Autonomous Flight Stacks Tested on Military Drones
The push for autonomous military aircraft is accelerating, with two key flight tests recently completed. The YFQ-44A unmanned platform successfully flew with a new mission autonomy software stack. Meanwhile, Shield AI's Hivemind autonomy stack completed its first flight as part of the Collaborative Combat Aircraft (CCA) program, demonstrating progress toward modular, AI-driven combat systems.
A key enabler for these tests is the Autonomy Government Reference Architecture (A-GRA), an open-systems approach designed to prevent vendor lock. This framework standardizes the interfaces between the drone's flight systems and the mission autonomy software, allowing the Air Force to plug-and-play different AI "pilots" on the same airframe. In a significant demonstration of this modularity, the Anduril YFQ-44A flew a single sortie using two different autonomy systems. The aircraft executed mission tasks under the control of Shield AI's Hivemind stack before seamlessly switching mid-flight to Anduril's own Lattice for Mission Autonomy software to perform the same maneuvers. Shield AI's Hivemind is designed as a mission autonomy system, distinct from a basic autopilot that follows preset waypoints. It's engineered to make complex tactical decisions, enabling the aircraft to reroute around threats, engage obstacles, and respond to dynamic conditions without direct human control. The YFQ-44A platform, originally developed by Blue Force Technologies as the "Fury" aggressor drone, is a high-performance unmanned aircraft. It is designed to fly at speeds up to Mach 0.95 and pull up to 9 Gs, roughly half the size of an F-16. These autonomous systems are a core component of the U.S. Air Force's Next-Generation Air Dominance (NGAD) initiative. The CCA program aims to pair at least 1,000 unmanned aircraft with crewed fighters like the F-35, intending to increase combat mass at a lower cost than traditional fighters. As these AI-driven systems take on more critical roles, the DO-178C "Software Considerations in Airborne Systems" standard is becoming the de facto benchmark for ensuring functional safety in military aerospace. This certification process is crucial for verifying the reliability of safety-critical software components that interact with autonomous functions.
