Distributed Decision-Theoretic Active Perception for Multi-robot Active Information Gathering

Abstract

Abstract. Multirobot systems have made tremendous progress in exploration and surveillance. However, information gathering tasks remain passive: the agents receive information from their sensors without acting in order to gather it. In this paper, we present a model and an algorithm for active information gathering using the information relevance concept. In this model, robots explore, assess the relevance, update their beliefs and communicate the appropriate information to relevant robots. To do so, we propose a distributed decision process where a robot maintains a belief matrix representing its beliefs and beliefs about the beliefs of the other robots. This decision process uses entropy and Hellinger distance in a reward function to access the relevance of their beliefs and their divergence with the other robots. This model allows to derive a policy for gathering information to make the entropy low and a communication policy to reduce the divergence between robot's beliefs. An experimental scenario has been developed for an indoor information gathering mission. Our model has been compared to two different systems : one without communication and one communicating each received observation. The results show that our approach is more efficient than both systems.

Planning-Aware Communication for Decentralised Multi-Robot Coordination

“…In most cases, communication planning has been performed where the messages are observations. The value of these messages can be measured by considering their effect on data fusion accuracy [10], [24], [2]. However, for decentralised planning the information value should instead measure the expected effect on the planning reward [25], [15], [26].…”