Bio-inspired multi-agent search algorithms for locating and sampling environment threats

A bio-inspired multi-agent search algorithm design is presented with the goal of improving search patterns and methodologies for autonomous radiological and multi-mode sample collection. Specifically, this paper describes a search algorithm and experiments with a search parameter of a Honey Bee inspired multi-agent swarm search technique. A performance analysis of a communication parameter is analyzed of this algorithm and shows that limiting the communication between agents results in approximately 20% more unique flowers sampled but at a cost of 61% more time. In this heterogeneous swarm, the desired worker bee percentage is shown to be about 90% with or without communication for maximizing the number of flowers (environment sites) visited.