Advanced Path Planning Algorithms in

Robotics
1. Introduction to Path Planning

Path planning is a fundamental concept in robotics and autonomous

systems that involves determining an optimal or feasible path for a robot to
move from a starting point to a goal point without colliding with any
obstacles in the environment.

Applications:

 Autonomous mobile robots

 Drones and UAVs
 Industrial automation
 Self-driving vehicles

The aim is to achieve safe, efficient, and real-time navigation, often

under dynamic or partially known environments.

2. Types of Path Planning

Path planning can be broadly classified into two categories:

a) Global Path Planning

This type of planning assumes that the entire map of the environment is
known in advance.

 Input: Global map, start point, end point

 Process: A global planner algorithm computes the optimal path by
analyzing the entire map.
 Output: Precomputed, optimal route from start to end

Advantages:
  Efficient for static environments
 Ensures shortest or cost-optimal path

Example Algorithms:

 A* (A-star)
 Dijkstra’s algorithm
 D* (Dynamic A-star)


The image shows how a global planner utilizes the global map and
start/end points to determine an optimal route, avoiding known obstacles.

b) Local Path Planning

This planning is dynamic and done in real-time using sensor data from
the robot.

 Input: Local map generated from real-time sensor input

 Process: A local planner constantly updates the path to avoid
unexpected obstacles
 Output: A safe, reactive path that adapts to the current surroundings

Advantages:

 Useful in dynamic environments

 Avoids moving or unknown obstacles

Example Algorithms:

 Dynamic Window Approach (DWA)

 Artificial Potential Fields
 Local trajectory planners (e.g., MPC)


As shown in the image, the local planner continuously adjusts the path in
the presence of dynamic obstacles using real-time sensor data.
3. Path Planning Basics

To compute any path, the robot must:

1. Understand the environment (mapping)

2. Determine start and goal positions
3. Search or optimize a path avoiding obstacles
4. Follow the path while adapting to changes

Key components:

 Search space: Area where the robot can move

 Obstacles: Must be detected and avoided
 Map representation: Often as occupancy grids

4. Path Coverage in Path Planning

Path coverage refers to how well the robot explores or navigates all areas
of interest in the map.

Types of Coverage:

 Complete Coverage: Robot covers every reachable point (used in

cleaning robots)
 Partial Coverage: Robot only covers required regions (used in
logistics)

Importance:

 Ensures area efficiency in cleaning, agriculture, and surveillance

tasks
 Optimizes sensor coverage and energy use

Coverage Algorithms:

 Boustrophedon path (lawnmower pattern)

 Spiral and zig-zag patterns
 Frontier-based exploration (SLAM)
5. Algorithms in Path Planning

Search-Based (Discrete Grid) Algorithms:

 A* Algorithm: Combines travel cost and heuristic for efficient search

 Dijkstra’s Algorithm: Finds the shortest path with uniform cost
 D* Algorithm: Suitable for changing environments

Reward-Based Algorithms:

 Based on reinforcement learning

 Learn path policies through trial and error
 Maximize reward while minimizing cost

Geometric Algorithms:

 Use geometry to plan paths (visibility graph, Voronoi diagrams)

 Provide smoother paths

Artificial Potential Fields:

 Simulate virtual forces

 Robot is attracted to goal and repelled by obstacles
 Simple but can get stuck in local minima

6. Map Representation: Occupancy Grid

Occupancy grids divide the environment into cells or grids:

 0 or white cells: Free space

 1 or black cells: Obstacles
 0.5 or gray cells: Unknown area

This representation allows algorithms to compute safe, collision-free paths

based on the known probability of cell occupancy.
7. Path Cost Function

The total cost of a path is calculated as:

Path Cost=Travel Cost+Heuristic Cost

 Travel Cost (g(n)): Actual distance traveled

 Heuristic Cost (h(n)): Estimated cost to reach the goal (based on
distance or obstacle density)

This cost-based evaluation helps the planner prioritize efficient and safer
paths.

8. Conclusion

Path planning is a critical task in autonomous robotic systems. It involves

selecting a suitable strategy (global or local), using appropriate algorithms
(search, geometric, or learning-based), and considering map representations
and cost functions. With the integration of sensor data, efficient maps,
and optimal planning algorithms, robots can achieve real-time, safe, and
effective navigation in both known and unknown environments.

Force Control System in Robotics

– Complete Explanation

🔷 Introduction
In traditional robotic systems, the goal is often to control the position of the
robot’s end-effector precisely. However, in many real-world applications,
position control alone is not sufficient. Tasks that involve physical
contact—such as polishing, grinding, welding, or assembling parts—require
control over the forces and torques exerted by the robot.

This is where Force Control Systems become essential. They enable the
robot to interact safely, adaptively, and precisely with its environment.
🔍 What is a Force Control System?
A Force Control System is a closed-loop feedback control system that
controls how much force a robot applies on an object or surface. Rather than
moving only to a predefined position, the robot adjusts its position
dynamically to maintain a target force.

🔁 Closed-Loop Concept

It uses a sensor (force/torque sensor) to continuously measure the force

being applied. This information is fed back into the controller, which adjusts
the motion of the robot in real-time to maintain the desired force.

🧠 Understanding Compliance in Robotics

🔹 What is Compliance?

Compliance means flexibility or adaptability in response to an external

force. In robotic systems, compliance is the ability of a robot’s end-effector
to "give way" or adjust when interacting with objects, instead of rigidly
resisting.

🔹 Types of Compliance:

✅ Passive Compliance:

 Achieved mechanically using flexible joints, springs, soft

materials.
 No feedback or sensors.
 Simple and low-cost, but limited in adaptability.

✅ Active Compliance:

 Achieved through sensors, feedback loops, and controllers.

 Uses force/torque measurements to adjust robot behavior.
  More intelligent, responsive, and suitable for dynamic environments.

🔲 Block Diagram – Detailed Breakdown

Here’s an expanded explanation of your uploaded Force Control Block
Diagram:

pgsql
CopyEdit
+-----------------+
| Force Sensor | ← Senses real force at contact point
+--------+--------+
|
Actual Force (F)
|
V
+---------------+ +------------------+
| Force Control | ←→ → | Desired Force | ← Input
from operator/task
+-------+-------+ +------------------+
|
Correction Signal (ΔF)
|
V
+------------------+
| Position Control | ← Converts force error into
position adjustment
+--------+---------+
|
V
+---------------+
| Robot Arm |
| (Actuator) |
+-------+-------+
|
V
+---------------+
| Workpiece |
 | (Surface) |
+---------------+

Component Roles:

Component Function
Detects actual force/torque at the contact
Force Sensor
point.
Compares actual force with desired force,
Force Control
calculates error.
Position
Adjusts robot position based on force error.
Controller
Robot
Moves the robot/tool accordingly.
Actuator
Work Surface The object being worked on.

📌 Where is Force Control Needed? (Use Case

Explanation)
In contact-based robotic applications, a robot might be:

 Grinding a surface with even pressure,

 Assembling parts without excess force,
 Welding with just enough contact pressure,
 Cleaning or painting where light touch is required.

Without force control:

 Too much force = breakage or damage

 Too little force = incomplete task or quality issues

Your uploaded image diagram demonstrates a robot tool sliding across a

surface, needing to maintain constant normal force:

 The robot is moving in one direction (sliding).

 Simultaneously, it must apply consistent downward force.
 Any change in surface height must be compensated for automatically
by adjusting the robot's position.
⚙️Types of Force Control Strategies
There are multiple methods for implementing force control, depending on
the complexity and requirement:

1. Impedance Control

 Controls dynamic relationship between force and motion (like a

virtual spring-damper system).
 Suitable for uncertain environments.

2. Admittance Control

 Takes external force as input and controls motion as output.

 Often used when force is sensed at the end-effector.

3. Hybrid Position/Force Control

 Controls position in some directions and force in others.

 E.g., control position along X/Y axis, and force along Z-axis.

💼 Applications of Force Control Systems

Application Role of Force Control
Surface Grinding /
Maintains even pressure for smooth finish.
Polishing
Applies just the right force to fit parts together
Assembly
without jamming.
Spot Welding Maintains contact force to ensure proper welds.
Fragile Object
Prevents damage by limiting applied force.
Handling
Maintains steady tool contact with varying
Deburring
surfaces.
Ensures safe contact during surgeries or patient
Medical Robotics
handling.
✅ Advantages of Force Control
 🌟 High Quality Output: Ensures uniform processing (e.g., even
surface finish).
 🤖 Adaptive Interaction: Adjusts to environmental changes like part
misalignment.
 💡 Intelligent Control: Makes robots capable of "feeling" the
environment.
 🧩 Versatile Use: Suitable for both rigid and soft object interaction.
 Prevents Damage: Protects both robot and object from excessive
force.

🚧 Challenges in Force Control

 Sensor Noise: Force sensors can introduce errors or delays.
 Complex Control Laws: Requires tuning of feedback parameters.
 Dynamic Environments: Needs fast adaptation to changing surfaces.
 Force Overshoot: Can occur if controller isn’t well-calibrated.

📈 Example – Real-Life Use Case

Let’s say a robot is grinding the surface of a metal component:

 Without force control: Surface imperfections or uneven areas would

cause too much or too little grinding.
 With force control: The robot adjusts its tool height in real-time to
keep pressure constant, ensuring perfect finish and no damage.

📑 Summary
Concept Description
A feedback control system to regulate
Force Control
contact force.
 The robot’s ability to adapt to external
Compliance
forces.
Measures real-time force at the contact
Force Sensor
point.
Active
Uses feedback and control algorithms.
Compliance
Assembly, polishing, welding, object
Applications
handling.

Manipulator vs. Wheeled Mobile

Robot & Components of Mobile
Robots

🔷 Manipulator vs. Wheeled Mobile Robot

Feature Manipulator Wheeled Mobile Robot
Operates in
Operates in a constrained
Workspa unconstrained
workspace (fixed working area
ce environments (indoors or
like a factory cell)
outdoors)
Position Has absolute knowledge of its Needs external sensing
Awarene position (e.g., joint angles, link (like GPS, LiDAR) to
ss lengths) determine position
Environm
Requires sensing to avoid
ental May or may not require sensors
obstacles and navigate
Perceptio to perceive surroundings
environments
n
Mobile robots, self-driving
Industrial robotic arms, surgical
Examples cars, autonomous delivery
robots
bots
🔧 Components of a Mobile Robot
A mobile robot is a blend of hardware and software subsystems. Each
plays a vital role in ensuring that the robot can move, sense, think, and act
in the environment.

🔹 Hardware Components

Subsyste
Function
m
Locomotio Enables the robot to move across environments (wheels, legs,
n tracks)
Measures robot's internal state and external world (e.g., IMU,
Sensing
LiDAR, cameras, GPS)
Control Executes the planned actions and ensures stable motion
Communic
Transfers data to/from operators or other robots
ation
Power Provides energy to all components (e.g., battery, solar panels)

🔹 Software Components

Subsystem Function
Reasoning / Maps sensory inputs into decisions (e.g., navigation, path
Planning planning)
Perception Interprets sensor data (e.g., object recognition, mapping)
Localization Determines robot’s position in the world
Decision Chooses actions based on goals and environment (e.g.,
Making obstacle avoidance)

🚗 Wheeled Mobile Robots (WMRs)

Wheeled Mobile Robots are robotic platforms that use wheels for locomotion
on flat or moderately rough terrain.

🔹 Examples:

 🤖 Justin – a mobile humanoid robot for research

 🚗 Self-Driving Cars – autonomous road navigation
  🚀 NASA Mars Rover – exploration in unknown terrains
 🏭 Mobile Manipulators – robot arm on mobile base for warehouse
operations

🔹 Applications:

 📦 Warehouse Automation
 🚓 Surveillance & Security Patrols
 🏥 Patient Assistance in Hospitals
 🚛 Material Transport in Factories
 🛰 Space Missions & Exploration
 🚗 Autonomous Navigation in Urban Roads

🔄 Types of Wheeled Mobile Robots

1️⃣Based on Number of Wheels

 Single Wheel (e.g., ball-balancing robots)

 Two-Wheel (e.g., differential drive robots)
 Three/Four-Wheel (e.g., delivery bots, mobile bases)
 Multi-Wheel (e.g., Mars Rover)

2️⃣Based on Wheel Type

Wheel Type Motion Description

Uni-
Tyre Wheel Common car-like wheel
directional
Bi-
Mecanum Wheel Allows side movement using rollers
directional
Ball Wheel / Omni- Can move in any direction; used in
Omni Wheel directional indoor robots

3️⃣Based on Terrain Capability

 Indoor/Smooth Terrain – warehouse robots

 All-Terrain – off-road and exploration bots
 On-Road Navigation – autonomous cars, delivery robots
🛞 Steering Mechanisms in Mobile Robots
Steering defines how the robot changes its direction.

🔹 1. Ackermann Steering (Car-like)

 Used in cars and self-driving vehicles

 Separate wheels for steering and power
 Efficient for high-speed, on-road navigation

🔹 2. Differential Drive

 Used in many two-wheel robots

 Each wheel has its own motor
 Turns by varying wheel speeds
 Compact and ideal for indoor use

🔹 3. Skid Steering

 All wheels are powered; turns by skidding

 Similar to tanks or bulldozers
 Used in rugged or tight spaces

⚠️Steering Constraints in Differential Drive

❗ Non-Holonomic Constraint

 A differential drive robot cannot move sideways.

 It cannot move directly to a position unless it's aligned in that
direction.

🧭 Solution:

 Must rotate first or combine rotation with translation (curved

paths)
 Common in indoor navigation and corridor movement