Decentralized multi-robot coordination for a risky surveillance application

This paper proposes a multi-robot control methodology that is based on a behavior-based control framework. In this behavior-based context, the robotic team members are controlled using one of 2 mutually exclusive behaviors: patrolling or intercepting. In patrol mode the robot seeks to detect enemy forces as rapidly as possible, by balancing 2 constraints: the intervention time should be minimized and the map coverage should be maximized. In interception mode, the robot tries to advance towards an enemy which was detected by one of the robotic team members. Subsequently, the robot tries to neutralize the threat posed by the enemy before enemy is able to reach the camp.