Path Planning and Patrolling for a team of Car-like Robots in a Campus Environment

For self-driving cars to traverse roads, it requires a robust path planning system. The planning system is responsible for high-level path planning such as route planning (which roads/ route to take to reach from point A to point B), path generation (generating a smooth curve to avoid abrupt acceleration/ deceleration and jerky behavior of car), and low-level path planning such as behavior selection (selecting behavior such as lane-keeping, intersection handling, traffic light handling), motion planning (generating the reference path along with reference speeds) and obstacle avoidance (reactive avoiding obstacles in its path).
  During my work at Indian Institute of Technology, Bombay, I have worked towards designing and developing several modules of a reactive path planning system for autonomous mobile vehicles, allowing the vehicles to navigate predefined areas from the world map. Figure below displays multiple vehicles traversing roads in our simulated environment.