WHITE PAPER

Virtual Commissioning
with Model-Based Design
How to implement virtual commissioning in your development process
With virtual commissioning, you use a simulation model of your manufacturing or industrial auto-
mation plant to test and validate system changes and upgrades before implementing them on actual
equipment.
• The use of virtual commissioning is increasing due to:
• More sophisticated software used in production equipment
• The need to operate equipment continuously and globally
• A lack of testing capabilities to verify the functional behavior of the machine
• The desire to eliminate errors early in the design process
Each of these factors increases the overall cost for commissioning due to project delays and the poten-
tial damage to your business reputation.
Instead of testing complex software after parts of the machinery or the complete production line is
assembled, engineers use a simulation model of the machine—a virtual machine. With simulation,
the interaction between mechanics, machine software, and the product can be tested, optimized, and
verified in different scenarios, even if the physical machine is not yet available.

Does Virtual Commissioning Pay Off?

A virtual machine lets you run most of the time-consuming and often expensive tests first in simula-
tion, reducing the time it takes for “real” commissioning of the physical plant. In addition, you can
test earlier and in parallel with other activities such as assembly of the mechanical and electrical
parts.
The main objection to introducing virtual commissioning is the effort required to develop a model of
the machine with sufficient accuracy, as well as the corresponding know-how needed by development
teams. So it’s important for companies evaluating virtual commissioning to compare the effort with
the expected benefits:
• What is the return on investment (ROI)?
• How much does traditional commissioning (without virtual commissioning) cost?
Consider staff needed onsite, travel costs, waste of material and energy, and any risks of
penalties in case of delay.
• What does the investment in virtual commissioning look like?
Consider creating models, building knowledge on the team, and licensing simulation
software.
• What are the technical requirements for the model?
• Level of abstraction: Incorporate simple control signals, dynamics of mechanics, sensor
signals, etc.
• Level of visualization: Consider whether visualization is needed. If so, you may want to
create graphs of values over time or even 2D or 3D animations.
• Real-time capabilities: Does the model need to run in real time?
• Communication interfaces: Connect the software/PLC to the virtual machine.

2
Companies that successfully implement virtual commissioning have two things in common:
• Realistic expectations about the model, the required accuracy, and which requirements it
can fulfill.
• A phased introduction of new processes. Steps that provide the best ROI are implemented
first and deliver first results. Thus, initial investments pay off early and facilitate later phases.
However, a model of the machine enables more than virtual commissioning. It can also play a central
role in the entire development of the machine and the machine software. Furthermore, you can use
the model throughout the lifetime of a machine in the form of a digital twin, such as for monitoring
machinery or performing predictive maintenance. Taking this into account, the effort and investment
in modeling is justified more easily and pays off in multiple ways.

Model-Based Design: Beyond Classic Virtual Commissioning

Even if virtual commissioning is an effective way of addressing the challenges of real commissioning,
it is not the most efficient way of using models. In contrast to the machine builder and production
equipment industry, companies in other fields like automotive and aerospace are already using
models from the first day of development. They create models of the plant (mechanics, electrical
drives, hydraulics, etc.) as well as the control software. This approach enables closed-loop system sim-
ulation from the very beginning of the product life cycle and allows simultaneous development of
mechanics, electronics, and software. Once the software algorithms are completed in design, they are
tested against the plant model. The production code for implementation on a PLC or other automa-
tion hardware is automatically generated. This workflow—Model-Based Design—shortens develop-
ment time and increases the quality of the product.

RESEARCH REQUIREMENTS

DESIGN

Environment Models
TEST AND VERIFICATION

Physical Plant Models

Control/Supervisory Logic Models

IMPLEMENTATION

C, C++ IEC HDL

MCU DSP PLC FPGA ASIC

INTEGRATION

Development workflow with Model-Based Design. The approach spans the entire development process
including the commissioning phase.

3
Modeling and Desktop Simulation
The machine model and controls model are designed based on requirements, measured data, CAD
models, and legacy machine software. The models are refined and then combined into a complete
system model. Tools such as Simulink® not only allow you to create models for the machine or system
with prebuilt elements, but also support the design of control logic and algorithms, as well as real-
time code generation for PLCs.

Key Capabilities
• Design of controls model and plant models Fast Modeling Using CAD Import
of the physical system, including CAD
import CAD models provide a comprehensive mechani-
• Prototyping of new functionality in combi- cal description of a machine. With Simscape
nation with legacy machine software Multibody™, CAD assemblies, including masses,
inertias, joints, constraints, and 3D geometry, can
• Automated system tests be imported to obtain the machine model. After
• Parameter optimization (such as software, the import, the user defines which joints are
mechanics, hydraulics) active (actuator driven), passive, and/or sensed
(provide position feedback). Depending on the
• Automatic code generation from models
requirements, electrical drive trains with motors
(IEC 61131-3 Structured Text and Ladder
and inverters are added.
Diagram, C/C++, HDL)

Benefits
• Simulate mechanical, electrical, and control
systems in the same software environment,
enabling complete system design
optimization
• Use system-level simulation to evaluate and
optimize the dynamic interaction of the System dynamics visualized in 3D.
electrical, mechanical, and control systems
and its impact on the embedded software
• Find design errors early in the design process
• Perform costly and safety-critical tests in simulation
• Avoid manual coding errors by automatically generating code
• Reuse models in future projects

Closed-loop system model of a robot arm and its control logic in Simscape™ and Stateflow®.

4
 Hardware-in-the-Loop and Real-Time Tests
A virtual version of the machine is created through automatic code generation from the machine sim-
ulation model. This virtual machine runs on a real-time hardware—a hardware-in-the-loop simulator—
which is connected to the industrial controller over an industrial fieldbus for testing, refining, and
optimizing the machine software.

Key Capabilities
• Emulate the behavior of the physical system
Virtual Commissioning Under
(plant model) in real time
Real-Time Conditions
• Design and test hardware-independent
functionality To obtain feedback about the functionality
and performance of the machine control soft-
• Debug real-time algorithms directly from
ware on the final hardware, it is helpful to run
Simulink (using “External Mode”)
the model under real-time conditions to test
• Connect the virtual machine to the PLC over the algorithms and the fieldbus communica-
an industrial fieldbus tion to the PLC.

Simulink Real-Time in conjunction with a real-

Benefits time hardware system from Speedgoat
• Test and verify the functionality of machine enables real-time execution of the machine
software before the physical machine is model and supports common industrial field-
available bus protocols used in the automation industry.

• Debug the machine software in operation

directly from the model
• Find integration errors earlier in the design
process by using real-time simulation for test-
ing and verification
• Simulink Real-Time™ and real-time hardware
from Speedgoat
provide an optimized solution for hardware-
in-the-loop simulation. Speedgoat hardware
provides interfaces to common industrial fieldbus protocols:

Protocol I/O Modules

PROFIBUS® IO641 (master), IO642 (slave)
PROFINET® IO751 (controller), IO752 (device)
EtherCAT
®
Via GbE ports on all real-time target machines; IO710, IO712, and IO713 (master); IO750 (slave)
Modbus TCP® IO753 (client), IO754 (server)
Modbus RTU® External gateway
EtherNet/IP™ IO755 (scanner), IO756 (adapter)

POWERLINK IO758 (controlled node)

5
Code Generation for PLC Platforms
The machine software can be implemented on several system architectures by automatically generat-
ing the code from tested and validated models. Depending on the selected hardware platform, C/C++,
IEC 61131-3 (Structured Text or Ladder Diagram), or HDL code can be generated. Automatic code
generation is available for all common PLC and industrial PC platforms. MathWorks cooperates with
several hardware partners to ensure compatibility with their integrated development environments
(IDEs).
The support of different implementation languages and PLC platforms enables engineers to design
and test the machine software independently from the hardware platform. This approach is especially
helpful when multiple PLC platforms are used or when the final decision for the hardware has not yet
been made.

Connections
Vendor IDE IEC 61131-3 C/C++
Partner
3S - Smart Software
Solutions
CODESYS®
• •
ABB / B&R
Industrial
Automation Studio™
• • •
Automation
Bachmann
Electronic
SolutionCenter
• • •
Beckhoff
Automation
TwinCAT®
• • •
Bosch Rexroth IndraWorks
• • •
Mitsubishi Electric CW Workbench
• •
Ingeteam Ingesys IC3
• •
Omron Sysmac® Studio
• •
Phoenix Contact PC WORX™
• • •
Rockwell
Automation
RSLogix™/Studio
5000
• •
Schneider Electric Unity Pro
•
Siemens TIA Portal/STEP® 7
• • •

Code generation support by PLC platform. MathWorks supports code generation for PLCs and industrial
PCs for all common PLC and industrial PC platforms and works closely with PLC vendors through
the Connections Program.

6
Key Capabilities
• Generation of C/C++, IEC 61131-3, or HDL code
• Integration of automatically generated code into the PLC software through the vendor’s IDE
• Online debugging from Simulink and Stateflow using “External Mode”

Benefits
• Perform hardware-independent design and test of PLC software
• Eliminate time and effort spent on manual coding
• Reduce coding errors by automatically generating code

Digital Twin
Although “classic” Model-Based Design and virtual commissioning focus on the design phase of the
machine or production plant, simulation models are more and more often used as the basis for
“digital twins.” The digital twin is a virtual representation of the equipment that runs in parallel to
the physical asset in operation and is supplied with measured data from the physical system. Digital
twins are typically integrated into an edge device or into the IT/OT infrastructure of the production
site.

Typical Applications
• Model-based health monitoring and predictive maintenance
• Reproduction of errors from field data
• Operator training on new systems

Benefits
• Reuse simulation models over the entire lifespan of the equipment
• Integrate into the IT/OT infrastructure at the production site and IoT platforms
• Continuously update the model based on measured in production data

Conclusion
While virtual commissioning provides machine builders and industrial equipment builders with a
valuable means for testing their embedded software early in the design process—before the physical
machine or prototype is available—it covers only a limited segment of the entire design process or
even lifespan of the equipment.
Model-Based Design enables users to benefit from their simulation models over the entire design pro-
cess, including automatic testing and code generation of software that later runs on a PLC, industrial
PC, or embedded controller.
Digital twins based on the simulation models developed for the design phase go even one step further
and serve as the basis for health monitoring, predictive maintenance, and many other valuable appli-
cations for in-production use.

7
Online Resources

Webinars and Videos

Virtual Commissioning with Simulink
See how you can test your software in early project stages by using virtu-
al commissioning with Model-Based Design.

Virtual Commissioning of Production Machines

Modern production machines contain increasingly complex real-time
software, typically running on a PLC, PAC, or industrial PC. To test and
verify the functionality before the physical machine is available, virtual
commissioning based on modeling, simulation, and code generation has
become a well-proven design workflow.

Model-Based Design in Industrial Automation

At Tetra Pak, using Model-Based Design has enabled the new flagship
packaging equipment lines to run at double the capacity of the previous-
ly fastest lines.

Verification and Virtual Commissioning of Configurable

Handling Systems
Manufacturing companies that aim to increase production rates and to
reduce costs need a method to virtually test system behavior. Reishauer
AG, a producer of high-precision gear grinding machines, addresses this
challenge by using Simscape to unite CAD designs with mechatronic
simulations.

Real-Time Simulation and Testing with Simulink Real-Time

Learn how to perform real-time testing for hardware-in-the-loop simula-
tion and controller design.

8
 Customer Statements

Metso Develops Controller for Energy-Saving Digital Hydraulic System for

Papermaking Equipment Using Model-Based Design
“Using Model-Based Design with MATLAB and Simulink, we achieved
multiple goals simultaneously. We developed a sophisticated control-
ler for digital hydraulics that is more reliable, accurate, and efficient
than previous systems, and we accelerated development, which
gives us a competitive advantage.” Papermaking equipment.
The machine’s calender is
— Kari Leminen, Metso controlled by a digital
hydraulic system.

FMTC Designs and Optimizes a Hybrid Hydrostatic Drivetrain

with Model-Based Design

“Model-Based Design supports a systematic approach to the design
of drivetrains and other complex mechatronics systems. Detailed
analysis of design alternatives based on the simulation of dynamic
physical models and optimal controllers enabled us to make in-
formed decisions early in the design phase.”

— Kristof Berx, FMTC The hybrid hydrostatic

drivetrain setup.

ENGEL Speeds Development of Injection Molding Machine Controllers

“Model-Based Design reduces the time needed to produce quality

control algorithms. Simulations help us understand the system, and
code generation enables us to maintain a single source for the de-
sign. The results are faster development and higher-quality systems.”
The injection unit, which is
— Hannes Bernhard, ENGEL
driven by four
synchronized drives.

Krones Develops Package-Handling Robot Digital Twin

“Simulations of the digital twin in Simulink enabled us to obtain data

and insights that would be either impossible to get via hardware
tests or simply too costly and time-consuming. Visualizing forces and
moments helped us to understand the effects of individual compo-
nents on a highly dynamic robot.”
The Krones Robobox T-GM
— Benedikt Böttcher, Krones package-handling robot.

© 2020 The MathWorks, Inc. MATLAB and Simulink are registered trademarks of The MathWorks, Inc. See mathworks.com/trademarks for a list of additional trademarks.
Other product or brand names may be trademarks or registered trademarks of their respective holders.

10/20