Coordination control design of heterogeneous swarm robots by means of task-oriented optimization

This paper is concerned with a framework to design self-organizing, self-reconfigurable robotic systems. We focus our attention on the algorithm of a multi-agent system called Swarm Chemistry, proposed by Sayama (Artif Life 15:105–114, 2009). In this model, a number of agents that have non-uniform kinetic properties coalesce into an excellent diversity of spatial structures and/or emergent behaviors, depending on the kinetic parameters provided. However, such bottom-up nature cannot be easily applied to the conventional and top-down design of artifacts. This paper presents a method of designing heterogeneous robotic swarms and finding solutions through a genetic algorithm. Simulation results with a few simple task examples demonstrate that the proposed framework allows us to acquire appropriate sets of kinetic parameters, i.e. recipes, creating swarm structures to perform a given task more effectively and efficiently. Such autonomous robots can be deployed for the purposes like disaster prevention, geographical survey, and subsea exploration.