Distributed multirobot exploration maintaining a mobile network

We present a behaviour-based architecture for multirobot exploration and mapping. The architecture is designed to guide the exploration in a decentralized fashion constrained to maintain local short-range communication in a mobile ad-hoc network. Exploration with multiple robots has been extensively studied but communication constraints have been ignored. We focus on the integration of such constraints and develop a behaviour-based approach. The behaviours are designed to enhance global performance and are triggered based on local information. Robots are encouraged to move together forming a mobile network and sharing relevant information for the team. Up to this date there is no research addressing the effects of communication and sensor constraints on exploration and mapping in structured (office like) environments. Our experimental results show that the increment in time due to the introduction of communication constraints is reasonable and decreases as the size of the team increases.