Autonomous Decentralized Coordination of Heterogeneous Unmanned Systems via Adaptive Auctions

As robotics and autonomous systems continue to rapidly evolve into highly advanced, multi-mission systems capable of manned-unmanned teaming efforts, opportunities to leverage practical real-world autonomous intervehicle coordination becomes feasible. Disaster relief efforts such as search and rescue are poised to take advantage of collaborating unmanned assets to provide risk reduction and necessary stand-off for first responders. Fully custom solutions are not necessary, as collaborative capabilities can be achieved by leveraging existing or emerging UAVs to perform the physical and cognitive tasks required. To achieve feasibility, Heron Systems proposes GLUE, a novel decentralized virtual auctioneer framework enabling autonomous resource and task allocation of coordinated, heterogeneous unmanned assets. Phase 1 work focuses on software algorithm research and development through conceptually demonstrating GLUE against key tasks that make up a search and rescue operation. Heron Systems will leverage its existing plug and play robotic architecture, the Multi-Agent Cooperative Engagement (MACE) framework, as a foundation for providing the necessary infrastructure to achieve a realistic SITL simulation environment. The proposed solution provides the necessary modularity ensuring robustness to future SoA advances in autonomy. This further provides an ability to remain vehicle agnostic enabling successful integration with a wide variety of COTS vehicles.