Mixed species flocking for heterogeneous robotic swarms

Abstract

Flocking models have numerous applications. Some include the simulation of traffic patterns such as the flow of cars on a freeway which has a flock-like motion, animating troop movement in real-time strategy games and in simulating mobile robot movement. Such numerous applications in different areas make the study of flocking efficiency a necessity. The originality of this paper lies in the extension of single species flocking models into the more complex mixed species flocking arrangement. The objective of this paper is threefold: (i) to provide a definition for flocking efficiency, (ii) to develop and analyze a first model for mixed species flocking, and finally (iii) to model and implement "heterospecific-attraction" between different agent types to improve the flocking efficiency in a mixed species flocking environment. The proposed model for mixed species flocking is highly efficient, accurate and consistent with the biological equivalents. The model offers benefit to heterogeneous and multi-robot systems.