Generating Collective Exploration Behavior of a Flock of Robotic Swarms Based on a Heterogeneous Boids Model

Abstract

Swarm robotics is one of the research field that uses a large number of simple autonomous robots and is expected to be applied to various operations in uncertain environment such as exploration activities at disaster sites. In this research, the task is to search a larger area along the way while the Boid-based swarms move together in an environment with obstacles. In a homogeneous swarm, all individuals move in a similar manner, often resulting in low exploring performance. Therefore, we generate a swarm by setting the collision avoidance distance of each robot from other robots to be heterogeneous with different advantages of following, spreading, and examine the effect of individual differences on the search performance. The results show that swarms with larger differences among individuals can improve search performance.