An approach to multi-robot cooperative SLAM

A novel multi-robot cooperation approach for simultaneous localization and mapping (SLAM) is proposed based on local submap strategy. Robots in the state of localization enhancing can improve its localization accuracy by the way of observing redundant landmarks or get help from other accurate robots. The robot team evaluates the candidate observing pose and then decides which robot to execute the task of localization enhancement by a competition mechanism. Except the robots which execute localization enhancing, the others go into the state of exploration. The greedy algorithm is introduced to allocate the task of exploration for the robots which are in the exploration state. In such a way, the multiple robots can achieve both efficient and accurate SLAM. The simulation results show the validity of the proposed approach.