Heterogeneous Teams of Robots using a Coordinating Model for Surveillance Task based on Cellular Automata and Repulsive Pheromone

Abstract

IACA-DI (Inverted Ant Cellular Automata with Discrete pheromone diffusion and Inertial motion) is a decentralized bioinspired coordination model recently proposed for a team of robots performing the surveillance task. It uses a two-dimensional cellular automata lattice to support the navigation moves and the interaction between the robots. Cells store physical information about the environment (walls, free and robot positions) and real-coded information that enables the stigmergy among the team members. Stigmergy emerges through a repulsive pheromone map that condenses the information about the recent paths navigated by the robots, helping the team to spread over the rooms. Some variations of this coordination model were previously investigated regarding the next position decision performed at each time step. Previous results have shown a trade-off between the resultant coverage and the performance of the team related to the number of visiting cycles of the rooms. Deterministic-decision robots returned a higher number of cycles of visits (with poor coverage) while stochastic-decision robots returned a better coverage (with a lower number of visiting cycles). Here, we investigate the application of heterogeneous compositions of robots, where a portion of the team members is stochastic while the other portion is deterministic.