Path and sensing point planning for mobile robot navigation to minimize the risk of collision

Abstract

A path planning algorithm for mobile robotics is proposed. When a mobile robot is navigating, the error in estimated position of the robot increases in proportion to the traveling distance. Because of this error, the robot has a greater risk of collision with objects in the environment. To reduce this error, a robot should obtain environmental information by using its environment sensors, and modify its estimated position accordingly. The authors propose a cost function for the mobile robot's path which minimizes the risk of collision and an algorithm to find the optimal path and sensing points. They have implemented the proposed algorithm on a real robot system and conducted several experiments to verify it.