Chain Formation in a Swarm of Robots
Nouyan Shervin, Dorigo Marco

Abstract:

In this paper, we present our first steps towards applying swarm intelligence methods for solving exploration and navigation tasks performed by a swarm of robots in unknown environments. Our approach consists in using chains of visually connected robots that collectively explore their environment. We adopt the idea of robotic chains from Goss et al., and realize our system stressing the swarm intelligence approach. We conducted a series of experiments in simulation and put the emphasis on evaluating the dynamics of the chain formation process. In particular, we analyse several aspects of the quality of the chains, such as the shape of the formed chains or the speed of the chain formation process, when varying robot group sizes and the values of control parameters. The results show that our simple control system can be easily tuned to obtain different behaviours at the group level.