Self-adaptive Coordination for Robot Teams Accomplishing Critical Activities

This paper presents a self-adaptive cooperation model for autonomous mobile devices, to achieve collaborative goals in crisis management scenarios. The model, which is based on the AMAS theory, allows dynamic team formation, task allocation and reconfiguration. The global behaviour emerges from interactions among individual agents. Task responsibility allocation is done by individual estimations of the degree of difficulty and priority to achieve the task. Then each peer exchanges its evaluation records with the others in order to find out the best suited peer to take the responsibility. Research work has been done in the framework of the ROSACE project. The experimental setting based on forest fire crisis management, and a working example are also described in the paper.