Towards Coordinated Multi-Robot Exploration under Bandwidth-constrained Conditions

Advances in robot locomotion algorithms, sensor technologies, planning frameworks and hardware designs have motivated the deployment of robot groups in unstructured environments. One of the most common tasks is autonomous exploration. During the multi-robot autonomous exploration task, it is desirable for robots to coordinate in order to steer them to different areas so as to increase the overall efficiency. Such coordination requires communication among peer robots. In some environments, such as tunnels, natural caves or urban underground spaces, it is not always feasible to maintain a stable communication network between robots. Thus, robust coordinated exploration solution must operate in unstable or constrained settings. In this work, we propose a decentralized and bandwidth-efficient algorithm to robustly coordinate robot groups to explore unknown unstructured environments. The proposed method's performance is evaluated in typical simulation environments and the experiment results show much reduced average bandwidth consumption compared with baseline algorithms. In addition, the proposed algorithms are implemented in a quadruped robot groups to demonstrate the effectiveness in a real world autonomous exploration task.