Adaptive self-configuration of robot swarms and its convergence degree

Abstract

This paper addresses the problem of the coordinated deployment of a swarm of mobile robots in an area with geographical constraints. The objective of our research is to investigate how to build an ad hoc network of robotic sensors in a real-world environment. For attaining scalability and robustness, it is desirable that each robot can configure themselves into an area through local interactions with adjacent robots, using only locally available information. Therefore, a decentralized deployment scheme is presented based on geometric ideas allowing robots to converge to the uniform spatial distribution by forming regular triangle lattices. The convergence of the deployment scheme is mathematically examined, and verified through extensive simulations and experiments. In addition, a new evaluation index is proposed, to estimate a relative convergence degree with respect to a desired geometric configuration. Our results indicate that the deployment scheme can be applied to the problem regarding the coverage of an area of interest by a swarm of mobile robotic sensors.