Multiagent Coordination by Stochastic Cellular Automata

A coordination mechanism for a system of sparsely communicating agents is described. The mechanism is based on a stochastic version of cellular automata. A parameter similar to a temperature can be tuned to change the behaviour of the system. It is found that the best coordination occurs near a phase transition between order and chaos. Coordination does not rely on any particular structure of the connections between agents, thus it may be applicable to a large array of sparsely communicating mobile robots.