Motion Planning for Self-Driving Cars

The course Motion Planning for Self-Driving Cars is the final course among four courses in Self-driving cars specialisation offered by the University of Toronto. The level of the course is advanced and is designed for the learners having a background of robotics and knowledge of controllers and models taught in course 1 of the specialisation.

In the course, the participant will be introduced to major planning related tasks in an autonomous mode of driving along with various other types of planning like behaviour planning, local planning, and mission planning. At the end of the programme, the participant will be able to find the shortest and direct path over a road network or a graph by making the use of A* algorithm and Dijkstra’s. They will be using machines in the finite state for selection of smooth paths, optimal designs, and identification of velocity profiles required for the navigation around the obstacles with safety and in accordance to the traffic rules. The participants will also get a chance to develop occupancy grid maps for the static objects prevailing in the environment and use them to check efficient collision.

The course will provide the opportunity to develop a complete self-driving planning solution that will provide a contemporary driving experience as well as safety while taking an individual from home to work. As a part of the final project, the candidate will be implementing the planner of hierarchical motion by navigating through the sequence of scenarios in the CARLA simulator, safe navigation during intersection, and avoiding the already parked vehicles in the lane.