SBIR/STTR Award attributes
The ultimate goal is to allow autonomous UAV teams to autonomously make decisions and coordinate their efforts in an intelligent and resilient way. This will require intelligent real-time information-sharing, planning, role allocation, and detailed path planning and autonomous, intelligent, adaptive behavior by the UAVs. Stottler Henke’s SCOUT, based on several existing mature technologies, and previously demonstrated on actual flying UAVs, completely autonomously creates path plans and automatically reacts to found targets by performing additional path planning as necessary. Furthermore, SCOUT can allocate search areas for individual UAVs based on their individual capabilities (in sensing and mobility) and autonomously plan around recently detected threats. Tactics also exist for employing multiple UAVs against a single target simultaneously. SCOUT is a general architecture that can be easily applied to different mission types and in Phase I, we will adapt SCOUT to electronic warfare domain, specifically in the A2AD environment. Missions will include deception/seduction of enemy sensors and penetration and suppression of IADS (similar to the SEAD mission which SCOUT has been proven to be effective with), which requires adapting and developing new AI techniques to automate all aspects of intelligently, automatically selecting appropriate plays, robustly assigning roles and planning routes, and adaptively executing each role, robustly and predictably in diverse environments, including intelligent information sharing. The existing SCOUT architecture provides an already-flight-tested and proven foundation for the development of the proposed EW-SCOUT. EW-SCOUT will enable fully autonomous UAS EW-related missions in the A2AD environment, allowing for allied manned aircraft to more safely penetrate the airspace. We will design the ultimate system and, to prove its feasibility, prototype all aspects of it in Phase I based on our current SCOUT architecture and demonstrate it with our already-existing Multi-Agent SimulaTor (MAST). MAST is able to simulate the movement of air and ground forces, factoring in environmental variables such as terrain and weather and executes Monte-Carlo-styled scenarios. It is capable of simulating UAV detection and tracking via sensors, human and mechanical, and communications over a variety of different technologies (e.g., visual, radar, IR, acoustic). In Phase I, we plan to make improvements to the communications feature of MAST to include new technologies such as low-energy lasers and to expand the set of DIL environment behaviors. We will also investigate integration with possible ALE platforms, representative sensors, and UAVs, supported by a group of highly qualified subject matter experts in EW operations.
