Mutual-Entrainment-Based Communication Field in Distributed Autonomous Robotic System — Autonomous coordinative control in unpredictable environment —

A mutual-entrainment-based communication field is proposed as a new control paradigm to realize autonomous coordination in a distributed autonomous robotic system. Its most interesting ability is self-organization of the control information field which indicates the functional relationship between each subsystem and the whole system, and it enables coordinative response to unpredictable changes of environment. By using such emergent information generation, we realized spontaneous and coordinative group formation in a multiple walking robot system as an example. As a result, we showed that the control information field is self-organized by the mutual entrainment between nonlinear oscillations, and it was encoded on the global phase relationship between walking rhythms in each robot. Furthermore, the information field was spontaneously reorganized corresponding to their environmental conditions. By interpreting this information, each robot could respond relevantly and in coordination to reorganize the group formation pattern as one whole system. These results suggest that mutual-entrainment-based communication field is a principle for realizing autonomous coordinative control in unpredictable environment.