Self-reconfigurable modular robots and their symbolic configuration space

Abstract

Modular and self-reconfigurable robots are a powerful way to design versatile systems that can adapt themselves to different physical environment conditions. Self-reconfiguration is not an easy task since there are numerous possibilities of module organization. Moreover, some module organizations are equivalent one to another. In this paper, we apply symbolic representation techniques from model checking to provide an optimized representation of all configurations for a modular robot. The proposed approach captures symmetries of the system and avoids storing all the equivalences generated by permuting modules, for a given configuration. From this representation, we can generate a compact symbolic configuration space and use it to efficiently compute the moves required for self-reconfiguration (i.e. going from one configuration to another). A prototype implementation is used to provide some benchmarks showing promising results. © 2011 Springer-Verlag.