Robotic path planning using rapidly-exploring random trees

Abstract

This study concerns the implementation of Rapidly-Exploring Random Trees (RRTs) algorithm for an autonomous robot path planning. RRTs possesses a number of advantages such as relatively simple, suitable for finding a path for a robot with dynamic and physical constraints, the expansion of RRT is heavily biased toward unexplored areas of search space and the number of edges is minimal. However, the planned path by using basic RRT structure might not always be optimal in terms of path length. Therefore, a path pruning method has been proposed to address this issue and improve the overall performance of the RRTs. Through simulations, the path pruning method has been proven to reduce paths lengths while preserving the aforementioned advantages of RRTs. A Graphical User Interface (GUI) has also been developed to demonstrate the RRTs technique in planning a path for an autonomous robot. The GUI package is designed to be interactive and user�friendly even for the users with minimal or no guidance and practice.