Leveraging Emergent Specialization in a Heterogeneous Multi-Role Swarm Control Architecture for Positional-based UAS Missions
Bradley Fraser, Claudia Szabo, Andrew J. Coyle, Robert Hunjet · 2020
Agent-based autonomous control of Uninhabited Aerial Systems (UASs) promises to improve system efficiency and reduce the burden on human operators. The autonomous control of swarming UAS is inherently difficult due to challenges of coordination, role allocation, and the engineering of emergent behavior. In addition, UAS platforms are capable of multiple, possibly heterogeneous tasks, adding another dimension to the existing spatio-temporal problem and further complicating task scheduling and allocation. In this paper, we propose an extensible multi-role swarm control architecture to appropriately allocate swarm resources in a UAS swarm. The architecture is based on pheromone influencing and multi-criteria decision making to leverage emergent specialization. Simulation results show that the proposed multi-role architecture reduces role latency by as much as 50% compared to using dedicated individual teams for each role.