Ants Based Control of Swarm Robots for Bushfire Fighting

Abstract

The use of a swarm of autonomous robots for simulated bushfire fighting is studied in this work. A distributed and behavioral control strategy, the Ants based Control of Swarm Robots (ACSR), is proposed. In this approach, each fire fighting robot can operate independently from decentralized controllers and the goal is accomplished by the individual robot's own ability. In particular, improvements on the conventional Ant System path planning method are made which enables the ACSR capable of tackling multiple goal problems. The ACSR is evaluated in the simulation of a bushfire fighting scenario comprising of forty swarm robots and multiple fire fronts. It is shown that swarm robots using ACSR are able to locate and extinguish fire efficiently in dynamic and unknown environments.