Agent-based coordination of human-multirobot teams in complex environments

Room clearing, in which building surveillance is conducted to search for criminals, continues to be a dangerous and difficult problem in urban settings, for both the military as well as for police. In a typical setting, an unknown number of hostile forces may be located in a building, and they may be armed. Furthermore, there may be innocent civilians. The goal of the friendly units is to enter the room and secure it, but without loss of life of friendly forces, hostile forces, and most especially of innocent civilians. It would be beneficial to allow robots to be a part of the friendly team, however it is very challenging to have robots that do not either slow down or obstruct their human teammate. This is especially difficult since nearly all robots in use by the military and police today are tele-operated. In this paper, we describe work we have developed in cooperation with the army, for the room clearing domain. We constructed an algorithm whereby multiple agents, in the form of robots, can accomplish a room clearing task. We augmented the agent algorithms to introduce Adjustable Autonomy, allowing cooperation with humans. We describe simulated results of the algorithm on building maps, and furthermore we describe how we intend to next conduct hardware tests, and eventual plans to field the system. This agent-based solution has great potential to increase the acceptance and leverage of robotics in complex environments.