Distributed tunneling reconfiguration of cubic modular robots without meta-module's disassembling in severe space requirement

Abstract This paper studies a tunneling-based reconfiguration algorithm for cubic modular robots. Tunneling-based reconfiguration is a promising approach for cubic modular robot reconfiguration in severe space requirements. This is because a tunneling modular robot only uses spaces occupied by the start and goal configurations. However, previously proposed methods have a limitation on the arrangement of the start and goal configurations, in which the overlapped part between them must be connected. We propose a tunneling reconfiguration algorithm that removes the limitation and is available for cases with multi-overlapped parts between the start and goal configurations. It is often the case that a tunneling-based reconfiguration assumes the use of a meta-module-based structure to maintain the connectivity and mobility of the robot structure. However, in previous methods, the meta-modules often come apart during the tunneling process, and each module belongs to a different meta-module before and after the reconfiguration. The proposed algorithm also solves this problem. We implement the algorithm in a distributed form and prove its completeness for assumed robot structures. We examine the proposed tunneling algorithm by simulation.