Applying the intrinsic principle of cell collectives to program robot swarms

Abstract

Many control algorithms for formation of robot swarms are often inspired by animal swarms. However, these algorithms require robots having sensing and computational capabilities and are not applicable to robot swarms working in extreme environments, such as at micro/nanoscale and in space. Here, we directly apply the differential adhesion hypothesis (DAH) of cell biology to the formation of robot swarms. Like cell collectives, swarms of sensor-less robots aggregate and sort in a self-organized manner. We quantitatively investigate the DAH principle in both swarms of cells and robots. We find that the sorting time is nonlinearly related to the levels of adhesion differences. This sheds light on the mechanisms of timing control in morphogenesis. Based on these findings, we program robot swarms to form functional morphologies by tuning their adhesion. This work advances swarm robotics in forming functional morphologies in a self-organized manner and enables us to investigate morphogenesis in cell collectives using robot swarms.