OptiSwarm: Optical Swarm Robots Using Implicit Cooperation

Abstract

Swarm robots have always served as verification platforms and deployment tools for swarming models that usually take relative distances, bearing angles, velocity directions, or differentiation of neighbors as inputs to regulate individual motion. With natural decentralization and high scalability, existing swarm robotic platforms based on purely implicit cooperation exhibit excellent potential for evaluating and deploying swarming models in denied environments. However, most of the current implicit cooperation platforms are limited in their adaptability because they are hardly capable of providing the above four inputs without external assistance. To address this problem, this article presents optiSwarm, a swarm robotic platform, where each robot (Cubot) within this platform is based on a novel implicit cooperation system. Cubot can not only directly obtain relative distances, bearing angles, and differentiation of adjacent robots through vision-based local rules but also has a relative velocity direction perception mechanism purely on implicit communication. The robot’s perception is proven to be reliable through real-world tests. Through implementing three collective behaviors guided by diverse inputs, optiSwarm is employed and evaluated in typical scenarios demonstrating that it is capable of verifying models of collective behaviors.