degraded sensor signals with one or more of the three domain signal
representations as may be available and useful. The visual camera images
representing the Spatial Domain shown in Figure 10.2 illustrate the occur-
rence of specific causes of degradation that are sufficiently different to enable
positive identification based upon the unique characteristics of each. Specific
degradation classes and types can then be created for training using machine
learning techniques.
A least two different classes of degradation are presented in Figure 10.2.
These can be classed as partial or full degradation based upon the extent of
sensor surface area affected. Partial degradation affecting only part of the
sensor image can also be seen in the center Radar image of Figure 10.4,
the center-left Sonar image of Figure 10.5, both the center and right Sonar
images of Figure 10.6 and the Sonar images on the right side of Figures 10.7
and 10.8; and the Lidar images at the bottom of Figure 10.9 and right side of
Figure 10.10. Full degradation that affects 50% or more of the sensor image
can be seen in many of the other figures. Degradation shown in Figure 10.2
can be attributed to damage, obstruction, obscurant and natural-fog and
natural-rain in the following order as shown in Table 10.2:
The examples cited represent identification of the causes of degradation
based only upon the unique characteristics presented from Spatial Domain
perspectives. This alone may be sufficient to accurately and consistently
identify particular types and sources of degradation. In cases as shown in
Table 10.2; frames 4, 7, 11 and 12 of Figure 10.2; additional insight may
be needed to determine proper classification and type. In such cases unique
characteristics associated with each degradation may also exist within the
Frequency Domain as the properties of each causal agent may filter specific
frequencies of light entering the camera and possibly introduce new frequen-
cies that are not naturally present. These inter-domain correlations can help
to provide greater specificity in identification. Examination of Time Domain
characteristics can help to identify the exact timing degradation began,
whether it occurred quickly or gradually, the extent of degradation, and
Table 10.2 Identification of Classes and Types of Degradation Shown in Figure 10.2
1. PARTIAL, DAMAGE 2. FULL, OBSCURANT 3. FULL, DEBRIS

4. FULL, OBSCURANT 5. FULL, FOG 6. PARTIAL, FOG
7. FULL, RAIN 8. FULL, DAMAGE 9. FULL, OBSCURANT
10. FULL, OBSCURANT 11. FULL, OBSCURANT 12. FULL, DAMAGE
whether it continues unchanged or if the sensor is recovering from the effects

of the degradation agent through evaporation, dislocation or other scenario.
10.8.2 Remediation
The capability to overcome the effects of sensor degradation may be
achieved through the use of sensor fusion techniques to derive the contents
and meaning of degraded data provided by one sensor by examining data
from other sensors to arrive at the same or similar conclusions. Such an
approach can provide greater resiliency in decision making by using sen-
sor data obtained from different portions of the electromagnetic spectrum
(Frequency Domain) over the same periods of time (Time Domain) while
providing fixed reference points for comparison between different sensor
data products (Spatial domain). Further, such an approach may provide a
second source for information that is normally gathered from another sen-
sor. Several examples are provided within this paragraph to demonstrate the
potential of this approach to insure greater resilience in sensor degradation
detection performance.
10.8.2.1 Visual, Infrared, Lidar, and Microphone

The fusion of multiple sensor data streams can be used to reinforce the con-
clusions reached by individual sensors and to eliminate ambiguities as to
which sensors may actually be degraded. Figure 10.16 illustrates snippets
of visual and infrared cameras as well as Lidar that have been degraded as
a result of heavy rain showers. Such data can be supplemented with audio
obtained from microphones to further reinforce conclusions. The simultan-
eous reporting of degradation by sensors that utilize dissimilar methods at
different points within the radio frequency spectrum can provide increased
confidence in degradation detection performance.
10.8.2.2 Visual and Weather

Visual camera imagery can reduce ambiguity in determining if sensor meas-
urements accurately depict the external environment. In the case of the wea-
ther sensor illustrated in Figure 10.19 showing winds measured at 107 knots,
a simple comparison with live imagery indicates a sea state much more appro-
priate to calm rather than hurricane conditions. The introduction of such
capabilities can be accomplished through training of visual camera neural
networks to readily discern between different sea states at various Beaufort
Wind Force Scale conditions.
Figure 10.19 Example Use of Sensor Fusion to Remediate Sensor Degradation.
10.8.2.3 Echosounder and ENC/E CDIS

Several references have been made throughout regarding the use of bot-
tom depth and contour tracking as a means to detect degraded Global
Positioning System and Inertial Navigation System sensors that provide
errors in geographical positioning.9,10 ENC/ECDIS hydrographic data is geo-
referenced to position by latitude and longitude and can be used as an alter-
native source of information to the echosounder regarding depths at any
particular location. Our approach utilizes these data to make comparisons
between charted depths and live echosounder readings to help determine
whether degradation may exist in either position sensors or the echosounder.
In well-charted areas with recent hydrographic survey and adjusted for tide
information these measurements should be relatively close. In areas lack-
ing recent hydrographic survey or after significant storms that may cause
shoaling and changes in seabed configuration these values may vary widely
warranting notification to navigation processes and remote controllers that
caution should be exercised.
10.8.2.4 AIS, Radar, and Visual/Infrared

Many vessels are not equipped with Automatic Identification System (AIS)
technology while others intentionally go dark by turning off their AIS tran-
sponders making it difficult to obtain a positive identification of vessels while
underway. Oftentimes AIS signals are spoofed such that a vessel may appear
at a location that is miles to thousands of miles distant from their actual loca-
tion. Further, denial of service attacks on AIS sensors may also make them
inoperable. All of these instances provide examples where AIS capabilities
are degraded through intentional or nefarious action even though the sen-
sors themselves are perfectly operational. In such cases it is possible to utilize
Radar, visual cameras and/or infrared cameras in an attempt to resolve ambi-
guity as to whether an AIS sensor data is degraded or not. They can also be
used to determine the presence or absence of a vessel and possibly to identify

such vessels.
10.8.2.5 Microphone, Visual, Infrared, Lidar, Weather and Satellite

Sound information obtained from microphones may reinforce indications
obtained from other sensors, or they may weaken and challenge the conclu-
sions reached through analysis of other sensor data. For example, sound can
boost assertions from visual and weather sensors that it is indeed raining
heavily, or snowing, or encountering hail. Likewise, the presence of unusual
sound or the absence of sound may indicate nefarious activity. In all cases
where sensor degradation is indicated, the use of microphones strategically
placed throughout the vehicle can help resolve ambiguities as to the source
and extent of degradation that may exist.
References
1 Jim Myer. NMERPAC to look at mariner shortage. Waterways Journal. 2
September 2022. National Merchant Marine Personnel Advisory Committee.
U.S. Coast Guard. www.waterwaysjournal.net/2022/09/02/nmerpac-to-look-
at-mariner-shortage/
2 R. Glenn Wright. In-Stride Detection of Sensor Degradation. GMATEK, Inc.
Final Report. Revision: A, 28 April 2022. Contract: N6833520G2005. Naval
Sea System Command, Washington, D.C.
3 Wright, R. Glenn. Signals Intelligence Automated Assessment of Test
Capabilities. IEEE Automatic Testing Conference. Washington, DC. September
2018.
4 John Boyd. New Security Technology Detects Malicious Cyberattacks on
Drones, Cars and Robots. 26 February 2019. https://spectrum.ieee.org/new-
security-technology-detects-attacks-on-sensors-controlling-numerous-applicati
ons-including-drones-cars-and-robots
5 Michael Koziol. New Ethernet Cyberattack Crunches Critical Systems. 15
November 2022. https://spectrum.ieee.org/cyberattacks
6 A. Aboulian, et al. (2019) NILM dashboard: A power system monitor for elec-
tromechanical equipment diagnostics. Trans. on Industrial Informatics, 15(3),
1405. DOI: 10.1109/TII.2018.2843770
7 DARPA Assured Autonomy Project. www.darpa.mil/program/assured-auton
omy
8 Fouling Ratings (FR) in Order of Increasing Severity. Naval Ship’s Technical
Manual. Waterborne Underwater Hull Cleaning of Navy Ships. S90860CQ-
STM-010, rev. 5. Chapter 081. Naval Sea Systems Command. 1 October 2006.
Table 081-1-1.14.
9 Wright R. Glenn and M. Baldauf (2016). Hydrographic survey in remote
regions: Using vessels of opportunity equipped with 3-dimensional forward-
looking sonar. Journal of Marine Geodesy, 39(6), 439–357. DOI 10.1080/
01490419.2016.1245266
10 Wright, R. Glenn. Enhanced MASS Situational Awareness Using Virtual

Navigation Aids, Second International Conference on Maritime Autonomous
Surface Ships (ICMASS 2019). Trondheim, Norway, 13–14 November 2019.
11 Advanced Navigation. GNSS Compass Reference Manual. Version 1.0. 23
January 2018, p. 36.
12 Maretron. Supported Display Data Types DSM570 NMEA 2000® Multi-
function Color Graphic Display, User’s Manual. p. 260, Revision 1.8.3.
13 U.S. Navy. Unmanned Maritime Autonomy Architecture (UMAA). PMS 406.
Unmanned Maritime Systems. Naval Sea Systems Command. www.navsea.
navy.mil/Portals/103/Documents/Exhibits/SAS2021/SAS-2021-Unmanned
MaritimeSystems.pdf
14 COLREGS, Rule 5. Look Out. Convention on the International Regulations
for Preventing Collisions at Sea, 1972, (72 COLREGS). Navigation Rules
COMDTINST M16672.2D. United States Coast Guard.
Chapter 11
Next-G eneration Sensing
The primary difference between conventional ships and unmanned and

autonomous ships is the latter incorporate enhanced sensor suites, sensor
data fusion technology, computers and communication systems that allow
them to operate without seafarers onboard. These technologies that benefit
efforts towards the automation of ship operations continue to be integrated
onto the bridges of conventionally staffed vessels to assist seafarers in the
performance of their duties.
One outcome of these events is the orders of magnitude increase in the
volume of imagery, signal and digital data generated by sensors on ships
in recent years that drive enhancements to hardware and software leading
to new and improved processes that make better use of these resources.
The promises of innovation in the scope of onboard sensors, their uses and
applications, and the data they generate have only just begun to be tapped.
These provide new perspectives from which enhanced situational awareness
can benefit all engineering functions and increase the safety of navigation,
cargo handling and transportation; and passenger safety, comfort and ser-
vices. Examples include revealing actual and pending deficiencies in sensors
and ship systems, predicting and avoiding nearby vessel traffic with greater
precision, improving ship routing and vessel efficiency, detecting hazards to
navigation, expanding the roles of sensors to achieve new capabilities, and
considering previously unknown mechanical, electrical and environmental
factors associated with vessel operations.
This chapter attempts to highlight several advances in sensor technology as
apply to present, developing and entirely new generations of conventionally
manned, unmanned and autonomous vessels. These sensor technologies and
new inventions must be considered in terms of functional, practical, legal
and ethical impact on future enhancements to the International Maritime
Organization (IMO) Safety of Life at Sea (SOLAS) Convention, the Polar
Code, the International Convention on Standards of Training, Certification
and Watchkeeping for Seafarers (STCW), Convention on the International
Regulations for Preventing Collisions at Sea (COLREGs) and other inter-
national codes, conventions, guidance documents and practices. Development
Doi: 10.1201/9781003377900-11
202 Next-Generation Sensing
of these enhancements and their implementation must be assisted by the

many maritime nation-states, shipping companies and labor organizations,
classification societies, insurers and non-governmental organizations.
11.1 Ship-B ased Drone Vehicles

There has been a massive increase in the use of drone technology throughout
all industries since 2016 upon the granting of broad exceptions by aviation
regulatory authorities allowing their routine operation by companies for
many uses. Total global shipments reportedly will reach 2.4 million in 2023
and are increasing at a 66.8% compound annual growth rate.1 This same
report cites Goldman Sachs forecasts of total drone market size being worth
$100 billion, supported by the growing demand for drones from the commer-
cial and government sectors. Drone use by the maritime industry is following
this same trend with many applications related to logistics and resupply at
sea as well as other uses for man overboard search and rescue, hull and cargo
inspection, scouting for navigation in ice and hazards to navigation.
One study published in 2020 supported by a North Sea Region EU grant
cited three areas where drone use would likely become accepted practice by
2025: multipurpose drones for vessel navigation support, document delivery
and man-overboard rescue.2 Although the timeline cited may be subject to
debate, their conclusions are generally in line with current industry events.
11.1.1 Logistics and Resupply at Sea

Rather than simply focusing on the delivery of documents, there is great
interest in drone use for the delivery of supplies for ships underway as well
as at anchor. One example occurred in January 2016 when Maersk Tankers
began drone testing by making a flight of 250 m from one of its barges
to a tanker and drop off a package, and again in March 2016 upon com-
pleting drone delivery to a ship to initiate a wider program to test whether
unmanned aerial vehicles (UAVs) can be used and implemented for supplying
vessels with spare parts, mail or medicine.3,4 In the latter case, a package was
successfully dropped from a height of 5 m above deck level. The use of small,
lightweight UAVs to deliver small packages was seen as a means to eliminate
the need for more conventional means of delivery as a way to cut costs.
In another test in 2019, Wilhelmsen Ship Service and Airbus completed a
trial shore-to-ship drone deployment to a vessel at anchor in Singapore.5 This
test originated from the Marina South Pier with a cargo of 3D printed con-
sumables where the Airbus Skyways drone navigated autonomously along
pre-determined aerial corridors in its 1.5 km flight to the Eastern Working
Anchorage. The drone took off, landed on the deck of Pacific Offshore anchor
handling tug supply vessel M/V Pacific Centurion Swire, deposited its 1.5-kg
cargo, and returned to land in 10 minutes.
Next-Generation Sensing 203
Similar efforts were undertaken by the U.S. Navy in 2021 using a small
drone to carry lightweight cargo between vessels to eliminate the need for
helicopter flights for a substantial fraction of its spare parts logistics.6 Their
plan is to use a commercial drone to carry up to 24 pounds of cargo for a
distance of 65 miles, which would be sufficient to airlift about 80% of the
Navy’s critical parts cargoes.
These details are indicative of one aera in which interest in maritime drone
use is flourishing. However, since this is a book on ship sensors, the rest of
this paragraph will focus on drone UAV use as a means to extend sensor
presence beyond the confines of the ship or vehicle to achieve other purposes.
11.1.2 Man Overboard Search and Rescue

The use of autonomous and remotely piloted air vehicles to locate people
in the water and possibly to drop flotation devices and other equipment has
long been a need amongst ship crewmembers and shipping companies alike.
With advances in drone sensor technology and increased drone payload
capacity, this need can now be filled and should take high priority for ship
owners and regulators in amending the IMO Convention on Safety of Life at
Sea (SOLAS) to mandate the installation and use of such equipment on board
medium to large vessels.
Worldwide incident statistics indicate 386 cruise and ferry passengers and
crew went overboard between 2020 and 2023.7 In the United Kingdom,
overall, 40% of man overboard incidents resulted in death as reported by
the UK Marine Accident Investigation Branch (MAIB). They examined 308
man overboard incidents between 2015 and 2022 from commercial and rec-
reational boats, both at sea and inland.8 The highest fatality rate was within
the fishing industry where, out of 58 reported man overboard incidents, 33
(56%) resulted in a fatality. The second highest fatality rate occurred within
the recreational boating sector, where 144 man-overboard incidents occurred
with 69 (48%) resulting in a fatality. Worldwide statistics are difficult to
obtain, but one study indicates a general assumption that approximately
1,000 people fall overboard annually.9 A good source in the literature deal-
ing with this subject for the period between 2011 and 2018 was advanced by
Gönel and Çiçek.10
There has been a great deal of research in developing and testing drone
solutions to the man overboard problem. One example is where members of
the United Kingdom Royal NavyX team tested a heavy-lift Minerva drone
UAV in delivering potentially life-saving equipment if someone falls over-
board.11 Over a several-month period in 2021, the team tested remotely
piloted systems which locate personnel in the water, drop life-saving equip-
ment and hover over the location until rescuers arrive. This was accomplished
using a Minerva T-150 at the Royal Navy’s Diving School at Horsea Island,
Portsmouth and also on a civilian boat using a Minerva T-80 drone. The
system was able to be controlled to take off the moving vessel and fly out to
sea, successfully locating a dummy in the water before returning to land itself
on a mat attached to the boat’s deck.
Remotely piloted and autonomous UAVs equipped with both visible light
and infrared cameras can play an essential role in speeding up the detection
and rescue of people who have fallen overboard. High-resolution imagery
can provide details of partial human forms in the water that are recog-
nizable by remote UAV pilots as well as artificial intelligence (AI)-based
smart cameras during daylight and low-light conditions in a variety of sea
states. Infrared cameras can supplement visible light cameras to also detect
human forms in the water using their heat signature that provides a sharp
contrast to the generally colder water. One researcher has also developed
and publicly released specialized annotated datasets for training and testing
AI-based detectors for this task. This MOBDrone benchmark is a collection
of more than 125K drone-view images in a marine environment under sev-
eral conditions, including different altitudes, camera shooting angles and
illumination.12
11.1.3 UAV Use for Sea Ice Navigation

Drone UAVs have become very handy in extending the vision of crewmem-
bers from the bridge to provide overhead views of sea ice forward of the ves-
sel to plan efficient routes through pack ice. Operations have been conducted
at opposite ends of the planet in both the Arctic and Antarctic. Two of many
examples are discussed in which the utility of UAVs to provide assistance in
ice navigation has been proven under operational conditions.
In 2015, the Antarctic Division of the Australian Department of Climate
Change, Energy, the Environment and Water used a quadcopter drone to
help navigate the research and supply ship Aurora Australis through the sea
ice on its annual resupply voyage to Casey station.13 Flown off the aft deck of
the ship, the UAV provided real-time imagery of the sea ice conditions ahead
of the ship to help the crew optimize navigation decisions. One interesting
aspect of this experiment was the fact that the drone’s compass was adversely
influenced by the ship’s thousands of tons of steel potentially causing the UAV
to behave unpredictably. The problem was solved by disabling the drone’s
GPS and flying it in full manual mode enabling the drone to operate success-
fully and deliver the required video imagery. During the nine-day voyage the
UAV made five flights voyage to Casey requiring eight minutes in the air for
each flight to collect the necessary data. A fixed-wing drone was also used to
survey the area around Casey station from the air, mapping 30 hectares (74.1
acres) in 20 minutes. Other observations included the UAV can be launched
in minutes, costs little to operate and saves valuable cargo space as it needs
only a 4 m × 4 m space for launch and recovery.
On a project managed by Fairweather, LLC of Anchorage, Alaska, in the

summer of 2016 a month-long voyage was successful in retrieving large
mooring (anchor) systems used for offshore drilling platforms from along
the Beaufort and Chukchi Sea floors.14 During this voyage, offshore tug/
supply ship Aiviq used an Aerovel Flexrotor drone launched off the helideck
to survey the vessel’s path guiding it through the least amount of ice. The
UAV flew five flights hovering at several thousand feet totaling 19 hours
and guided the fleet through otherwise slow seas choked with ice, saving
the ship more than two weeks in their schedule.15,16 Also used on this voy-
age were Sonar and underwater Remotely Operated Vehicles (ROV’s) to
precisely locate the anchors’ positions prior to the arrival of the anchor
handling vessels.
11.1.4 Ship Inspection via UAV

In 2020, cargo claims consultancy WK Webster performed an initial damage
survey of the container ship One Apus using a remotely piloted drone while
en route as the ship transited Osaka Bay.17 Company inspectors in Kobe,
Japan, also conducted a remote drone survey upon arrival. The imagery was
taken by two drones using ultra-high-definition cameras that clearly showed
the scale and detail of the devastation on board.
The trend towards more autonomous operations is a continuation of
an approach already being used in shipyards. In September 2015, the
Remontowa Shiprepair Yard in Gdańsk Poland remotely inspected internal
spaces of a ship following an overhaul.18 By using a high-definition camera
Remontowa was able to record video and take still images of one of the cargo
tanks to examine the condition of protective coatings and detect corrosion
and cracks. In June 2016, the UK Royal Navy used drones to identify defects
in HMS Diamond’s exteriors and reach inaccessible areas. Such capability
was seen as being able to reduce time and cost during maintenance periods
and allow ships to diagnose their own defects while still at sea. In the United
States, a quadrotor measuring 23 in wide and 12 in high gathered informa-
tion relating to onboard fire-fighting on a former US Navy ship. The UAV
was able to navigate a smoke-filled location to map areas and locate fire,
providing a clear safety benefit. These three examples illustrate the untapped
potential for smaller drones not only on large unmanned vessels but across
all vessels of all types and performing various missions and functions.
Another land-based mission is the use of drones to monitor vessel emissions
into the air and water. This is being tried in Istanbul by the Metropolitan
Municipality’s Marine Services Directorate to monitor the Bosphorus by
taking photographs and video footage. Under a proposal by the European
Maritime Safety Agency (EMSA) and the European Space Agency, drones
could also be equipped with specialized sensors to detect and classify specific
types of stack emissions such as sulfur and carbon dioxide in the English
Channel, the North Sea, the Baltic Sea and the Gulf of Bothnia.
11.2 Forward-L ooking Navigation Sonar (FLS)

Although Forward-Looking Navigation Sonar (FLS) has existed for decades,
this valuable and highly overlooked resource has been virtually ignored as a
proven means to help prevent groundings, collisions, allisions and the result-
ing loss of life, property and large-scale destruction of the environment. This
paragraph describes FLS technology and presents a case study of one of many
similar accidents where it would have positively detected the hazard that
existed, and bridge watchstanders could have been notified well in advance
in sufficient time to avoid disaster.
An important aspect of basic and enhanced training in ship handling
and navigation courses is the familiarization with and handling of equip-
ment, including transferring knowledge and experience on the capabilities,
functionalities and constraints of their use. Bridge navigational equipment
encompasses many and diverse systems to support safe and efficient navi-
gation and protection of the marine environment. The present situation is
characterized by an increasing level of integration of sensors, technical sys-
tems, displays and sophisticated decision support systems combined with
complex alerts to ensure sufficient situational awareness of the bridge team.
However, despite the presence of such sophisticated systems, accidents
in the form of groundings and collisions still happen. Regardless of the
behavior, actions taken or not taken by the bridge team, the grounding of
Costa Concordia with the bridge equipped with very modern Integrated
Navigational Systems (INS) can be seen as another prominent case where
alerts implemented in the most modern and highly sophisticated navigation
systems failed to raise the attention of those tasked to take action and avoid
an accident.1
11.2.1 The Problem of Blind Navigation

The COLREGs and STCW require an adequate watch be maintained.19,20
However, there is presently no established means to maintain a watch below
the waterline to directly detect the presence of hazards to navigation. A key
assertion of this author’s research is that bridge watchstanders are generally
provided only with indirect and/or secondary information from which safety-
critical decisions are routinely formulated. Although the tragic grounding of
Costa Concordia provides a convenient and fresh example to illustrate this
assertion, the problem has and continues to exist on a daily basis during virtu-
ally every vessel transit. For example, nautical charts and even most modern
Electronic Chart Display Information System (ECDIS) provide navigation
information as a secondary reference created by survey at some time in the

past that was accurate at the time it was made but not necessarily reflecting
actual conditions at the present time of passage. Radar indicates only those
nearby targets from which the proximity to underwater hazards to naviga-
tion should be deduced. A traditional echo sounder provides depth directly
below the keel, but no indication of depth directly forward of the bow. These
and other methods such as GNSS merely infer positional relationships to
geographic locations of interest and concern, but none are capable of directly
detecting the physical underwater Hazard to Navigation (HtoN) consisting
of Scole Rocks off Isle del Giglio, or any other such hazard prior to ground-
ing. In addition, there are many floating hazards that cannot be charted such
as drifting shipping containers, debris fields, whales, etc., having little or no
presence on Radar, other types of Sonar systems, or other navigation sensors.
Another assertion is that the technology needed to detect hazards to
navigation in real time at and below the waterline currently exists in the
form of FLS and is available as commercial off-the-shelf equipment ready
to be installed on vessels at the next scheduled drydocking. This does not
imply, however, that this technology is sufficiently tested, capable of being
integrated directly into existing navigation systems in its present form,
or adequate training exists for its use. Merely adding yet another sensor
and display system to an already complex bridge environment without
adequate engineering, planning and training is likely to make matters
worse rather than improving them, resulting in increased risks to naviga-
tion. This was illustrated by the introduction of Radar that led to the 1956
Radar-assisted collision between M/Vs Stockholm and Andrea Doria. Lack
of proper training in the operation, use and interpretation of Radar equip-
ment combined with a lack of procedures for implementing corrective
actions based upon Radar indications created a scenario where established
procedures for navigating in restricted visibility were not followed, result-
ing in the collision.21
The capabilities and limitations of FLS technology are discussed in terms
of their scope and capability to detect hazards to navigation, along with its
suitability in terms of vessel type, speed and installation requirements. Also
covered is the complexity of using FLS for vessel navigation describing pre-
sent capabilities as well as future requirements in terms of standards for INS
and bridge alerts. This includes the introduction of an electronic Virtual Aids
to Navigation (VAtoN) as a means to display hazards to navigation on elec-
tronic chart display and information systems (ECDIS). Integral to this pro-
cess is the use of simulation technologies to test, verify and validate system
processes, procedures and training requirements well in advance of the intro-
duction of this technology to vessels by the introduction of carriage require-
ments through the IMO.
11.2.2 Forward-Looking Sonar (FLS) Technology

Active Sonar, commonly referred to as an echo sounder by the IMO, is used
by vessels to determine the depth between the keel and the sea bed.22 A vari-
ation of the echo sounder is FLS used to detect bottom features and objects
within the water column forward of the bow. Despite its usefulness and the
availability of this technology in the commercial marketplace it is rarely
included within the ships’ complement of navigation sensors.
The methods used for detecting bottom features, objects and soundings
by determining range, azimuth and elevation information can generally be
described as variations on transmitting a steerable Sonar signal ahead along
the path of the vessel or transmitting a single ping from which snapshots of
the environment are obtained.23,24,25,26 A mosaic of the bottom topography
and specific targets is then created as the vessel progresses on its course.
The range of FLS can extend from eight to twenty times the depth ahead,
depending on bottom and target conditions. It is most effective when the
bottom topography slopes upwards and when targets are large and consist
of hard rock and/or coral that provide good acoustic signatures. FLS prod-
ucts are available that provide both two-dimensional (2D) as well as three-
dimensional (3D) representations that provide a more realistic portrayal of
the course ahead.
FLS systems have been developed with different capabilities supporting
both autonomous underwater vehicle and vessel applications. Of those
designed for use on vessels, most are intended for pleasure and small fishing
boats. There are systems with range and resolution that make them suitable
for use on larger vessels such as workboats, offshore service vessels, merchant
and passenger vessels. However, operational constraints may create limita-
tions on their usefulness. For example, effective range may be limited by
tradeoffs in transducer design to minimize water resistance and drag. A sum-
mary and examples of several FLS systems, presently available on the market
and each utilizing a different type of transducer, and their specifications are
given in Table 11.1.
11.2.3 Approach for Estimating Advance FLS Warnings

As a very first step for establishing and further development of definitive
parameters for the issuance of warnings to watchstanders using FLS, certain
guiding principles may be cited to achieve an estimate useful for broad guid-
ance on the subject. For example, a calculation of the relative time afforded
a vessel to begin corrective action such as altering course, reducing speed, or
otherwise reacting to an alarm generated by an FLS would be based, at min-
imum, upon the following factors:
newgenrtpdf
Table 11.1 Comparison between Forward-L ooking Sonar (FLS) Systems
Specifications FarSounder Argos 1000 a Furuno CH-2 70 b EchoPilot c
Maximum Detection Depth 85 m (278 ft) 100 m (328 ft) 100 m (328 ft)
Maximum Detection Range (MDR) 1,000 m (3,200 ft) 800 m (2,500 ft) 200 m (656 ft)
Operating Frequency 61 kHz 180 kHz 200 kHz
Vertical Coverage 60 o 180 o 90 o
Horizontal Field of View 60 o, 90 o, or 120 o 360 o 60 o
Maximum Transmit Power <1,500 Wrms 800 Wrms Not specified
Angular Accuracy 1.6 o Not specified Not specified
Roll/P itch Compensation ±20 o Not specified Not specified
Roll/P itch Accuracy 0.5 o Not specified Not specified
Maximum Vessel Speed 25 kts. 10 kts. Not specified
Screen Refresh Rate 0.3–2 s 8s 1–2 s
Advance Warning at MDR @ 10 kts. a 183 s (3.1 min) 139 s (2.3 min) 28 s (0.5 min)
Next-Generation Sensing
Sources: EchoPilot 3D Forward Looking Sonar; www.echopil ot.com/u ser/i mage/f ls3d-b rochu re.pdf; Furuno CH-2 70 Searchlight Sonar; www.fur
uno.com/e n/b usine ss_p rodu ct/p df/m ari ne/c h270.pdf; FarSounder 1000 Navigation Sonar; www.fars ound er.com/f iles/f 31566_3 d-s onar-b roc hure _
3.0.pdf
Notes:
Calculations by author.
a
United States Patent 8,717,847; Blake, 6 May 2014.
b
IMO Resolution MSC.232(82): 2006, Adoption of the Revised Performance Standards for Electronic Chart Display and Information Systems
(ECDIS).
c
IMO Resolution A.1021(26), Code on Alerts and Indicators, 2009.
209
MDR ( m ) - SRR (sec ) - APT (sec ) -WRT (sec )

AdvanceWarning (sec ) =
 m
Vs 
 sec 
where logical arguments may be made to establish values for each of these
parameters:
Speed of Vessel (Vs) =10 knots
Screen Refresh Rate (SRR) =2 sec, or as noted
Maximum Detection Range (MDR) =obtained from Table 11.1
Alarm Processing Time (APT) =4 sec
Watchstander Response Time (WRT) =5 sec
1 m/s =1.9438 knots
Although some systems can perform at much higher speeds, a value of 10
knots was selected for the Speed of Vessel (Vs) to provide a common basis
for evaluating the reaction time to FLS system alarms. A value of two seconds
for Screen Refresh Rate (SRR) on the display was assumed based on system
performance specifications. However, SRR for the Furuno CH-270 system
may be as high as eight seconds or more as this system provides general
coverage that extends beyond the area directly ahead of the bow. The value
for Maximum Detection Range (MDR) is obtained from the performance
specifications of the individual FLS units.
Alarm Processing Time (APT) is the speed at which FLS data can be eval-
uated by signal processing and alarm generation algorithms to determine
whether a condition exists that breaches predetermined vessel-specific safety
contours and depths. This includes draft, course, maneuvering capabilities,
lateral clearance margins and other factors pertinent to safety of navigation.
An assumption is made that such criteria for FLS are likely to be similar to
those established by IMO for ECDIS.27 Integral to this factor are time delays
to prevent normal operating conditions from causing false alerts because of
normal transients that may exist in the FLS data.28 This can add an additional
one to several seconds to ensure target persistence, eliminating the gener-
ation of an alarm due to a single occurrence or short duration or transient
target. A value of four seconds was selected based on FLS refresh rates as well
as estimates of processing times for interface and communications systems
handling software that may be required by an Integrated Navigation System.
Watchstander Response Time (WRT) is that needed for the Officer of
the Watch (OOW) to acknowledge an alert and take appropriate corrective
action based on the nature of the alarm. This must take into account time
lags necessary to assess rates of change in processes, such as changing ves-
sel course against targets’ movements.29 The time required for the OOW to
confer with other bridge watchstanders and lookouts and to issue orders to
the helmsman must also be factored into this calculation. Add to this the
effects of fatigue and various other human factor elements, one can easily see
that this factor is the most subjective and imprecise in the equation. A value
of five seconds was chosen in part based on the author’s direct observations
of bridge practices used on several vessels under similar conditions using
ECDIS and Radar indications.
Using these criteria, the EchoPilot 3D FLS provided approximately ½-
minute warning, the Furuno CH-270 provided approximately 2¼-minute
warning and the FarSounder 1000 provided approximately a 3-minute warn-
ing when considering their maximum detection range. The speed of 10 knots
may appear somewhat slow for most vessels while underway. However,
if a vessel is operating in unknown, poorly charted, or known-hazardous
waters it is prudent to increase safety margins by proceeding at a slower-
than-normal pace.
It should be noted that such advance warning calculations generally pro-
vide “best case” scenarios under ideal conditions and that actual conditions
and response times must be expected to reduce these margins –significantly
in some cases. Actual conditions must also take into consideration both
human and technological factors that can result in major deviations from
these response times. Technological factors can include water turbidity, poor
acoustic reflection qualities of potential HtoN and even growth on the hull
that may reduce FLS sensitivity. Human factors can range widely from dis-
tractions on the bridge, unfamiliarity with the equipment and general lack of
training, proficiency or currency in watchkeeping procedures.
11.2.4 FLS Installations

To achieve the most reliable and accurate performance FLS transducer(s) must
be located on an area of the hull that is free of turbulence from obstructions
located forward of the mounting position. Each transducer must be provided
an unobstructed view both horizontally towards the bow and vertically from
the waterline to the bottom to achieve the best accuracy and effectiveness.
The EchoPilot 3D Forward Looking Sonar uses two transducers mounted
athwartships, one on either side of the bottom equidistant to the keel –ideally
on the rear third of the vessel. The interior installation of these transducers,
as visualized from within the hull, is illustrated in Figure 11.1 (left).
The FarSounder 1000 Navigation Sonar uses one transducer mounted on
the bow with a fairing tube inserted within an existing or standard bulb, or
as a separate installation. The exterior installation of this transducer as seen
from ahead of the bow, is shown in Figure 11.1 (center and left).
The Furuno Searchlight Sonar, Model CH-270 (not pictured) uses a single
transducer that can be lowered from and retracted into the hull at speeds
above 10 knots. This unit may be mounted anywhere on the vessel where an
unobstructed view is available horizontally through 360°.
Figure 11.1 FLS Transducer Installations.

Note: Dual EchoPilot transducers mounted athwartships (left) and single FarSounder bow-
mounted transducers (center and right).
11.2.5 Integrated Bridge Alerts

The information available from the FLS can be useful in alerting watchstand-
ers as to potential HtoN present in the path directly ahead of the bow. Visual
indications seen on the FLS display are one form of alerting mechanism.
However, this data may also be shared as part of an integrated approach
using ECDIS as well as the Alert management, module C of an INS accord-
ing to latest IMO Performance Standards. This, moreover, coincides with
existing STCW training requirements and could utilize existing alarm mecha-
nisms with which watchstanders are already familiar.
The capability to provide those warnings and alarms can be even inde-
pendent of operator interpretation and be based upon analysis of the infor-
mation obtained from the FLS. This appears to be available only from the
FarSounder system. Taking into account the characteristics of the sea area
(e.g., approaching coastal waters, navigating in ice, etc.), the operator can
configure parameters within the area being observed by setting alarm values
by depth, minimum range, maximum range and field-of-view angle width as
illustrated in Figure 11.2.30 The alarm volume is shown in red. An additional
setting is the number of hits detected before triggering the alarm.
Figure 11.3 shows an example of FLS information integrated as an ECIDS
overlay. For this prototype Human Machine Interface (HMI) of integrated
FLS information, the three major requirements valid for display of Radar
information in ECDIS are applied with respect to the new technology.
Taking the potentials of FLS technology and results from first experimental
implementations it is concluded that watchstanders are provided both visual
and audible notifications as announced via the centralized Central Alert
Management (CAM)- HMI and supported by presentation in FLS target
overlay in an ECDIS as suggested in the Figure 11.3.
Furthermore, analysis of information available from FLS can be used to
trigger alerts with different levels of priority. In accordance with the defi-
nitions provided by IMO performance standards the lowest level of such
Figure 11.2 F orward-L ooking Sonar Alarm Volume Setting. (image: FarSounder,
Inc.).
Figure 11.3 FLS Electronic Navigation Chart (ENC) Overlay. (Image: FarSounder,

Inc.).
an alert is a caution to just raise awareness of the bridge team to a certain

unusual situation. In respect to the integrated use of FLS information this
could be a situation, where an obstacle is detected at a larger distance ahead,
e.g., due to a cross-track error.
Triggering a warning, the second priority level of alert and requiring imme-
diate attention by the bridge team could be linked to criteria within the usual
maneuvering range of the ship to avoid contact to any detected HtoN. The
time frame is to be configured as to allow the OOW to additionally check
the FLS display but also match available ECDIS information (e.g., to proof
approaching shallow water, a wreck or rock, island).
Finally, an alarm could be triggered characterizing conditions requiring
immediate action when, for example, the ship is approaching the lower
maneuvering limit (e.g., an emergency, hard rudder action would be required)
to avoid contact with an FLS-detected object/HtoN. This last stage of an alert
could be connected to the actual ship status and the ship’s maneuvering cap-
abilities. That information can be provided by dynamic predictions of the
ship’s future path for an emergency evasive maneuvers and check of the pre-
dicted track, for instance, against the available water depth.
11.2.6 FLS Information and Warnings

For the purposes of visualization and discussion of FLS-based warnings , a
simulation of the circumstances encountered during the grounding of Costa
Concordia in 2012 is used as a hypothetical case. Data are taken from the offi-
cial accident investigation report31 and simulation is attempted through extrac-
tion of the ground track using the AIS data record, then overlaying FarSounder
1000 FLS cone coverage onto the ground track to identify key events that
may have provided opportunities to enhance situational awareness. Four note-
worthy positions in the transit just prior to grounding identified as A, B, C and
D are illustrated in Figure 11.4 and recorded in the below chronology.
Ground Track Position A (10°56′32″ E, 42°20′41″ N; Speed:

15.3 knots; Course 325°)
This position lies approximately 1,600 meters from Scole Rocks at a depth in
excess of 100 meters and a mud bottom.32 This location is outside the range
of FLS to detect bottom information and is not likely to have provided any
information of significance to navigation.
Segment A to B (Elapsed Time: 00:01:08, Rate of Yaw: 11°/m in)

During most of this segment, the bottom depth would have remained in
excess of 100 meters with a mud bottom. However, just prior to arriving at
Position B, the maximum range of the FLS would have crossed the 100-meter
Figure 11.4 Ground Track of Costa Concordia, Annotated with Coverage for FLS
Navigation.
Source: Marine Casualties Investigative Body, Cruise Ship Costa Concordia, Report on the
safety technical investigation, p. 61, Ministry of Infrastructures and Transports (Italy).
bottom depth contour directly ahead and to the port of the centerline. It is
possible that early indications of an upslope bottom may have been able to
be detected. Note that during this segment, the rate of turn is approximately
11 degrees per minute.
Ground Track Position B (10°56′15″E, 42°20′52″N; Speed 15.6

knots; Course 335°)
This position lies approximately 1,100 meters from Scole Rocks at a depth
in excess of 100 meters with a mud bottom. The 100-meter bottom contour
is approximately 500 meters further along the course, with upslope bottom
rising to the 10-meter bottom contour at the maximum range of the FLS off
the starboard bow. The bottom is also transitioning from mud to rock, with
a resultant increase in acoustic reflectivity of the bottom material.
Clear indications of the approaches to Scole Rocks would have appeared
on the starboard bow on the FLS display. This probably would have initially
alerted the Master to the existence of a problem since the appearance of Scole
Rocks would have been expected off the port bow based on his passage plan-
ning.33 Furthermore, the depth contour displayed on the FLS would clearly
indicate insufficient depth in an unexpected location.
Segment B to C (Elapsed Time: 00:01:04, Rate of Yaw: 10°/m in)

During this segment bottom depth is steadily decreasing across the range of
the FLS from in excess of 100 meters to the surface. The bottom consistency
also changes from mud to solid rock, resulting in a large acoustic reflection
and indications of strong targets.
The appearance of a solid wall leading up to the surface and continuously
decreasing in distance would have loomed prominently on the FLS until it
consumed two-thirds of the display from far port to starboard of center.
Despite orders from the master to turn from 330° to 350° throughout the
1-minute segment duration, the rate of turn appears to have been half that at
10 degrees per minute.
Ground Track Position C (10°56′01″E, 42°21′05″N; Speed 16.0

knots; Course 340°?)
This position lies approximately 530 meters from Scole Rocks with a depth
of 95 meters just past the 100-meter contour with a rock bottom. The bot-
tom contour would reflect an upslope bottom rising to the surface around
500 meters ahead. Less than one-third of the display off the starboard bow
would now show clear water. There would be very little room to maneuver
at this point, but it was still possible to avoid disaster.
Segment C to D (Elapsed Time: 00:01:03)

This is the terminal segment of the voyage. Bottom consistency is solid rock
resulting in a large acoustic reflection and indications of strong targets.
The appearance of a solid wall leading up to the surface and continuously
decreasing in distance would have continued up to the point of impact.
Ground Track Position D (10°55′40″E, 42°21′21″N; Speed 15.3

knots; Course???)
This position lies approximately 30 meters from Scole Rocks with twelve
seconds remaining before contact.
Hypothetic Course of Events – Applying FLS

Using a speed of 16 knots obtained from the above chronology as the value
for Speed of Vessel (Vs) in equation (1), FLS could have provided approxi-
mately 1.8 minutes advance warning to the Master of Costa Concordia of
the pending HtoN consisting of the approaches to Scole Rocks. The warning
would have been issued when the vessel was just past position B, as shown
in Figure 11.4.
Had FLS been installed and operational on the bridge as discrete navi-
gational equipment without the benefit of inclusion within the INS, visual
indications would have shown an unmistakable wall directly in the path
of the vessel. Existing alarms available both integral and external to the
equipment would have also been activated, further reinforcing the severity
of the situation. Consistently, at all noted positions on the final approach
(Figure 11.4: positions A through D), the FLS would have indicated clear
water was present off the starboard bow.
It is assumed that the 1.8 minutes prior warning would have provided
sufficient advance notice to plan and execute evasive maneuvers that may
have lessened the severity of the grounding or averted it entirely. Halting
the forward momentum of the vessel would not have been possible as this
would require around 1,300 meters, with the vessel moving at 16 knots, and
this distance was not available.34 However, slowing the vessel combined with
executing a hard turn to starboard upon receiving the warning commen-
cing approximately 800 meters prior to Scole Rocks would have significantly
reduced the damage incurred in the event of grounding such that the vessel
may have remained afloat and lives may have been saved. Indeed, and even
hypothetical, the accident may not have happened at all.
11.2.7 FLS Potential to Avert Disaster

Vessel groundings and collisions with HtoN can and do cause untold suffer-
ing, loss of life and property, and often result in devastating environmental
damage. Since the Costa Concordia accident in 2012, many ships have
continued to run aground primarily under circumstances where situational
awareness was lacking by watchstanders of conditions surrounding the vessel
below the waterline, oftentimes as their vessels were making good yet devi-
ating from the planned track. The names of these ships are many, and a par-
tial list includes: Coastal Isle (2012), Tundra (2012), USS Guardian (2013),
Ovit (2013), Maersk Garonne (2015), Hamburg (2015), Nova Cura (2016),
Roger Blough (2016), Molly Manx (2016), USS Antietam (2017), Kea Trader
(2017), Universal Durban (2017), Marbella (2017), Celtica Hav (2018)
Priscilla (2018), Leda Maersk (2018), Wakashio (2020), Kaami (2020), Ever
Given (2020), Beaumaiden (2021), USS Connecticut (2021), Ever Forward
(2022) and Grande Senegal (2023).
Costs incurred as a result of groundings must include property and liabil-
ities arising from the accident itself as well as the cleanup. Such costs escalate
as vessels increase in size and cargo capacity, especially for cargos that
include toxic chemicals. Larger passenger vessels with greater carrying cap-
acity create even greater risk. The effects of such accidents are amplified in
areas that are remote where search and rescue and salvage efforts are prob-
lematic at best.
FLS can provide a means to directly detect bottom contours threatening
to rise up to meet the hull, HtoN that are not present on navigation charts
and can supplement existing methods to effect greater urgency to adverse
circumstances and heighten situational awareness. The technical capabilities
exist to make a substantial difference now. The establishment of carriage
requirements as well as fusion into integrated navigation systems still needs

to take place. The costs involved are minimal, especially in light of the con-
sequences of not having FLS available for navigational use. The salvage cost
of Costa Concordia alone amounted to over $2 billion US.35 This amount by
itself would cover the cost to purchase and install FLS equipment on 75%
of the world’s merchant fleet greater than 25,000 gross tons.36 These same
comments are just as applicable today as they were in 2014. Very little has
changed.
11.2.8 Acknowledgements
Some of the results and parts of the investigations presented were performed
under the European Interreg IVb- project –ACCSEAS Accessibility for
Shipping, Efficiency Advantages and Sustainability. Portions of this paper
were originally published by Michael Baldauf and R. Glenn Wright at the
10th International Symposium on Integrated Ship’s Information Systems
(ISIS 2014), German Inst. of Navigation, Deutsche Gesellschaft für Ortung
und Navigation e. V. (DGON), Hamburg, Germany, 5 September 2014.
11.3 eCelestial Navigation System (eCNS) 2

This paragraph describes a modern implementation of our greatest legacy
method by using the sun, moon and stars to guide ship navigation at sea,
and across inland bodies of water. It can also supplement pilotage techniques
when physical aids to navigation are not visible or are unavailable, and
when satellite methods such as Global Navigation Satellite System (GNSS)
are stymied by degradation, denial of service attacks and spoofing events.
Technology has progressed to the point where the development of an auto-
mated electronic Celestial Navigation System (eCNS) suitable for use by ves-
sels of all types is entirely feasible. Conventionally manned, unmanned and
autonomous vehicles can benefit from an independent source of positioning
information for primary navigation and as a reference for Inertial Navigation
System (INS) alignment.
The approach taken in original and ongoing research performed by the
author uses high resolution, commercially available passive sensors respon-
sive to visible, far-red and near-infrared light capable of imaging celestial
objects, including the sun, moon and stars as well as manmade satellites
during both daylight and nighttime conditions. Multi-megapixel cameras
combined with machine learning and deep-learning artificial intelligence (AI)
techniques make possible the continuous measurement of large numbers of
sky object altitudes approaching one-arcsecond accuracy that can provide
positioning to 31 meters at the Earth’s surface. These sensors reside on a
gyro-stabilized platform supplemented with software camera stabilization to
provide isolation from vessel movement in moderate seas –all contained
within a ruggedized waterproof housing topside approximately 0.75 cubic

meters in volume and weighing less than 38 pounds.
11.3.1 eCNS Technical Specifications

A wide field of view of 360° horizontally and 180° vertically is possible with
precision measurement maintained between 15° and 75° from the horizon
ensuring system resilience across a wide range of weather conditions that
may partially obscure the sky. Sensor power is provided and sensor data
routed to a below-deck controller consisting of a power supply, computer
and computer interfaces requiring a small amount of space (3U, or less) in
a computer rack. Celestial object and satellite information are contained
within databases that facilitate precise spatial-domain identification of indi-
vidual objects as well as the making of multi-star, Sun and Sun-Moon fixes,
and positioning based upon the orbital paths of observed satellites. An artifi-
cial intelligence-based automated process utilizing data fusion from multiple
sensors establishes an artificial horizon and assists in sky object identification
during obscured meteorological conditions and guides continuous observa-
tions and positioning updates at intervals of every few seconds throughout
a voyage. This makes possible the identification of spatial and time-domain
relationships to discern satellite behavior from those of aircraft, birds and
other environmental features. Further, the design of the system readily detects
and identifies eCNS degradation due to natural phenomena and distinguishes
it from nefarious influences effects that may result from third-party actions.
A capability also exists to capture and forward for remote human review one
or more small snippets of sensor data via a low bandwidth communication
channel as may be useful for off-vessel performance monitoring.
Design models, as well as prototypes of critical hardware components
provide the means to assess system trades that include size, weight, power
consumption, cost and projected system performance. In addition, proto-
type eCNS hardware mounted on an autonomous maritime vessel testbed
research vessel evaluates system utility, installation options and limitations,
and initial performance capabilities while underway.
11.3.2 eCNS Problem Identification

An investigation was undertaken into three distinct aspects of the celes-
tial navigation problem. The first relates to the identification and specifi-
cation of sensors and the design of an overall system architecture through
which the eCNS perceives the physical environment external to the vessel
as well as internal vessel capabilities to fulfill navigation requirements. This
included how sensor data and the physical and environmental characteristics
they represent are understood and the fusion of these data through which
a comprehensive near-real-time (within fractions of a second) assessment
of positioning awareness was created. Commercial-off-the-shelf hardware

is used to create the complete eCNS. Mounted topside at a location with
an unobstructed view of the sky in all quadrants is the ruggedized water-
proof housing with a transparent dome attached to a stabilized platform.
Within the housing are installed independent high-resolution visible light and
infrared cameras with wide-angle optical lenses suitable for full sky coverage
for each camera. Also contained within the housing are cooling and heating
apparatus sufficient to prevent condensation from occurring on the inside of
the housing and on the transparent dome. The platform is stabilized in two
axes to compensate for ship’s motion in terms of roll and pitch. Additional
hardware is associated with the computer that forms the heart of eCNS com-
putational capabilities and the stabilization controller that isolates the eCNS
camera/optics housing from ship’s movements.
Second, methods and processes used in celestial navigation are examined
across many individual situations and operational scenarios considering rea-
soning with uncertainty and potential sensor degradation. A key element to
ensure success was the use of existing commercial-off-the-shelf products to
the greatest extent possible and not have to re-invent already proven tech-
nology in developing the eCNS architecture. The line of research dictated
the use of numerous algorithms to perform celestial navigation in the mari-
time domain in terms of maritime piloting functions to provide data inputs
through an automated process. The same issue applies to satellite databases,
where reusing existing software provides significant cost and time savings
and reduces overall program risk. The software chosen for this application
contains details on more than 4,800 satellites, including over 4,000 Low
Earth Orbit (LEO) satellites currently orbiting Earth and provides orbit
information including apogee, perigee, inclination, period and other oper-
ational details.
The capability to acquire and process visible and far red/near-infrared
camera imagery to derive position information is an essential aspect of this
research. This is being accomplished using software functions to sample
video images at a standard rate, detect prominent image contents (points of
light) likely to represent navigational bodies (Sun, Moon, stars and satellites)
and store these as data objects for subsequent use in a two-stage effort for
object identification. Histograms of sky body images assist in the identifica-
tion of individual observed object spectral contents through comparison with
known objects. Supplementing this software are machine learning algorithms
that identify persistent objects (i.e., not satellites) based on their position in
the sky along with their proximity to other objects. The results of these two
steps generate a probability of identification as a particular object, which
is then assessed to either reinforce this conclusion or to choose another,
more likely object identification. Such methods are essential under partially
obscured skies when only portions of constellations may be visible. Another
aspect of this effort is the use of machine learning techniques to establish an

artificial horizon during obscured conditions through training with known
horizon characteristics under unobscured conditions for the eCNS installa-
tion. This is an essential element in establishing a pixel-width precision base-
line by which a pixel-by-pixel correlation may be made between the position
of the camera’s sensor element and the position of an object in the sky that is
essential to achieve highly accurate object altitude measurements that is far
beyond the capability of human observers.
Another aspect of this development effort is to again use machine learn-
ing to detect and differentiate between relatively persistent objects, such as
stars and planets and moving objects in the sky that include satellites and
aircraft. This is accomplished through frame-by-frame comparison to reveal
moving objects, and histogram analysis of moving object images that detect
aircraft strobe and navigation light characteristics that are not present in
satellites. A third aspect includes detection of the distinct and unique light
characteristics of anomalies and behavior that may represent sensor degrad-
ation due to natural conditions and those resulting from the nefarious acts
of adversaries.37
11.3.3 Present Development Status

Ongoing efforts continue with refinement of AI-based software and conven-
tional algorithms that implement eCNS user interface features. Also included
is further development of interfaces with celestial almanac and satellite data-
bases, provisioning of INS positioning and timing data, interacting with
other onboard systems as appropriate, and system anomaly handling as may
be needed. Future presentation of case studies and validated test results is
anticipated.
11.4 Virtual Aids to Navigation (VAtoN)

This paragraph describes new technology for implementing truly Virtual
Aids to Navigation (VAtoN) that have no need to rely on the presence of
VHF radio physical infrastructure within line of sight to establish the pres-
ence of navigation aids through Automatic Identification System (AIS) tech-
nology. Rather, VAtoN exist on the navigation chart at designated positions
defined by waypoints that are georeferenced (or anchored, so to speak) to
local bottom topography. They require no maintenance other than periodic
hydrographic surveys to ensure the bottom depths and contours on which
their positions are based remain accurate so that a proper watch may be
maintained –a prudent task regardless of whether physical or virtual aids to
navigation are deployed.
11.4.1 VAtoN Need and Application

Earlier in Chapter 3, Navigation, there was mention of the use of satellite
navigation where in practice almost universal reliance upon this single source
of PNT data is unwisely made. GNSS sensor data has been integrated and is
shared throughout all bridge systems and across sensor data bus interconnec-
tions throughout the vessel. However, the susceptibility of GNSS to spoof-
ing and denial of service attacks and the subtle degradation in positioning
accuracy and reliability that may ensue may catch watchstanders unaware
and lead to loss of situational awareness and even disaster. Attempts to miti-
gate and overcome these frailties are many and include the development of
enhanced encryption, anti-spoofing techniques and other methods to over-
come denial-of-service attacks. However, adversaries who are intent on mali-
cious behavior continue to make advances just as fast as countermeasures
can be developed.
11.4.2 VAtoN Development

A means was developed to maintain situational awareness sufficient to safely
navigate vessels without the need to rely on GNSS technology whereby crew-
members and the vessels on which they serve may again be self-reliant and
not at the mercy of external entities for the real-time information they need
to navigate and operate. This is envisioned as involving active navigation
systems such as Radar, AIS, Sonar and physical AtoN; as well as a passive
navigation system that facilitates precise positioning by adapting traditional
methods for pilotage to present day and evolving environmental sensing cap-
abilities that exist within the vessel itself. This approach also takes advantage
of high-resolution hydrographic vector data products provided by unmanned
survey vehicles, crowd-sourced data produced by participating surface ves-
sels, and by already available assets established by national hydrographic
agencies and other available civilian resources. The underlying concept is
to create a navigation system that considers a wide range of environmental
and vessel factors to update watchstanders INS guidance information. The
approach acknowledges the obtaining and fusion of data from multiple,
independent onboard sensor systems that include:
• Traditional single beam echosounder to determine real-time depth and

bottom contour information for comparison with legacy charted depths
contained within ECDIS to determine probable location.
• Side-scan Sonar for seabed feature detection and identification as physical
aids to navigation for pilotage on the sea bed and to precisely identify
known positions.
• Electronic Celestial Navigation System (eCNS) for day and night deter-
mination of position fixes from astronomical objects, including stars, the
sun, moon, planets and satellites.
• Visual piloting using visible light and infrared cameras to detect and iden-
tify physical aids to navigation present at the sea surface, including buoys,
lights and ranges plus landmarks and terrain features.
• Virtual AtoN (VAtoN) georeferenced to the seabed environment as way-
points along unmarked routes.
11.4.3 Sensor Utilization

The basic single-beam echosounders that exist onboard vessels by design and
regulation through mandated carriage requirements provide data needed to
help determine the depth of water that exists between the bottom of the
vessel and the seabed and is vital to ensure it does not run aground. Sailors
have developed the means to navigate using bottom contours and, aided by
computer automation, this capability has become much more accurate and
is limited only by the resolution and accuracy of hydrographic surveys and
their results codified in modern navigation charts. The use of numerical data
searches of ECDIS database contents to detect patterns within vector naviga-
tion chart depth soundings corresponding with continuous live Sonar meas-
urements of depth, change of depth and rate of change of depth patterns can
greatly aid in identifying actual present vessel position.
Supplementing the single-beam echosounder with side-scan Sonar helps
detect, recognize and positively identify seabed features such as significant
natural and/or manmade landmarks along with bottom consistency (e.g.,
rock, sand and mud). All such physical features and landmarks are them-
selves georeferenced to the physical environment enabling them to function
as AtoN, and waypoints that are also georeferenced to the physical environ-
ment enabling them to function as Virtual AtoN; both of which are verifi-
able as watching properly by having the proper characteristics and being in
the correct position. This approach extends traditional pilotage that takes
advantage of hills and bluffs, water and transmission towers, river mouths
and other natural and manmade landscape features to determine location;
to extend below the waterline where known boulders, reefs, fissures, wrecks
and possibly manmade aids to navigation disguised as natural features can
mark precise locations capable of being used as a position reference for
inertial navigation. The task of AtoN detection, recognition and positive
identification is one that is perfectly suited for machine learning and has
been demonstrated by the author and others as being practical. Similar to
the echosounder, side-scan Sonar (available as a built-in feature on many
low-cost commercial echosounders and fish finders) can assist when position
error crosses a certain threshold and when the vessel is believed to be in rela-
tively close proximity to a known undersea feature.
Additional considerations include the effects of tides, winds, sea state and
other external factors affecting a voyage; combined with their effects on
performance-based on vessel motion in terms of heave, sway, surge, roll,
pitch and yaw. Also considered can be internal sensors and external meteoro-
logical and oceanographic (METOC) sensor data content and availability to
derive error correction vectors that compensate for progressive error exhib-
ited by INS resulting from a lack of positional references through GNSS or
known physical location. This capability utilizes software developed through
machine-learning techniques to model natural forces and tailor their effects
on performance to the unique characteristics of a specific vessel. Deep-
learning AI can also be used to discern unknown factors, trends and events
present in the data that may also influence vessel performance.
The combination of such techniques can provide a means for a navigation
system to constantly learn new things about the environment and the vessel
itself and to improve its own operational capabilities on an ongoing basis. All
such methods and capabilities listed above are possible today using existing
computing and AI technology. Each method is subject to its own limitations
and accuracy characteristics. However, the fusion of multiple sources of com-
plementary data can reinforce conclusions reached regarding positioning and
achieve accuracies greater than those of individual sources.
An autonomous vehicle would not have the constraints of human senses
and attributes but, without a human in the loop, would lack the knowledge
and decision-making necessary to properly execute safe passage. Machine
learning and optical character recognition can be used to train an automaton
to recognize basic shapes and colors of standard buoys and other AtoN, and
to read identifying numbers and letters that would enable positive identifica-
tion of individual navigation aids. An automated process is used to correlate
with ECDIS and the Light List along with performing cross-checking with
Radar and GNSS indications to verify that the detected aid is watching prop-
erly in the correct position and exhibiting proper characteristics. However,
less than ideal orientation of buoy numbers and letters, fouling by birds, low
light conditions and other natural and man-made phenomena can drastically
obscure, alter physical and light characteristics and reduce the certainty of
what is being seen. Other charted AtoN, such as lighthouses, piers, water
towers, building and landscape features are much less standardized, but a
mariner would routinely detect and identify can be much more challenging
or outright impossible for automated processes.
11.4.4 VAtoN Experiments

Figure 11.5 illustrates an experiment performed to detect and identify AtoN
to assist in the automated navigation of a vessel across a setting that includes
different aspects of open sea, harbor and inland navigation.38 This was accom-
plished entirely without the benefit of GNSS. The methodology incorporates
the use of machine learning, reinforcement learning and deep learning AI
techniques to create image, data and information analytics for multiple sen-
sor modalities for a cohesive approach to Maritime Autonomous Surface
newgenrtpdf
Figure 11.5 Aid to Navigation (AtoN) Detection and Identification Experiment Route.
225
Ship (MASS) navigation. However, issues associated with the use of AI in

safety-critical MASS applications cannot be examined in uncertain, unstruc-
tured and dynamic environments which lack mathematical frameworks to
verify the correctness of their solutions and validate the appropriateness of
their actions.
The location of this experiment was the Chesapeake Bay and Severn River
near Annapolis, Maryland in the United States, where vessel traffic includes
large oceangoing ships, tugs, barges and many small recreational boats.
Charted landmarks abound that are useful as AtoN, including buildings,
domes and natural terrain, including 27 buoys and fixed aids, 2 bridges and
4 buildings selected for this experiment. The transit route was approximately
11 nautical miles in length with an average depth of 31 meters (east) and 5
meters (west).
A ten-meter research vessel shown in Figure 10.1 configured as a maritime
autonomous system test bed was used to perform this experiment. Sensor
data acquisition (e.g., compass, Radar, AIS, echosounder) was accomplished
in real time during the transit on the navigation computer with subsequent
fusion during post-processing. Daylight, low light and infrared imaging data
along with Radar and FLS data were captured using video data recorder and
NMEA-2000 bus, with subsequent fusion achieved during post-processing.
Automated AtoN detection and identification were performed live and
recorded on the navigation computer during the experiment. Post-processing
correlated and verified all sensor data with electronic chart tracking infor-
mation and GNSS positioning acquired independent of the experiment as a
secondary check. For safety reasons, vessel navigation was performed uncou-
pled from the navigation system, but with visual indications provided at all
times to the helm and followed throughout the voyage.
Updated results from this experiment were reported at the International
Conference on Maritime Autonomous Surface Ships (ICMASS) in Trondheim,
Norway, in November 2019.39 Of the 27 marine AtoN (buoys) present along
the route, all were detected and 18 were positively identified. In addition, two
bridges and four buildings along the route were detected and positively iden-
tified. Nine buoys were detected but not identified due to distances being too
great and beyond sensor resolution, poor buoy orientation with identifying
characteristics positioned away from camera view, and fouling preventing
optical recognition of buoy identification features.
Conclusions taken from this experiment indicate that smart sensors com-
bined with well-trained software can positively identify manmade critical
AtoN and provide adequate situational awareness in an operational environ-
ment under defined circumstances for onboard computers. However, remote
operators may be hampered by bandwidth and data latency issues. Further
conclusions indicated that top-level (Captain) decision-making capabilities
must further mature before it can be determined whether overall system
resilience in managing distributed smart sensors can be assured.
Pilotage navigation of a vessel through restricted waters has traditionally

involved the use of terrain features, landmarks and other AtoN such as buoys
visible to watchstanders marking proper channels and hazards to navigation.
Awareness of the depths of the waters below the keel to ensure safety of navi-
gation is also an essential part of pilotage first achieved in antiquity using the
leaded line, wire drags and subsequently through the use of the single-beam
echosounder beginning in the 1930’s.40 In 1974, under SOLAS, echosounders
were specified under IMO mandatory carriage requirements for larger vessels,
and they are now standard equipment found on just about all vessel types
and sizes larger than a rowboat. The relatively recent introduction over the
last 50 years of multibeam and side-scan Sonar systems has greatly advanced
hydrographic survey practices and vastly improved the accuracy and reso-
lution of soundings data on navigation charts. These inventions, along with
FLS, now enable individual vessels to obtain highly accurate visibility into
the seabed configuration not only below the keel but ahead of the vessel’s
path to instantly warn of shoals, reefs and other hazards to navigation and
allow sufficient time to take evasive action prior to running aground.
One method of piloting using an echosounder is the use of depth contour
following to determine when the vessel enters deeper or shallower water. This
is accomplished using waypoints to establish desired entry and/or exit points
along a depth contour that may be used to roughly establish position when
anticipating crossing a bank or upon approaching a seacoast featuring grad-
ually shallowing depths. Another method is to follow a contour at a specific
depth until reaching a prominent landmark or AtoN that may lead to the
destination.
Enhanced Sonar capabilities, to directly assess bottom configurations
and conditions in real time, provide new and redundant precision naviga-
tion and vision capabilities to provide terrain tracking capabilities and high-
resolution imaging of prominent seabed features in the same way as surface
landmarks to aid navigation in ways never before possible.41 Such capabilities
are essential for Unmanned Surface Vehicle (USV) navigation without GNSS
but can also apply to conventional vessels in the future. Navigation Sonar
with forward-looking capabilities can provide high-resolution bathymetry
that may be compared to electronic navigation charts (ENCs) displayed on
ECDIS for backup navigation, detection of hazards to navigation and obsta-
cles, and avoidance of large marine mammals.42 These data are sufficient to
crowdsource navigation chart bathymetry.43
Figure 11.6 illustrates the combined use of visual AtoN detection and iden-
tification with navigation by depth contour tracking. Originally developed
for Arctic navigation in poorly charted and uncharted regions, the principles
involve the correlation of three aspects of bathymetry to determine position-
ing: the depth soundings themselves, changes in depth soundings along the
planned course, and the rate of change of depth soundings as they occur.44
newgenrtpdf
228
Figure 11.6 AtoN Detection and Identification Combined with Depth Contour Following.
This approach formed the basis to implement truly Virtual AtoN (VAtoN)
that do not require a corresponding VHF radio transmitter but are georefer-
enced to the physical seabed environment as waypoints. Course and heading
information developed by the navigation computer using multi-sensor data
fusion from cameras, Radar, Sonar and other sensors are provided directly to
the helmsman and autopilot to execute the voyage plan.
The findings of this experiment note the significant contribution of depth
contour following and the use of VAtoN depicted as red circles to enhance
adherence to the planned route over the initial portion of the experiment
course where physical AtoN in the form of buoys are sparse with great dis-
tances between buoys that are essentially beyond camera range. Further along
the course shown in Figure 11.6 where the channel was well marked with
physical AtoN, the use of depth contour following using VAtoN was gener-
ally effective in maintaining a proper course within the channel. Note that
the illustrations provided are shown using raster chart depictions for clarity.
This experiment was performed using vector charts and supplemented with
live soundings obtained using the FarSounder ARGOS 350 forward-looking
navigation Sonar (FLS) to provide more accurate data in much greater detail
than is available using vector charts.
11.5 Quantum Sensing

Welcome to the weird, wonderful world of the quantum domain, where seem-
ingly the laws of physics that govern the behavior of familiar objects give
way to the principles of quantum mechanics that describe interactions at the
atomic and subatomic levels. I’ll go into this in more detail a little later. But
first, you might wonder what quantum sensing has to do with ship sensors.
Quantum sensors can measure time, dynamics (i.e., forces, acceleration and
rotation) and fields (i.e., gravitational, electromagnetic and mechanical) with
much greater precision and stability than any presently available technology.
These properties have boundless potential for improving existing ship sen-
sors and sensor types, and inventing previously unimaginable sensing tech-
nologies that can provide new solutions to current problems and difficulties.
11.5.1 Quantum Ship Sensor Applications

There are many different areas on ships and within ship functions where
quantum sensing can make significant improvements in existing and future
sensing applications. Examples include:
• Ship Navigation:45 A quantum sensor with potential applications in

GNSS-free navigation, developed at Imperial College London, has been
tested in collaboration with the Royal Navy. This sensor represents a
new type of accelerometer that displays quantum properties at ultracold
temperatures. By combining changes in an object’s velocity over time with

rotation measurements and the initial position of the object, the current
location of a ship can be calculated with great precision.
• Temperature Sensing:46 Quantum dots, which are three- dimensional
semiconductor nanoparticles ranging in size from 2 to 10 nm, feature
fine-tunable optical characteristics and can be highly sensitive to envir-
onmental conditions such as temperature. These specific properties make
them useful for a wide range of applications, including optical and opto-
electronic devices for temperature sensing for machinery and machinery
spaces, cargo and environmental monitoring.
• Imaging:47 Quantum sensors can be combined with imaging devices to
perform new types of ultra-high resolution imaging capabilities that far
surpass those of present-day camera and video, Radar, Sonar and Lidar
technologies. This can include detecting and building 3D models of stealth
or “invisible” objects by sensing gravitational changes in the local area.
These devices can include a probe that scans the position of the sensor,
an array of sensors, or a beam of electromagnetic waves prepared in a
quantum state to perform high-resolution macroscopic (i.e., large-scale)
imaging with unprecedented sensitivity. Also possible are new features
that can see around corners and through walls without any adverse effects
on materials or living creatures.
• Non-Destructive and Non-invasive Testing:48 Combining groups of atoms
and a web of lasers, scientists can monitor the fluctuations in individual
atoms within this web. This can culminate in large amounts of accurate
data that can be used to detect metal fatigue, failures and pending failures,
and other changes within rotating machinery such as propeller shafts,
mounts for deck and hull equipment, and other mechanical and electro-
mechanical devices.
• Quantum Twin Duplication:49 The quantum property of entanglement
can be incorporated into sensors where two particles within the apparatus
would be entangled or have interdependent quantum states. This can have
the effect of being able to duplicate events taking place at one location for
viewing or taking action at another, such as between a ship and a shore
location. Entanglement can occur theoretically without regard to distance.
• Sensor Security:50 Quantum metrology leverages quantum entanglement
to surpass existing physical limitations and achieve great measurement
precision. Quantum states also have a non-deterministic nature leading to
uncertainties in their characterization and can be employed in quantum
cryptography protocols to provide information and data security.
These are just a few of the many areas of research that can be applied dir-
ectly to the maritime environment and, specifically, to ships and unmanned
vehicles or all types and kinds.
11.5.2 How Quantum Sensing Works

Quantum sensors use the principles of quantum mechanics to measure phys-
ical properties at the atomic and subatomic levels, which behave quite dif-
ferently than the normal physical objects we deal with every day. In the
quantum domain something can be either one thing or another at the same
time –until you look at it. Then it becomes one thing. Many principles of
quantum mechanics achieve capabilities that do not necessarily make sense
using conventional physics. Let me explain.
The “quantum bit” (qubit) is very much like the “bit” of conventional
computing that stores information as either a 1 or 0. This is where the simi-
larity ends since a qubit can also be in a combination (or superposition) of
both states or somewhere in between at the same time. However, when you
observe the quantum bit then it becomes a 1 or 0. This is sort of like the
cat of someone named Schrödinger who put it in a sealed box (metaphoric-
ally, of course) where it could either be dead or alive (or both?) at the same
time.51 However, when you peek inside the box, then you know one way or
the other. Now that you understand this basic principle you have graduated
to the level of quantum engineer (third class) and are now ready for the
next level.
• Sensors using quantum mechanics can be linked together (or entangled) in

such a way that they share the same fate. This allows quantum sensors to
achieve even higher sensitivity and to operate independently yet produce
the same results regardless of distance.
• Quantum sensors can observe phenomena by tunneling particles with a
wave function defined over all space that can go through barriers impos-
sible for classical objects to pass through. This allows quantum sensors
to measure very small changes in electric fields and distinguish between
physical properties with great precision.
• The ability to reduce unwanted noise caused by photons popping in and
out of existence in empty space is referred to as quantum squeezing.
This occurs by reducing uncertainty in one property at the expense of
increasing the uncertainty in another property, allowing quantum sensors
to achieve even higher precision.
These are just a few of the principles of quantum mechanics that are used
in quantum sensors. As quantum sensing technology continues to develop, it
is likely that even more ways will be discovered to use quantum mechanics
and to achieve even higher sensitivity and precision in sensors.
Congratulations! Understanding these basic principles advances you
to quantum engineer (second class). However, this is as far as you can go
because the rest of the curriculum has yet to be invented.
11.5.3 Future Directions of Quantum Sensing

Quantum technology is expected to change and improve secure communi-
cation and information processing, improve the sensitivity of detectors and
sensors, provide new functional materials for future electronic devices and
change the way we approach computing.52 In theory, quantum sensors are
able to achieve much higher sensitivity and precision than conventional sen-
sors, making them well-suited for a wide range of applications including navi-
gation, process and environmental monitoring, and even space exploration.
There are many areas in quantum mechanics that are yet to be explained.
Albert Einstein called quantum entanglement “spooky action at a distance”,
where two particles potentially separated by light-years great distances but
connected through a mysterious and inexplicable link. At present, there are
about 30 different models still trying to explain it.
Notes
1 In this paragraph (11.2) INS is used for Integrated Navigation System, NOT
Inertial Navigation System.
2 In this paragraph (11.3) INS is used for Inertial Navigation System, NOT
Integrated Navigation System.
References
1 Drone market outlook in 2023: Industry growth trends, market stats and fore-
cast. Insider Intelligence. 7 January 2023. www.insiderintelligence.com/insig
hts/drone-industry-analysis-market-trends-growth-forecasts/
2 Matthew Spaniol. (2020). Drones on Ships. Periscope. 10.13140/
RG.2.2.18715.28969. www.researchgate.net/publication/347521511_Drones_
on_ships
3 Aishwarya Lakshmi. Maersk to Use Drones to Resupply Its Fleet of Tankers. 6
June 2016. MarineLink. www.marinelink.com/news/resupply-tankers-UAS410
695.aspx
4 Gary. How Drones Are Changing the Maritime Industry. 18 May 2016. Ship
Technology. www.ship-technology.com/features/featurehow-drones-are-chang
ing-the-maritime-industry-4865807/
5 Wilhelmsen and Airbus Trial Drone Deliveries to Singapore Anchorage. 16
March 2019. The Maritime Executive. https://maritime-executive.com/article/
wilhelmsen-and-airbus-trial-drone-deliveries-to-singapore-anchorage
6 Carrier USS Ford Tests Out Air Cargo Drone. 8 August 2021. The
Maritime Executive. www.maritime-executive.com/article/video-carrier-
uss-ford-tests-out-air-cargo-drone
7 www.cruisejunkie.com/Overboard.html
8 Marine Accident Recommendations and Statistics, 2015– 2022. Marine
Accident Investigation Branch (MAIB), United Kingdom.
9 A.S. Selmy, (2016). The Need of Man Overboard (MOB) Detecting and
Tracking System Descriptive Analyses. Conference: The 9th China International
Rescue & Salvage Conference, China.
10 Orhan Gönel and Ismail Çiçek. ÖNEL Statistical Analysis of Man Overboard
(MOB) Incidents. December 2020. www.researchgate.net/publication/
348266442_STATISTICAL_ANALYSIS_OF_MAN_OVER_BOARD_MOB_
INCIDENTS
11 Martin Manaranche. Royal Navy Tests Drones in Man Overboard Trials.
Naval News, 06 July 2021. www.navalnews.com/naval-news/2021/07/royal-
navy-tests-drones-in-man-overboard-trials/
12 Donato Cafarelli, et al. MOBDrone: A Drone Video Dataset for Man OverBoard
Rescue. 15 March 2022. 22st International Conference on Image Analysis
and Processing. ICIAP 2021.arXiv:2203.07973. https://doi.org/10.48550/
arXiv.2203.0797
13 Mike Ball. Antarctic Supply Ship Uses Drone to Navigate Sea Ice. Unmanned
Systems Technology , 23 December 2015. www.unmannedsystemstechnology.
com/2015/12/antarctic-supply-ship-uses-drone-to-navigate-sea-ice/
14 Royal Dutch Shell Scope of Work Narrative. Anchor Retrieval Project
(Location: Beaufort and Chukchi Seas). Fairweather, LLC. Anchorage, Alaska.
USA. 1 August 2023.
15 Drone Guided Ship through Arctic Sea Ice. 16 November 2016. iHLS. https://
i-hls.com/archives/72848
16 Matt McFarland. Ship relies on drone to avoid ice blocks in Arctic waters.
CNN Business. 5 November 2016. https://money.cnn.com/2016/11/05/technol
ogy/arctic-drone-ship-navigate/index.html
17 Drone Survey of ONE Apus Container Collapse. 16 December 2020. The
Maritime Executive. www.maritime-executive.com/article/video-survey-of-
one-apus-container-collapse-made-public-by-wk-webster
18 Gary. How drones are changing the maritime industry. 18 May 2016. Ship
Technology. www.ship-technology.com/features/featurehow-drones-are-chang
ing-the-maritime-industry-4865807/
19 International Regulations for Preventing Collisions at Sea 1972 (COLREGS),
Rule 5: Look-out.
20 Standards of Training, Certification and Watchkeeping (STCW) Convention
and Code, Chapter VIII, section A-VIII/2, part 3 –Watchkeeping at Sea.
21 Samuel Halpern, An Objective Forensic Analysis of the Collision Between
Stockholm and Andrea Doria, www.titanicology.com/AndreaDoria/Stockh
olm-Andrea_Doria_Collision_Analysis.pdf
22 Resolution A.918(22), IMO Standard Marine Communication Phrases, p. 67,
B1/1.3.
23 United States Patent 5,390,152; Boucher, et al.; 14 February 1995.
24 United States Patent 5,675,552; Hicks, et al.; 7 October 1997.
25 United States Patent 7,173,879; Zimmerman, et al.; 6 February 2007.
26 United States Patent 8,717,847; Blake, 6 May 2014.
27 IMO Resolution MSC.232(82): 2006, Adoption of the Revised Performance
Standards for Electronic Chart Display and Information Systems (ECDIS).
28 IMO Resolution A.1021(26), Code on Alerts and Indicators, 2009.
29 Guidance Notes on Ergonomic Design of Navigation Bridges, October 2003,

American Bureau of Shipping,
30 Operation Manual, FarSounder-1000 Navigation Sonar, F31552 (Rev. 2.9.0)
2014, FarSounder, Inc.
31 Marine Casualties Investigative Body, Cruise Ship Costa Concordia, Marine
casualty on 13 January 2012, Report on the safety technical investigation,
pp. 3, 13, 14; Ministry of Infrastructures and Transports (Italy).
32 Isle del Giglio, www.cruisersforum.com/forums/f134/chart-accuracy-and-det
ail-comparison-74805.html
casualty on 13 January 2012, Preliminary Report, Slide 14, 18 May 2012, pre-
sented at MSC90, London.
casualty on 13 January 2012, Report on the safety technical investigation,
p. 50, para. 2; Ministry of Infrastructures and Transports (Italy).
35 Telegraph 2014, Costa Concordia salvage: the £1.2bn price tag of the disaster;
The Telegraph, 14 July 2014, London, www.telegraph.co.uk/news/worldnews/
europe/italy/10966308/Costa-Concordia-salvage-the-1.2bn-price-tag-of-the-
disaster.html
36 Equasis Statistics; The World Merchant Fleet in 2011, Table 1, p. 6.
37 GMATEK, Inc. In-Stride Detection of Sensor Degradation. Final Report. 28
April 2022. Contract N68335-20-G-2005. Principal Investigator: R. Glenn
Wright, Ph.D.
38 R. Glenn Wright, (2019). Intelligent autonomous ship navigation using multi-
sensor modalities. International Journal on Marine Navigation and Safety of
Sea Transportation, 13(3). September 2019. DOI: 10.12716/1001.13.03.03
www.transnav.eu
39 R. Glenn Wright. Enhanced MASS situational awareness using virtual naviga-
tion aids. International Conference on Maritime Autonomous Surface Ships
(ICMASS) in Trondheim, Norway in November 2019.
40 NOAA. History of Hydrographic Surveying. Office of Coast Survey. National
Oceanographic and Atmospheric Administration. www.nauticalcharts.noaa.
gov/learn/history-of-hydrographic-surveying.html
41 Wright R. Glenn and Baldauf M. (2016). Correlation of Virtual Aids to
Navigation to the Physical Environment. TransNav, The International Journal
on Marine Navigation and Safety of Sea Transportation, 10(2), 287– 299.
DOI: 10.12716/1001.10.02.11
42 R. Glenn Wright and Ian Russell (2017). Navigation Sonar Use in Maritime
Frontier Exploration. Soundings. Hydrographic Society of the UK.
Summer 2017.
43 FarSounder 2018. FarSounder Joins NOAA as a Trusted Node. Press Release.
17 October 2018. www.farsounder.com/about/press_releases
44 R. Glenn Wright (2017). 4-dimensional Virtual Aids to Navigation for Arctic
Maritime Transit. IEEE/ MTS Oceans Conference. Anchorage, AK. Session
AK.10. Marine Safety and Security with Diminishing Ice. September 2017.
45 Hayley Dunning. Quantum Sensor for a Future Navigation System Tested

Aboard Royal Navy Ship. TechXplore. 26 May 2023. https://techxplore.com/
news/2023-05-quantum-sensor-future-aboard-royal.html
46 Natalia Doskaliuk, Yuliana Lukan, and Yuriy Khalavka. Quantum Dots for
Temperature Sensing. April 2023. DOI: 10.58332/scirad2023v2i2a01. www.
researchgate.net/publication/370142487_Quantum_dots_for _temperature_
sensing
47 Quantum Technology: Sensing and Imaging. Australian Army Research Centre.
2 September 2020. https://researchcentre.army.gov.au/library/land-power-
forum/quantum-technology-sensing-and-imaging
48 Chen-Wei Liang, Ehsan Ahmad, Ertan Balaban, James Sexton, and Mohamed
Missous. A Quantum Well Hall Effect Magnetovision System for Non-
Destructive Testing. 19th World Conference on Non-Destructive Testing 2016.
www.ndt.net/article/wcndt2016/papers/mo1f1.pdf
49 L. Pezzè. Entanglement-enhanced sensor networks. Nature Photonics 15, 74–
76 (2021). https://doi.org/10.1038/s41566-020-00755-x
50 M.T. Rahim, A. Khan, U. Khalid. et al. (2023) Quantum secure metrology for
network sensing-based applications. Science Report 13, 11630. https://doi.org/
10.1038/s41598-023-38802-6
51 John Gribben. In Search of Schrödinger’s Cat. Bantam Books. 1984. ISBN 978-
0-553-34253-6 www.randomhousebooks.com
52 Tamara Dietrich. Curious about Quantum? 16 February 2023 https://new.nsf.
gov/science-matters/curious-about-quantum
Index
ABS 67, 161, 162, 167 Aurora Australis 204

acceleration 12–14, 21, 24, 46, 55, 56, automatic identification system see AIS
66, 79, 229 automatic picture transmission see APT
accelerometer bias 46 automatic radar plotting aid see ARPA
active sensors 88 autonomous vehicle 29, 37, 41, 42, 102,
actuators 11, 157 117, 118, 120, 123, 168, 169, 218
AI 5, 29, 30, 41, 60, 74, 77, 79, 83–6, autopilot 28, 31, 140, 158, 229
88–100 auxiliary systems 14
aids to navigation see AtoN
Airbus 202, 232 background lights 117
air flow 12 Baltic and International Maritime
air pressure 55 Council see BIMCO
AIS 2, 6, 29, 32, 33, 38, 41, 43–5, 49, bandwidth 23, 26, 35, 102, 103, 131,
51, 80, 82, 103, 105, 107, 125, 126, 133, 135, 138, 170, 194, 219, 226
150, 157, 158, 165, 178, 190, 192, Baron WxWorx 36
193, 198, 214, 221, 222, 226 bauxite 57, 68
AIS-ATON 45 Bayometric 77, 81
Aiviq 205 Beaufort Wind Force Scale 197
alarm 23, 26, 30, 75, 79, 80, 188, 208, Beaumaiden 217
210, 212–14 BeiDou 34
alarm processing time 210 bell 29, 101, 122, 124, 125, 193
allision 29, 39, 170, 172, 174, 206 BIMCO 161, 167
American Bureau of Shipping see ABS biofouling 173, 180
ammonia 15 black box testing 92, 177
Andrea Doria 37, 207, 233 Bluetooth 6, 24, 78, 129, 134–7, 141–3,
anemometer 16, 32, 47 153, 158, 160
Annapolis, Maryland 48, 226 broadband 3, 30, 101, 102, 104, 158,
APT 6, 36 170, 194
ARPA 29–32, 37, 158 Broadcast Notice to Mariners 108
artificial intelligence see AI bubble screens 183, 185
Assured Autonomy 177, 199 Bulk Jupiter 57, 68
atmospheric ducting 109 Bureau Veritas 64
atomic force microscopy 4, 9
AtoN 29, 36, 39–45, 51, 73, 117, 118, calcium carbonate 58
127, 222–9 cameras 7, 22, 23, 42, 43, 54, 61, 63,
attack surface 155, 156, 166 73, 74, 78, 79, 81, 85, 87, 88, 94,
audio 12, 31, 32, 41, 42, 122, 124, 151, 104, 117, 122, 124, 138, 146, 151–3,
193, 197 155, 157–9, 166, 171, 175, 178, 180,
Index 237
193, 197, 198, 204, 205, 218, 220, converters 23, 26

223, 229 convolutional neural network see CNN
CAN 24, 129, 134–40, 143, 152, 158 coolant 16–18
capacitance 19, 20, 25, 86 COSPAS-SARSAT 80, 115
CAPTCHA 93 Costa Concordia 97, 206, 214–18, 234
carbon monoxide 14 COTS 176
car carriers 52, 56 counterfeiting 72, 75
cargo control systems 154 crewmember credentials 71
C-band 6, 103 crowd control 129
CCNI Arauco 56 CSC 55, 67
CCTV 73, 79, 81 CSSC Cape Town 58, 68
celestial navigation 47, 219, 220 Curtis Bay Coal Terminal 58
Celtica Hav 217 cyberattack 141, 144–6, 148, 149,
Central Alert Management 212 152–5, 169
centrifuge 145 cybersecurity 2, 9, 75, 92, 145–7, 154,
certification of cargoes 57 155, 162–4
ChatGPT 84, 93, 98 cylinder liners 12, 157
chemical obscurants 180, 181
chemicals 53, 146, 174, 217 DARPA 177, 199
chip log 28, 46, 49 data rate 131–3, 135, 137, 140, 142
chronometer 46, 47, 51 dayshapes 101, 115, 116
class A AIS 44 DCSA 161, 162, 167
class B AIS 44 dead reckoning 28, 46; see also DR
classification 8, 89, 103, 170, 196, 202 Decca 43, 44
closed circuit television see CCTV decision trees 89
the Cloud 3, 74 deep learning 29, 60, 74, 77, 83, 86,
cloud services 160 90–3, 170, 171, 173, 175–7, 195,
clustering 89 218, 224
CNN 90, 99 Defense Advanced Research Projects
coal 53, 58 Agency see DARPA
coal dust 58 denial of service see DOS
Coast Pilots 28, 42, 96 Department of Homeland Security 81,
Coastal Isle 217 106, 108, 126, 127, 164
COLREGS 29, 32, 40, 49, 50, 96, DGPS 35, 44, 49, 188
115–17, 122, 127, 128, 200, 233 differential GPS see DGPS
combustibles 52 Digital Container Shipping Association
commercial off the shelf see COTS see DCSA
communications-denied Digital Select Calling see DSC
environments 171 Doppler radar 88
compaction 57 Doppler shift 37
condensation 19, 124, 220 DOS 35, 47
condensers 11, 157 DR 45–7
conductance 12, 19 drone 150, 181, 202–5, 232, 233
containers 45, 52–6, 64, 65, 85, 153, DSC 30, 32, 80, 105, 110, 114, 115,
155, 158, 160, 207 126, 158
controller area network see CAN dust 21, 22, 58, 62, 79
controllers 23, 35, 146, 147, 157, 160,
171, 198 ECDIS 28, 30–2, 41–3, 45, 149, 150,
Convention on Facilitation of 157, 162, 193, 198, 203, 207,
International Maritime Traffic 58 209–12, 214, 222–4, 227, 233
Convention on the International eCelestial Navigation System see eCNS
Regulations for Preventing Collisions EchoPilot 179, 180, 183–5, 209,
at Sea see COLREGS 211, 212
238 Index
echosounder 172, 178–80, 184, 198, Felicity Ace 63

222, 223, 226, 227 fiber optic 17, 23, 63, 102, 135,
eCNS 218–22 137, 138
electromagnetic spectrum 5, 197 filtering 23, 25, 38, 47, 87
electronic celestial navigation system fingerprint readers 78, 152
see eCNS fire detection 21, 22, 27, 59, 63
Electronic Chart Display and FIRESAFE 64, 69
Information System see ECDIS Firewall 133, 146, 156, 159
electronic control unit see ECU firmware 41, 147, 152, 157–60
Electronic Industries Alliance see EIA 5G 10, 36, 50, 103, 146, 158
electronic navigation chart see ENC flame 22, 121
Elera 103 flammability 16
eLORAN 30 flammable liquids 52
ELVA-1 mmRadar 39, 50 flow moisture point 57, 58
Emergency Position-Indicating FLS 30, 88, 180, 206–18, 226, 227, 229
Radiobeacon see EPIRB fluorspar 58
e-Navigation 24, 49 fog 1, 2, 30, 35, 47, 121, 169, 174, 180,
ENC 31, 190, 193, 198, 213 181, 196
encryption 76, 133, 151, 164, 222 forward-looking savigation sonar
enhanced Long-Range Radio Navigation see FLS
see eLORAN fouling rating 180, 199
Enhanced Maritime Mobile Service 49 frequency domain 87, 88, 170, 196, 197
entangled 230, 231 frequency modulation 37
EPIRB 6, 31, 43, 49, 115 friction 16, 57
Ethernet 129, 134–8, 142, 171, 176, Furuno 178, 182, 183, 188, 209–11
179, 199
Eurocargo Trieste 63, 64, 70 Galileo 34, 115
Euroferry Olympia 63 game of chess 88, 95
European Coastal Monitoring System GAN 90
see ECMS Garmin 178, 182–4, 188
European Maritime Safety Agency gasoline 61–3
see EMSA gas turbine power control system 148
European Maritime Safety Information generative adversarial network see GAN
System see EMSIS GEO 23, 33, 35, 51, 65, 67, 78, 102–4,
European Space Agency 205 114, 115, 119, 127, 182, 188, 198,
evaporator 11, 157 199, 207, 221, 223, 229
Ever Alot 53 Geostationary Earth Orbit see GEO
Ever Forward 217 Global Maritime Distress Safety System
Ever Given 217 see GMDSS
exhaust gas 12, 157 Global Navigation Satellite System see
explosion 16, 58–62, 68, 69, 151 GLONASS
explosive gases 14, 55 Global Navigation Satellite Systems
explosive vapor 61, 153 see GNSS
Extremely High Frequency see EHF Global Positioning System see GPS
eye scanning 76, 78 Global Xpress 103
GMDSS 4, 6, 30–2, 49, 102, 105, 114,
facial recognition 71, 74–6, 78, 81, 146, 115, 127
152, 158, 166 GNSS 6, 28–35, 42–4, 47, 49–51, 53,
Fairweather, LLC 205, 233 55, 56, 80, 105, 110, 113, 132, 140,
FarSounder 98, 179, 183, 185, 186, 145, 150, 153, 157, 172, 174, 175,
209, 211, 212, 214, 229, 234 178, 188, 189, 200, 207, 218, 222,
Fatigue 14, 17, 27, 61, 69, 211, 230 224, 226, 229
Index 239
Golden Ray 63 IACS 161, 162, 167

gong 122, 124, 125 IALA 51, 117, 118, 127
Google 51, 93, 95, 98 IALA A 117
GPS 32, 34, 35, 39, 50, 51, 115, IALA B 117
150, 157, 174, 175, 178, 188–90, IAPH 144, 161, 162, 167
204 IBS 130, 142
Grande America 63, 69 IC 75
Grande Costa d’Avoria 62 icing 174
Grande Europa 63 ID cards 72, 73, 75, 76
Grande Senegal 217 IEC 60945 137
grounding 57, 170, 172, 174, 206, 207, IEC 61158 137
214, 217 IEC 61162 13, 24, 137, 140, 158
gunfire 101, 122 IEC 11623 158
gyroscope 46, 56, 79, 80, 170, 188 IEEE 802.11 136, 137
gyroscope drift 46 imbalance 20, 137
IMO 4, 21, 24, 27, 29, 31, 32, 49, 55,
hall effect 20, 235 58, 63, 67, 68, 96, 114, 115, 128,
Hamburg 217 130, 161, 167, 201, 203, 207–10,
hard fouling 180 212, 227, 233
hazardous cargo 54 IMSBC code 57
Hazard to Navigation see HtoN Inclinometer 56
heating ventilation and air conditioning Indian Regional Navigation Satellite
see HVAC System see IRNSS
HF 6, 32, 34, 43, 49, 76, 102, 105, inductive 20
109–11, 113–15, 122, 158, 194 inertial measurement unit 80
high atmospheric ionization 114 inertial navigation system see INS
high frequency see HF information and regulatory marks 118
Highland Chieftain 104 information technology see IT
high moisture content 57 infrared cameras 42, 63, 85, 87, 122,
high-pass filter 26 153, 158, 171, 197, 198, 204,
high seas forecasts 110 220, 223
HMI 212 infrared see IR
HMS Diamond 205 Inmarsat 6, 36, 49, 102, 103, 115, 122,
Höegh Xiamen 62, 69 125, 127, 171–5, 178, 188, 189,
horn 29, 101, 121–5, 193 206, 207, 212, 216, 218, 221, 222,
hospitality 1, 71, 83, 158 224, 232
HtoN 207, 211, 212, 214, 216, 217 INS 30, 31, 46, 47, 130, 132, 171
Hubble Space Telescope 3 insufficient righting arm 66
hull fouling 173, 180 integrated bridge system see IBS
human activity 5, 61, 72, 73, 79, 129 integration drift 46, 172
human error 27, 61, 66, 83 intelligent network edge devices 104
human machine interface see HMI intensity 12, 20, 76, 77, 80, 86, 121,
humidity 12, 13, 18, 19, 24, 32, 47, 52, 125, 153, 181
53, 55, 58, 86, 131, 158, 175, 192 interference 35, 38, 40, 86, 104, 105,
HVAC 11, 14, 157 131, 132, 134, 145, 152, 154, 171–4,
hydraulic 13, 15, 170, 174, 176 179, 180, 182, 187, 188
hydrogen 21, 61 internal friction 16
hydrographic survey 42, 198, 199, 221, International Association of
223, 227, 234 Classification Societies see IACS
hydrostatic 15 International Association of Marine
hygrometer 47 Aids to Navigation and Lighthouse
hysteresis 14 Authorities see IALA
240 Index
International Association of Ports and latency 30, 42, 103, 131, 133, 226
Harbors see IAPH laterite 57
International Code of Signals see ISC L-band 6, 34, 38, 39, 57, 103
International Convention for Safe leaked fuel 65
Containers see CSC leaked oil 65
International Convention for the Safety Leda Maersk 217
of Life at Sea see SOLAS LEO 102–4, 115, 126, 220
International Convention on Standards level sensing 12, 157
of Training, Certification and LF 6, 43, 44, 76
Watchkeeping for Seafarers see STCW Lidar 7, 28, 37, 39, 50, 87, 88, 94, 171,
International Electrotechnical 172, 174–6, 178, 185–8, 196, 197,
Commission see IEC 199, 230
International Maritime Organization life jackets 73
see IMO light flashes 123
International Maritime Solid Bulk Light Imaging Detection and Ranging
Cargoes see IMSBC see Lidar
International Organization of Masters, light rhythm 118
Mates and Pilots see IOMM&P light signals 17, 117, 121, 122
International Standards Organization liquefaction 52, 57, 85, 158
see ISO liquid natural gas see LNG
Internet 23, 30, 48, 56, 81, 84, 102, lithium-ion batteries 21, 52, 63, 64
103, 129, 134, 137, 138, 144, 159 load monitoring 12, 157
Internet of Things see IoT local area network see LAN
interoperability 92, 94, 133, 147, 148 log book 46
IoT 23, 56, 144, 151–3, 155, 159, long-range identification and tracking
160, 165 see LRIT
IP camera 138, 153 long-range radio navigation see LORAN
IR 7, 42 LORAN 29, 33
Iridium 36, 102, 104, 126 Loran-C 43, 44
iris scanners 152 loss of resolution 175
IRNSS 34 low bandwidth communication
iron ore fines 57 194, 219
ISC 118 low-bandwidth data links 170
ISO 55, 80, 82, 137–9, 143 low earth orbit see LEO
ISO 11898 139, 143 lower explosion limit see LEL
ISO/IEC 7498-1 139, 143 low frequency see LF
IT 94, 102, 146, 147, 151, 157–9, 161, low-light video cameras 178
162, 164 L, S, C, X, Ku and Ka-bands 103
lubricant oil 15
James Webb Space Telescope 3, 9
jamming 35, 174 machine learning 28, 41, 60, 74, 77, 79,
86, 89–93, 97, 171, 173, 175–7, 196,
Kaami 217 218, 220, 221, 223, 224
Kea Trader 217 macrofouling 180
Ku-band 6, 103 Maersk Garonne 217
Maersk Honam 58
LAN 6, 24, 129, 134, 135, 137, 138, Maersk Tankers 202
143, 150, 154, 158, 160 magnetic compass 46
large wave strikes 174 magnetic resonance imaging 3
laser strikes 175 magnetometer 56, 170
laser weapons 174 MAIB 59, 68, 203, 232
Lash Fire 63, 69 man overboard 34, 80, 202, 203
Index 241
Marbella 217 MSC Irina 53

Marine Accident Investigation Branch MSC Loreto 53
see MAIB MSI 36, 113
marine growth 172 MS Ore Brasil 53
marine safety information see MSI MT Pablo 60
Marine Transportation Security MTS 162, 163
Clearance see MTSC M/V Pacific Centurion Swire 202
Marine Transportation System 108 MV Tonsberg 53
Maritime Autonomous Surface Ship
see MASS Natanz, Iran 145
Maritime Institute of Technology and National Institute of Standards and
Graduate Studies (MITTAGS) 96 Technology see NIST
Maritime Mobile Service Identity National Marine Electronics Association
see MMSI see NMEA
Maritime Security (MARSEC) Card 72 National Science and Technology
Maritime Security Identification Card Council see NSTC
(MSIC) 72 National Transportation Safety Board
MASS 4, 99, 200, 226, 234 see NTSB
MatoN 45, 51 Nationwide Automatic Identification
Maximum Detection Range (MDR) System see NAIS
209, 210 natural gas 15, 21, 53
Mayflower 104, 126 NavalDome 144
Mayhew, Tracey 96, 99 navigational bodies 220
measurement bias 132 navigational text messages see NAVTEX
medium earth orbit see MEO navigation chart 42, 46, 96, 117, 137,
medium frequency see MF 145, 217, 221, 223, 227
MEO 33, 102, 104, 115 navigation divider 46
Methane 21, 61 NAVTEX 6, 30, 32, 110, 113, 115, 158
Metropolitan Area Network (MAN) neural networks 89, 197
102, 125 NIST 49, 75, 78, 81, 82, 127, 161, 163,
MF 6, 43, 44, 49, 109, 110, 113, 115 164, 167
microphone 22, 42, 74, 85, 122–5, 151, NMEA-0183 24, 135, 137, 139, 140,
178, 193, 197, 199 158, 188–90
Microsoft 95 NMEA-2000 24, 134, 135, 137, 140,
microwave 79, 102, 155 158, 179, 188, 190, 191, 226
middleware 147, 157–60 NMHS 30
millimeter radar see mmRadar noise 22, 25, 46, 85, 86, 89, 123, 131,
mineral concentrates 57 132, 170, 171, 175, 231
misalignment 174 Notices to Mariners 28, 42, 96, 108
mmRadar 5–7, 37, 39, 152 not making way 117, 124
MMSI 31, 32, 44, 105, 110 not under command 116, 124
MOBDrone benchmark 204 Nova Cura 217
mobile aid to navigation see MAtoN NTSB 64–6, 69, 70
modem 157, 159 NSTC 163, 167
moisture 12, 22, 57, 58
Molly Manx 217 object recognition 74, 94
mooring buoys 118 occupancy detection 71, 158
Morse code 43, 118, 121 Ocean Data Acquisition System 45
motion 12, 14, 20, 47, 56, 74, 78, 79, oil consumption 16
80, 220, 233 Omega 33, 43, 44
MSC Daniela 56 One Apus 205
MSC Flaminia 56 OneWeb 102
242 Index
operational technology see OT positioning, navigation and timing

operations centers 170 see PNT
optical 17, 20, 22, 58, 77, 87, 101, 169, position sensor 19, 20, 198
178, 220, 224, 226, 230 post-quantum cryptography 164
Orbcomm 102 potable, ballast, bilge and black
Orchestra 103 water 15
oscilloscope 176 powder 16
OSI model 138, 139, 142 precipitation 35, 38, 47, 169, 174, 180
OT 144, 146, 148, 151, 157, 159, precision time signals 113
161, 162 pressure sensor 17, 54, 79
Ouster OS2 Lidar 39 Priscilla 217
outliers 131 PRM 56, 62, 217
Ovit 217 prognostics 173
oxygen 14, 21, 55, 58 programmable logic array (PLA) 160
programmable logic controller
PAB 80 (PLC) 160
parallel ruler 46 progressive error 171, 173, 224
parametric rolling movement see PRM propulsion 11, 14, 27, 123, 148, 149,
parametric test 180 157, 170, 172–4, 176
particle pore spaces 37 protect surface 155, 156, 163
particle size distribution 37 proximity 20, 39, 40, 43, 153, 169, 188,
passenger credentials 71, 72, 147, 152 189, 207, 220, 223
passive infrared (PIR) 79 psychrometer 47
passive sensors 218 pulse modulation (PM) 13
pattern recognition 78, 89 pulse width 37, 86
PCS 154 pyrites 58
perimeter Sonar 153 pyrotechnic 121
peripherals 141, 157
personal Automatic Identification quantity 12, 87
System beacon see PAB quantum 164, 229–32
personal locator beacon see PLB Quasi-Zenith Satellite System see QZSS
personal protective equipment 73 QZSS 34, 39
phase shift 37
photodiode 58 Racon 41
photoelectric 20 radar-assisted collision 207
physical damage 170, 172, 174, 180 radar beacon see Racon
physical obstructions 151, 178, radial 41
180, 181 radioactive materials 52
physical ship infrastructure 157 radio beacon 43
P&I club Gard 58, 68 radio direction finding see RDF
piezoelectric 20 radiofacsimile 113
piezoresistive 20 radio frequency (RF) 10, 15, 155
Pillars of Creation 3, 9 radio frequency identification see RFID
piloting 36, 40–2, 73, 220, 223, 227 radiotelegraphy 118
pixilation 175 radiotelephony 119
PNT 32, 34, 43, 46, 171, 172, 174, RDF 6, 43
175, 222 red mud 57
Polar Code 201 reluctance 17
port control system see PCS Remontowa Shiprepair Yard,
port digital infrastructure 160 Gdańsk PL 205
Port Security Identification Card remote desktop protocol (RDP) 160
(PSIC) 72 remotely operated vehicle see ROV
Index 243
remote operating center (ROC) 104 Shinshu Maru 53

repeatability 84 ship digitalization 27
Rescue 21 108, 109, 126 shock 52, 55, 56
resistance 12, 15–20, 24–6, 86, 175, signal flags 119, 120
176, 208 signaling devices 121
resolution 4, 5, 28, 37–9, 42, 43, 76, sirens 193
77, 87, 88, 122, 133, 146, 161, 169, Sirius XM Marine Weather 36
175, 180, 184, 204, 208, 218, 220, SkyMate 36
222, 223, 226, 227, 230 smart camera 79, 81, 104, 124,
retina 76, 77 146, 204
RFID 55, 67, 73, 75, 76, 81 smart sensor 2, 4, 9, 11, 27, 60–3,
Rizika, Robert 14, 144 65, 66, 83–6, 91–5, 98, 148, 153,
Roger Blough 217 160, 226
Ro-Ro 5, 53, 54, 62–4, 66, 69, 158 smoke 21, 22, 55, 59, 63–5, 71, 79,
routers 138, 157, 159 101, 121, 122, 153, 158, 205
ROV 184 Smooth Sea 22 62
RS-485 129, 135–7, 140, 141, 143 snippets 193, 194, 197, 219
rudder position 11, 18 Society of Automotive Engineers (SAE)
runaway polymerisation 60 134, 142
soft fouling 180
Saab 102 software defined radio see SDR
safe water marks 118 SOLAS 4, 21, 27, 32, 96, 101, 105, 108,
safety of navigation 3, 9, 33, 38, 47, 114, 130, 142, 201, 227
118, 125, 149, 168, 171, 180, 201, sound signaling 118, 122
210, 227 SpaceX 102, 125
Safetytech Accelerator 63, 69, 166 spatial domain 87, 91, 170, 173, 176,
Sagar Nandini-2 61 177, 195–7, 219
Sailing Directions 42 special marks 118
S-AIS 30, 42, 103 spectrum analyzer 176
SAR 6, 37, 80, 115, 127 spontaneous combustion 21, 52, 58
SARSAT 80, 115, 127 spoofing 35, 47, 150, 152, 174, 175,
SART 6, 32, 39, 115 190, 218, 222
SateliotX 102 SP Technical Research Institute of
satellite AIS see S-AIS Sweden AB 64
Satellite-Based Augmentation System stability 14, 53, 57, 65, 66, 70, 132,
(SBAS) 115 148, 178, 229
S-band 6, 34, 38, 39, 103 stabilized platform 218, 229
Scole Rocks 207, 214–17 Starlink 102, 103, 125
SDR 176, 188 statistical 3, 74, 77, 89–92, 131,
Seafarers International Union 96 177, 233
search and rescue see SAR steering 11, 33, 117, 148, 149, 157,
search and rescue transmitter see SART 170, 172, 176
sea state 56, 79, 85, 170, 197, 204, Stena Rederi AB 64
223 Stockholm 36, 207, 233
secure protocols 133 Stolt Focus 60
self-test 180 Stolt Groenland 59, 68
sensitivity 14, 25, 46, 76, 88, 131, 193, Strain 12–14, 17, 20, 24, 25
211, 230–2 styrene monomer 60
sensor analytics 3, 9, 86, 173 supervised learning 89–91
sensor fusion 11, 28, 39, 47, 80, 92, 94, Svitzer Hermod 104
99, 197, 198 swath bathymetry 88
sensor noise 46 synthetic aperture radar 37, 50
244 Index
TCP/IP 138, 142 VAtoN 207, 221–4, 229

Telecommunications Industry velocity 12, 13, 16, 17, 20, 21, 34, 35,
Association (TIA) 141 46, 79, 152, 172, 188, 230
TEU 53, 56 verification 74, 77, 91–3, 161, 173, 177
thermal runaway 63, 64, 175 very-small-aperture terminal see VSAT
thermistor 19 vessel traffic monitoring and
thermocouple 18, 25 information system VTMIS
thermometer 47 vessel traffic services see VTS
tide 42, 48, 51, 108, 198, 223 vibration 4, 11–13, 20–2, 27, 52, 53,
time domain 87, 90, 91, 170, 173, 176, 131, 157
196, 197, 219 virtual aid to navigation see VAtoN
time-triggered Ethernet 171 visual signals 115
torque 12–14, 17, 157 voyage data recorder (VDR) 32,
traceability 92, 177 150, 165
training datasets 91–3 VSAT 103, 126, 166
transducer 3, 87, 88, 169, 172–5, 179, VTS 6, 44, 103, 150
180, 183, 185, 208, 211, 212
transfer function 14 Wakashio 217
transmission control protocol/iInternet Wallenius Wilhelmsen 63
protocol see TCP/IP WAN 101, 155, 158, 160
transportable moisture limit 57 waste oil 15
Transportation Worker Identification watchstanders 29, 30, 83, 96, 108, 116,
Card (TWIC) 72 120, 122, 123, 149, 169, 206, 208,
Tundra 217 210, 212, 217, 222, 227
twenty-foot equivalent unit see TEU wave conditions 56
wear 11, 20, 80, 157, 173
ultrasonic sensors 77 whistle 101, 122, 123, 193
UMAA 190, 200 wide area network see WAN
Universal Durban 217 Wi-Fi 78, 85, 102, 129, 134–7,
Universal Leader 53 142, 152
Unmanned Maritime Autonomy Wilhelmsen Ship Service 202
Architecture see UMAA winch torque 12, 157
unmanned surface vehicle see USV wind vane 47
unmanned vehicles 3, 9, 94, 101, wind velocity 152
173, 220 Windows operating system 149
unsupervised learning 89, 91, 92 Wogan, Ford 144
USS Antietam 217 World Meteorological Organization
USS Bonhomme Richard 59, 68 36
USS Connecticut 217 Wright, R. Glenn i, 49, 87, 98, 199,
USS Guardian 217 200, 218, 234
USV 4, 39, 89, 227
X-band 6, 32, 38, 39, 41, 103
vacuum 12–14, 17, 18, 37, 38
validation 91–3, 173, 177 zero trust 155, 156

Ship Sensors 2024

Uploaded by

Copyright:

Available Formats

Ship Sensors 2024

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ship Sensors 2024

Uploaded by

Copyright:

Available Formats

Ship Sensors

When combined with artificial intelligence, advanced computing architec-

R. Glenn Wright holds a BS in Electrical Engineering, MS in Computer

Conventional, Unmanned and

List of Figures xii

1 Introduction to Ship Sensors 1

4.4.2 Explosion and Fire 60

5 Crewmember and Passenger Sensors 71

6 Artificial Intelligence in Sensor Systems 83

7 Shore and Intership Communications 101

7.2.1 Very-​Small-​Aperture Terminal (VSAT)

8 Intraship and Internal Communications 129

8.2.3 Bandwidth 133

9 Cybersecurity and Ship Sensors 144

10 Vessel Sensor Degradation 168

10.4 Sensor Degradation Detection 175

11 Next-​Generation Sensing 201

11.2.4 FLS Installations 211

6.1 Radar Image in Pixels, with Associated Radar Chirp

10.17 Transmission of Snippets via Email Using High Frequency

1.1 Electromagnetic Spectrum Allocation to Maritime Sensors 6

Automaton1 (noun) au· tom· a· ton

1 A mechanism that is relatively self-​operating, especially a robot.

1 A sensor that conditions measured signals before transmitting them to the

My eternal love, thanks and appreciation are extended to Magdalena, whose

ABS American Bureau of Shipping

ENC Electronic Navigation Chart

MAtoN Mobile Aid to Navigation

List of Acronyms and Abbreviations xxi

SBAS Satellite-​Based Augmentation System

Introduction to Ship Sensors

Ship Without A Soul

1.1 Ship Sensors

1.2 What Exactly Is a Sensor?

environment.1 In practice, the terms “sensor” and “transducer” are often

1.3 The Sensor Revolution

computing, today makes possible non-​invasive visualization of the interior

1.4 Non-​t echnical Factors Driving Maritime Sensor

1.5 Electromagnetic Spectrum Allocation to Maritime

communicate great volumes of related information across large portions of

1.6 Types of Ships and Maritime Vehicles

• Passenger Ships • Dry Cargo

Introduction to Ship Sensors

These exemplify the vast majority of sensor applications found on modern

1.7 Sensors Discussed, and Not Discussed

1.8 Flow of the Chapters in This Book

Chapters 2–​10 focus on the present state of sensor technology, describe

6 David Sumi. V and E Bands Key to the 5G Revolution. July 2019.

Engineering sensors have evolved to support multiple roles and functions

2.1 Ship Function, Health and Performance

2.2 Types, Functions and Applications

• Pressure • Angle • Capacitance • Fluid • Open/​

Some sensors have the capability to measure multiple characteris-

Table 2.1 Example Engineering Sensor Functions a

Function Signal Descriptions

Alternating Voltage, Current, Power, Frequency, Average Frequency, Average L-​

applications and the characteristics measured in monitoring specific func-

2.3 Performance Parameters

• Transfer Function: A mathematical function that models sensor output

2.4 Engineering Sensors

List of Figures xii

1 Introduction to Ship Sensors 1

5 Crewmember and Passenger Sensors 71

6 Artificial Intelligence in Sensor Systems 83

7 Shore and Intership Communications 101

7.2.1 Very-Small-Aperture Terminal (VSAT)

8 Intraship and Internal Communications 129

9 Cybersecurity and Ship Sensors 144

10 Vessel Sensor Degradation 168

11 Next-Generation Sensing 201

1.1 Electromagnetic Spectrum Allocation to Maritime Sensors 6

1 A mechanism that is relatively self-operating, especially a robot.

SBAS Satellite-Based Augmentation System

computing, today makes possible non-invasive visualization of the interior

1.4 Non-t echnical Factors Driving Maritime Sensor

Chapters 2–10 focus on the present state of sensor technology, describe

• Pressure • Angle • Capacitance • Fluid • Open/

Table 2.1 Example Engineering Sensor Functions a

Alternating Voltage, Current, Power, Frequency, Average Frequency, Average L-

• Piezoresistive: Solid-state devices embedded within a sensor change their

• Electro-mechanical: Two dissimilar metals are physically bonded together,

• Synchro/Resolver: A synchro converts an angular position into an elec-

expressed in cycles-per-second or Hertz (Hz). The amplitude between vibra-

the operating bandwidth of the receiver, low-pass filters reduce or elim-

Table 3.1 Example Navigation Sensor Functions

• Global Positioning System (GPS) –United States.

Increasing bandwidth for ship satellite communications allows many real-

3.3.2.2 Radar Operating Frequency Trade-o ffs

3.3.2.3 Radar-L ike Devices

Direction-finding has traditionally been accomplished using low frequency