Experimental Investigation of Mutual Collision Avoidance Behavior for Multiple Mobile Robots

Abstract

Introductions of multiple mobile robots into real environments have been reported. In the environments where multiple robots co-exist together, a mutual collision-free motion planning with awareness of each other’s movements is important. This study focuses on a distributed system where each robot avoids obstacles and other robots without communication. Dynamic avoidance methods based on the prediction of the movement of moving objects have been proposed. However, when robots applying such methods confront each other, an oscillation problem where robots synchronously jiggle their avoiding direction is prevalent. This study proposes a way for realizing multi-robot collision avoidance eliminating or alleviating oscillation problems by extending the dynamic path planning method. We aimed to eliminate or alleviate the oscillation problems by finding the key parameters that affects and leads to different interaction behavior when plugged with different values. In this study, three parameters: sensing range, update frequency of path planning and robot’s maximum angular velocity, that may affect the robot’s behavior of obstacle avoidance are picked up. We experimentally verified the behavior of robots applying different parameters of the dynamic obstacle avoidance method when they face each other. From the experimental results, it was confirmed that the oscillation problem can be eliminate or alleviate by applying different parameters to the confronting robots.