A Comparative Analysis of Foraging Strategies for Swarm Robotics using ARGoS Simulator

In the field of exploration strategies for teams of autonomous vehicles, one relevant set of solutions build upon the so called foraging algorithms, which mimic the foraging strategies of animals and insects, such as bugs and/or ant colonies. In the literature, it is most often observed that the choice of the foraging strategy to be applied for a specific swarm robotics problem does not rely on quantitative and objective selection criteria but, rather, it is guided solely by qualitative guidelines. Hence, this paper proposes a quantitative review of four popular foraging strategies, namely solitary foraging, behavioural matching, stigmergical foraging and signalling. A quantitative evaluation of their performance in terms of collectible or goal acquisition in different operating scenarios is proposed together with a comparison of their computation times when the size of the swarm changes. The comparative simulations presented to provide evidence of the different approaches efficiency have been implemented with the ARGoS simulation tool.