Honeybee-like collective decision making in a kilobot swarm

Abstract

Drawing inspiration from honeybee swarms’ nest-site selection process, we assess the ability of a kilobotrobot swarm to replicate this captivating example of collective decision making. Honeybees locate the optimalsite for their new nest by aggregating information about potential locations and exchanging it through theirwaggle dance. The complexity and elegance of solving this problem rely on two key abilities of scout honey-bees: self-discovery and imitation, symbolizing independence and interdependence, respectively. We employa mathematical model to represent this nest-site selection problem and program our kilobots to follow itsrules. Our experiments demonstrate that the kilobot swarm can collectively reach consensus decisions in adecentralized manner, akin to honeybees. However, the strength of this consensus depends not only on theinterplay between independence and interdependence but also on critical factors such as swarm density and themotion of kilobots. These factors enable the formation of a percolated communication network, through whicheach robot can receive information beyond its immediate vicinity. By shedding light on this crucial layer ofcomplexity—the crowding and mobility conditions during the decision making—we emphasize the significanceof factors typically overlooked but essential to living systems and life itself.