Scout-explorer multi-agent framework for terrain coverage

Abstract

Multi-Agent systems enable us to simulate swarm behaviour wherein intelligence emerges from the collective instead of the individual. The problem of terrain coverage forms a prominent field of research in multi-agent systems with wide-ranging real-life applications. We propose a new multi-agent framework called the Scout-Explorer framework which utilizes global terrain knowledge to cover a previously unknown terrain while being energy effective. The paper demonstrates results obtained via simulations over multiple metrics comparing standard terrain coverage algorithms with a spatiotemporal implementation of the proposed framework.