AI-Driven Collaboration Mode Selection for Manned-Unmanned Teaming in Future Combat Air Systems

The integration of autonomous systems and artificial intelligence (AI) is reshaping Air Dominance in Future Combat Air Systems (FCAS) by enhancing manned-unmanned teaming (MUM-T). This research presents STAR (Sharing Tasks with Autonomous Resources), a system for managing MUM-T operations within Command and Control (C2), utilizing multi-modal sensor-fusion data. STAR enables AI-driven task allocation to improve mission efficiency based on interdependence and capacities of human and autonomous agents. A human-in-the loop simulation with ten participants in a Threat Detection mission demonstrated the system's ability to optimize task allocation. Seven supervised machine learning models, were trained and tested with the highest accuracy being the Extreme Gradient Boosting (XGBoost), which achieved 89% in predicting the best collaboration mode among autonomous search (Scout), escort (Guardian), shared tasking (Collaborative), human-controlled (Follow), and backup (Request Back-up). This approach enhances interoperability within MUM-T and facilitates resource coordination across agents. STAR may also become a new COI (Community of Interest) service for FCAS, as an enabler for dynamic tasking using FCAS data during mission execution.