Multi-agent Modelling for Autonomous but Cooperative Robots

Modelling a society of cooperative and autonomous mobile robots can be viewed as a multi-agent issue, i.e., the task to determine a multi-agent architecture for a robot society according to this two requirements. In this paper, we motivate a multi-layered approach to fulfill cooperation and autonomy. We propose an agent model together with an interaction model based on this approach. Features of this model are the clear distinction between social and individual capabilities and knowledge, and the possibility to provide interactions at several levels of abstraction to guarantee cooperation between agents.