Relative position estimation in a group of robots

Abstract

This work presents a new approach to the problem of relative position estimation for multi robot systems. It also creates the basis for establishing and maintaining a common co-ordinate system for a group of robots. The information of a laser scanner system mounted on each of the robots is used to calculate the relative position of each of the surrounding robots. The measured movement of each robot is compared to the reported movement and laser readings, which are communicated between the robots. From this comparison, a co-ordinate transformation is calculated. Once the transformation is calculated, a Kalman Filter is used to track the robots positions. The algorithm is implemented on a real multi-robot system. Preliminary results of real world experiments are presented.