A Distributed Self-Assembly Planning Algorithm for Modular Robots

Abstract

A distributed modular robot is composed of many autonomous modules, capable of organizing the overall robot into a specific goal structure. There are two possibilities to change the morphology of such a robot. The first one, self-reconfiguration, moves each module to the right place, whereas the second one, self-assembly docks the modules at the right place. Self-assembly is composed of two steps, (1) identifying the free positions that are available for docking and (2) docking the modules to these positions. This work focuses on the first step. This paper presents a distributed planning algorithm that can decide which positions can be filled and can create any 3D shape, including shapes with internal holes and concavities. Our algorithm consider kinematic constraints and prevents positions from being blocked. Each module embeds the same algorithm and coordinates with the others by means of neighbor-to-neighbor communication.