Title:Achieving Non-Uniform Densities in Vibration Driven Robot Swarms Using Phase Separation Theory

Abstract:In robot swarms operating with severely constrained sensing and communication, individuals may need to use direct physical proximity to facilitate information exchange, perform task-specific actions, or, crucially, both. Unfortunately, the sorts of densities that are most appropriate for information exchange may differ markedly from densities that are apt for performing the task at hand. We envision a scenario where a swarm of vibration-driven robots - which sit atop bristles and achieve directed motion by vibrating them - move somewhat randomly in an environment while colliding with each other. We demonstrate that such a swarm of brushbots can predictably form high-density robot clusters along with simultaneously co-existing regions with lower robot densities. Theoretical techniques from the study of far-from-equilibrium collectives and statistical mechanics clarify the mechanisms underlying the formation of these regions. Specifically, we capitalize on a transformation that connects the collective properties of a system of self-propelled particles with a fluid system which is passive and classical, thereby inheriting the rich theory of equilibrium thermodynamics. This deeply surprising connection is a formal one and is a relatively recent result in studies of motility induced phase separation; it is previously unexplored in the context of robotics. Experiments are presented for a swarm of differential-drive like brushbots.