A Leader Based Coating Algorithm for Simple and Cave Shaped Objects with Robot Swarms

Abstract

Self-organizing swarm systems promise an essential progress in several research areas. They can find application in health context, such as through nano-robots, or material context, such as for health-healing properties. Swarm behavior in nature is well known e.g. bird flocks that protect individual birds, ants that collaboratively transport bigger objects, or predators who hunt in packs. Applying these strategies to future robot swarms promises to open further opportunities. In this paper, we address the problem of object coating, in which a swarm of “subjects” aims to position itself as tightly as possible around an object. We introduce a leader-based coating algorithm that promises near-optimal performance. A leader is elected to coordinate the actions of the swarm; thus, it scans the entire object and plans the placement of the individual swarm members upfront. Through its coordination, each individual subject is placed around the scanned object. While simple shapes are easy to master, a strong emphasis of this paper lies on dealing with cave-shaped objects in which passages or trails might be blocked in a naive approach. We elaborate how to deal with such problems and how our algorithm addresses this issue. In the evaluation we showcase that our approach is able to coat cave shaped objects with a 100% success rate.