Multirobot Navigation Using Cooperative Teams

Mobile robots may utilize interrobot sensing and communications to improve the cooperative navigation capabilities of a group based on optimal estimation and control algorithms [12]. In this paper, we show that cooperative teaming strategies based on these principles may be used to further improve navigation. The partition of the group into virtual teams is used to optimally balance the net speed of motion with the uncertainty in navigational position. Analysis of the CNS principles for partitioned groups leads to simulation experiments in which alternative teaming strategies are evaluated based on an overall time-uncertainty cost function. The results suggest that smaller teams of robots are preferred when interrobot sensing is more accurate, since the use of other teams as position references is more effective. Similarly, a round-robin motion plan among teams yields the best compromise between speed and uncertainty. The resulting analysis and strategies are directly applicable to cooperative robot applications in which some robot subgroups adopt navigation service tasks to assist in overall navigation of the group.