Replanning for Evasive Autonomy to Counter Threats

The contract seeks the development of an AI-driven autonomy system capable of enabling resilient, decentralized coordination of UAS swarms under contested conditions by integrating multi-agent planning and adversarial reasoning grounded in game theory. The system must generate and execute strategically optimized Courses of Action for both Blue (friendly UAS swarm) and Red (adversarial threat) forces, where each side’s optimal strategy dynamically depends on the other’s actions. It must be built using a government-approved modeling framework and demonstrated in a contractor-chosen simulation environment while incorporating the Commander’s Intent into a mission utility function. The solution must scale to manage at least 200 UAS simultaneously, function reliably under electronic warfare interference, kinetic threats, and adverse weather, and support decentralized decision-making with communication-aware planning and peer-to-peer fallback mechanisms for operations in denied environments. The capability must align with AFSOC’s Adaptive Airborne Enterprise and FANTOM Futures initiatives and be suitable for integration into a maritime F2T2EA demonstration over 100 nautical miles offshore against targets protected by notional integrated air defense systems. The system must operate effectively without centralized control, autonomously adapting to dynamic threats and environmental degradation. This is a Small Business Set-Aside solicitation under the SBIR/STTR mandate, targeting businesses with fewer than 500 employees, and is managed by the Department of Defense’s Special Operations Command. Proposals are due by July 22, 2026, and must address all viable system design options with detailed specifications to ensure technical feasibility and operational viability.