Planning in a known environment with Stationary obstacles: Potential fields, Dynamic programming, Graph search, Depth-first, Breadth-first, and Djikstra, A*;
Unknown environment and stationary obstacle: Weighted A, Anytime A*, D* Lite;
Complex high dimensional environment: Sampling based algorithms, Visibility Road maps and randomized trees, graphs: RRT and RRT*;
Realistic robot models: Reed-Shepps Car, Dubins Car, Differential drive robot and kinematic constraints, articulated arms;
Planning in the presence of uncertainty: MDP based approaches;
Planning in the absence of system model: Machine Learning and Reinforced Learning based approaches.
Suggested Lab Sessions:
· Introduction to ROS (2 or latest stable version) – ROS Basic Concepts: Nodes, topics, parameters, services – Simple ROS programs to publish and subscribe messages. Simulation of robot systems in ROS: Manipulators, wheeled robots in scenarios such as in a maze etc., legged robots and UAVs in various environments.
· Implementation of motion planning algorithms in MATLAB/ROS/Equivalent.
· Understanding Kinematic models for Mobile Robots, Maneuverability, Dynamic Path Planning, Scenario based control, path planning and sensor fusion, Workspace & Motion control, Sensors & Actuators for Mobile Robots, Sizing and Torque Calculations.
· Develop Autonomous Robotic System as a physical working prototype.