Target search with a swarm of UAVs

Niki Trigoni ( SEP )

In this talk, I will discuss the problem of searching for a target in a partially unknown environment with a swarm of autonomous UAVs. I will first present our key assumptions regarding UAV initial deployment, flight characteristics and sensor capabilities, and will then propose a series of performance metrics that could be used to compare alternative search methods. I will then focus on two simple search methods, one based on the Ants approach, and the other on the Potential Field Approach. I will briefly review previous work in these two areas pointing out strengths and weaknesses of the two approaches. I will then explain how they can be applied to the multi-UAV target search problem, and will provide illustrative examples in a simulation environment.