Cooperative Multi-Robot Path Planning

In risky applications it is advantageous to use multiple robots in order to improve performances. In a research or patrol scenario robots have to move cooperatively in order to avoid collisions and to improve performance. In this paper we describe the implementation of two variations of a cooperative planning algorithm: Cooperative A* and Cooperative Voronoi A*. The task is decoupled into a series of single agent searches. The individual searches are performed in a 3D space-time search space and take into account the planned routes of other agents. The method is able to plan in 2D and 2.5D environments by incorporating traversability information. The algorithm can also handle single and multiple waypoints. These can be chosen by the user or automatically generated in function of an a priori known map and the characteristics of the detection sensors. A simulator has been developed in order to rapidly evaluate and compare algorithms and to analyse the influence of configuration parameters. A summary of the results and their discussion are presented in this paper.