Multimodal Central Place Foraging

Abstract

Implementations of central place foraging using multi-robot systems need efficient mechanisms for searching for resources and gathering them in a central home-nest location. We propose an approach that partitions the search space and assigns agents multiple behavioral roles inspired by honey bee colonies as a way of organizing the search and gather operation. Through simulation we demonstrate that this approach minimizes spatio-temporal congestion that results from many robots sharing a common space near the home-nest. We compare our role based algorithm to the Distributed Deterministic Spiral Search Algorithm (DDSA) in a high fidelity simulation environment built using ROS and Gazebo.