Summary

In my report titled "RRT and RRT* Motion Planning for a Robot in a Parking Lot," I explored the implementation and comparative analysis of two motion planning algorithms, Rapidly-exploring Random Tree (RRT) and RRT*. My focus was on navigating a robot through a simulated parking lot environment, filled with static obstacles, using these algorithms. I utilized a kinematic bicycle model for the robot's movement and integrated a PD controller for path tracking. The simulation environment was created using Pygame, providing a visual representation of the parking lot scenario. I developed both RRT and RRT* from the ground up, with RRT based on a tutorial and RRT* implemented independently. My findings indicated that although RRT* tends to generate shorter paths, it does so at the expense of increased computational time compared to RRT. This study highlights the balance between path efficiency and computational resource requirements, suggesting future explorations in advanced RRT* algorithms, alternative path generators, and controllers, as well as more dynamic scenarios such as moving obstacles. This work underscores the potential and complexities involved in robotic navigation within structured environments like parking lots.

Project documents

The report:

• Download PDF

Code implementation

Code for this blog:

• WEBSITE

Code to reproduce the results:

• PROJECT CODE