AUTONOMOUS CUEING WITHIN HETEROGENEOUS ROBOT SWARMS

Abstract

When integrated into military operations, autonomous swarm technology has the potential to alleviate some of the challenges presented by demanding and highly dynamic environments and problems. Autonomous aerial swarms composed of heterogeneous robots can further increase flexibility by providing a broader range of capabilities. Heterogeneous swarm effectiveness is currently limited, however, by the system’s ability to assign robots to tasks in a globally optimal manner. We propose a market-based approach to the unmanned aerial system’s decision-making for allocating the most suitable robots to the available tasks. Specifically, we evaluated the performance of an auction algorithm in task allocation for an area search problem. In addition to addressing swarm heterogeneity, our implementation accounts for situations in which secondary contact investigation tasks are generated asynchronously. Experiments were conducted with the Naval Postgraduate School Advanced Robotic Systems Engineering Laboratory’s aerial swarm system with various fixed-wing and rotary-wing unmanned air vehicle configurations ranging from three to ten robots. Our research and testing demonstrate that the auction algorithm is a scalable approach to task assignment in area search and suggests that our implementation can efficiently scale to swarms with arbitrary capability distributions to address highly complex problems.