Multi-Vehicle Cooperative Control for Teams of Air and Sea Vehicles Performing Littoral Operations

Aurora Flight Sciences proposes to apply algorithms for cooperative tasking of multiple unmanned vehicles, which have been significantly matured in UAVs, to a variety of USV littoral search, inspection, and force protection missions. Because of our mature starting point, the focus will be to address real-world issues such as distributed implementation over intermittent communication networks; dynamic, stochastic environments; and management of vehicle loss and other multi-vehicle health management issues. Aurora’s existing collaboration with MIT researchers will be expanded to include Woods Hole Oceanographic Institute (WHOI), who together with Aurora researchers with seminal experience in Autonomous Surface Craft will work toward transitioning technologies from the UAV to the USV realm. Cooperative USV/UAV/UUV teams will be considered, in which some members play the role of communication relays (UAVs for host ship-to-USV comms, USVs for host ship-to-UUV comms. Existing multi-vehicle real-time simulations with communication emulation are already available for these studies, and the MIT-developed Robust Decentralized Task Assignment (RDTA) algorithms can be used to optimize planning for flexible, diverse unmanned teams with diverse sensor sets. Extensions to incorporate recent results in multi-vehicle health management and human interfaces to reduce operator work load will also be incorporated.