Robust and Autonomous Management of Mission Systems (RAMMS)

SSCI proposes RAMMS: Robust and Autonomous Management of Mission Systems. RAMMS incorporates autonomy into drive Defensive Mission Systems, tasking these systems to improve operator survivability and lethality while simultaneously reducing operator workload by switching from an “operator in the loop” paradigm to an “operator on the loop” paradigm. RAMMS key technologies include (1) A fuzzy state representation of the mission state, which allows for operator-understandable representations of uncertainty and risk, (2) a genetic fuzzy tree for selecting “Plays” for particular mission states, which allows for formal verification (in support of flight-qualifying the RAMMS software) and demonstrated superhuman performance in relevant mission sets including many-vs-many air-to-air combat, while also being human-understandable (3) an online shielded genetic algorithm for optimizing Plays for specific threat pictures in situ, while guaranteeing compliance with key constraints such as keep-out zones and ROEs. RAMMS has gained the support of AFLMC and AATC, as the RAMMS system both serves an immediate need for further maturation of the capabilities of 4th generation aircraft, while also acting as a risk-reduction exercise for injection of autonomy into 5th generation aircraft. The RAMMS system is architecturally aligned with the USAF Autonomy Government Reference Architecture (A-GRA), ensuring an ability to rapidly deploy software from this R&D effort to the DAF.