In-Place Distributed Heterogeneous Reconfiguration Planning

Abstract

This paper describes a distributed planner that allows a self-reconfiguring robot consisting of heterogeneous modules to change configuration without using extra space. This work builds on previous work on reconfiguration planning for ho-mogeneous robots and our recent work on heterogeneous reconfiguration planning that requires temporary working space to execute. Motivated by a specific algorithm for the distributed Crystal robot, we describe and analyze an O(n 2)-time reconfigu-ration algorithm for systems of modules with assigned types that uses a constrained amount of free space.