Pheromone Interactions in a Cellular Automata-Based Model for Surveillance Robots

This work investigates a coordination model based on a two dimensional cellular automata applied to a team of surveillance robots. The synergy among the robots emerges from the indirect communication performed by repulsive pheromone interactions. Five strategies are evaluated for the decision-making related to the next-cell selection: three stochastic (pure, elitist and inertial), one random and one deterministic. The performance of the team performing surveillance are evaluated in respect to two aspects: the number of task cycles (visiting all the rooms) completed in a fixed interval of time and the homogeneity of the environment coverage. Experimental results corroborate the importance of the cooperative pheromone and shows that the decision-making strategies have different inherent skills that can be explored for distinct situations.