Decentralized Asset Manager for Layered Sensing

ABSTRACT: Toyon Research Corporation proposes to develop a decentralized asset management framework which will provide near-optimal dynamic cooperative control of diverse assets within the Layered Sensing Architecture. We propose to develop an asset manager which will perform allocation, assignment, vehicle routing, and sensor tasking for a collection of assets while taking into account other asset managers and assets. The proposed design creates a"team of teams"behavior, which can accommodate large numbers of diverse assets with loose synchronization. Interfaces will accept information needs from multiple sources and will be"plug-n-play"so that organic assets can come on-line at any time to help fulfill mission objectives. We will build upon Toyon"s existing fingerprinting, fusion, prediction, tasking, and routing algorithms and the Tracked Object Manager (TOM). In Phase I, we will develop the prototype asset manager, build a realistic contemporary operational scenario, and demonstrate real-time closed-loop plug-and-play operation of multiple asset managers controlling assets simulated using SLAMEM. In Phase II we will advance the prototype asset manager and perform live closed-loop testing of the new decentralized control framework using multiple asset managers and real unmanned aerial vehicles at Camp Roberts. BENEFIT: The proposed decentralized asset management will provide loose closed-loop control of diverse sensor systems operating in a layered sensing architecture. The new technology will enable organic assets and sensor systems to simultaneously support a multiple missions and common operating pictures (COPS). The developed test cases used for evaluation can also be used by AFRL to evaluate other layered sensing concepts and technologies.