Distributed intelligent sensor agent system for environment mapping

This paper discusses a multi-agent system consisting of a limited set of mobile intelligent sensor agents that are exploring an environment with the goal of minimizing the environment mapping uncertainty, i.e. the entropy. A novel tree in-motion mapping method combining simplicity and speed of computation with low storage and communications requirements is proposed for the management of a network of robotic agents, each possessing limited sensing, processing and communicating operational entities. Simulation and experimental results demonstrate the efficiency of the proposed system architecture and mapping method.