Coordination of Heterogeneous Autonomous Robots in Tunnels and Subterranean Environments (CHARTS)

Missions of importance to the Defense Threat Reduction Agency (DTRA), particularly missions countering weapons of mass destruction (WMD), are difficult to conduct in subterranean environments. While there has been substantial work in using individual robots to remotely explore these environments, the difficulties presented in the subterranean realm provide an opportunity for multiple robots to cooperate in their exploration. We propose to design and demonstrate Coordination of Heterogeneous Autonomous Robots in Tunnels and Subterranean Environments (CHARTS), a framework that provides a multi-robot system with behaviors and contextual reasoning to map, explore, and discover objects in a communications-limited environment. In our approach, we will: (1) extend our existing TRL-4 Swarm Coordination Framework (SCF), which focuses on coordination of decentralized, heterogeneous agents and robots, to the challenges of counter-WMD missions in subterranean environments; (3) design mission models and behaviors that will allow multi-robot systems to coordinate their activities and achieve mission objectives; and (4) design a contextual reasoning and distributed learning component that lets agents reason under uncertainty and flexibly learn from each other’s experiences to improve group performance. We will simulate and evaluate CHARTS and identify opportunities to commercialize and transition this work, especially in service to the future of DTRA missions.