Coordination of Robot Agency by Fuzzy Rules

An optimal navigation strategy for a robot agency is related to the team colony coordination, environment areas subdivision and the best exploration strategy of each robot citizen taking into account the government political stance. The last item depends also on the assigned sub-area features: mission environments can be empty, with obstacles, with walls and corridors etc. Adopting the Metaphor of Italian Politics (MIP) architecture, we have developed a robot agency coordination strategy based on fuzzy rules. The MIP architecture takes inspiration from the political organizations of the democratic governments, where the leader isn't only one robot but it is constituted by a government of robots. In the MIP architecture the robots team is coordinated by a Prime Minister, a Minister of the Defense and a Minister of Communication while a second group of robots, the Robot Citizens, are the executors of each mission. The model of the agents is hybrid (reactive and deliberative), so every robot can assume every political position inside this dynamic structure. In our approach the MD and the navigation selection module is based on fuzzy rules. The fuzzy module is based on apposite merit factors choosen in order to guarantee the exhaustive evaluation of the efficiency of the citizen exploration. To validate the effectiveness of the our approach we have developed a framework based on the Mission Lab software developed at the Mobile Robot Lab of the Georgia Institute of Technology.