RepAtt: Achieving Swarm Coordination through Chemotaxis

Abstract

Swarm foraging is a common test case application for multi-robot systems. In this paper we present a novel algorithm for improving coordination of a robot swarm by selectively broadcasting repulsion and attraction signals. Robots use a chemotaxis-inspired search behaviour based on the temporal gradients of these signals in order to navigate towards more advantageous areas. Hardware experiments were used to model and validate realistic, noisy sound communication. We then show through extensive simulation studies that our chemotaxis-based coordination algorithm significantly improves swarm foraging time and robot efficiency.

Swarm Foraging Under Communication and Vision Uncertainties

“…This paper describes the Repulsion-Attraction (RepAtt) algorithm, which uses simple communication and a chemotaxis-inspired behaviour to improve coordination in a swarm of foraging robots. RepAtt was first proposed in our previous work [5]. This paper extends our previous work by using real-world vision experiments and deep neural networks for object detection as a basis for modeling a probabilistic robot vision system (more details in Section III-B).…”

“…The RepAtt algorithm discussed in this paper was first introduced in [5]. RepAtt is inspired by the chemotactic search behaviour observed in micro-organisms such as the Escherichia coli bacterium and Caenorhabditis elegans nematode.…”