Agent-oriented robot task transformation

This paper describes a new approach to modelling robot systems by a large number of cooperating agents solving a robot task. On the one hand the agent model submits to use results of the distributed artificial intelligence (DAI), on the other hand, typical aspects of robot systems must be integrated. Therefore, an agent model will be presented serving the special requirements of robot systems and offering many advantages like robustness, high efficiency, and the chance of integrating reactive and deliberative behavior. The agent model also enables the use of different programming paradigms, ranging from complex over hierarchical to object-oriented agents. Finally, a typical set of agents is presented, forming a multi-layered, distributed and reactive (MDR) robot task-level transformer