On April 6, 2023, General Atomics Aeronautical Systems (GA-ASI) successfully executed operational air combat maneuvers using AI pilots to control an MQ-20 Avenger Unmanned Aircraft System. These collaborative maneuvers between human and AI pilots were achieved through GA-ASI’s Live, Virtual, Constructive (LVC) combat aircraft ecosystem, utilizing a Low Earth Orbit (LEO) satellite communication (SATCOM) datalink provided by an IP-based Mission Beyond Line of Sight (BLOS) solution. The endeavor utilized two L3Harris Technologies RASOR Multi-Functional Processors (MFPs), with one hosting the transceiver card and the other controlling the BLOS Active Electronically Scanned Array (AESA).
Notably, this represents the first deployment of a LEO SATCOM provider connection on an operationally relevant unmanned combat aerial vehicle (UCAV) platform. The demonstration also showcased GA-ASI’s ability to swiftly retrain and redeploy AI pilots during flight using the LEO SATCOM connection. The test aircraft was equipped with a Ball Aerospace BLOS AESA system, enabling full duplex operation. This milestone highlights GA-ASI’s dedication to advancing collaborative combat aircraft capabilities, integrating cutting-edge technologies such as AI pilots, the LVC ecosystem, and L3Harris and Ball Aerospace BLOS datalink solutions.
According to GA-ASI’s Senior Director of Advanced Programs, Michael Atwood, “The flight demonstrated GA-ASI’s unparalleled expertise in autonomous flight operations with real-world, tactically relevant unmanned combat aerial vehicles. It showcased the effective command and control capabilities provided by BLOS collaboration between three defense primes, underscoring the rapid development of our CCA mission system suite and bringing us closer to delivering this groundbreaking capability to the warfighter.”
This achievement was made possible through GA-ASI’s comprehensive CCA ecosystem, integrating third-party capabilities, human-in-the-loop control, and autonomous functionalities to enable efficient human-machine teaming in modern conflicts. Performance data from the AI pilots was collected and transmitted to the ground, where agents were retrained for improved performance and swiftly redeployed via LEO SATCOM. Operator commands, captured via hands-on throttle-and-stick (HOTAS) controls, were transmitted to the AI pilots using Reinforcement Learning (RL) algorithms. The AI pilots autonomously tracked and maneuvered around dynamic entities based on HOTAS specifications. Operators received updates from the AI pilots through a heads-up display in the cockpit and could dynamically reassign tasks via HOTAS as the mission progressed. These developments represent ongoing advancements in technology and autonomous flight, funded through internal research and development, to validate critical concepts for unmanned aircraft systems.
GA-ASI, an affiliate of General Atomics, is renowned for its expertise in designing and manufacturing reliable Remotely Piloted Aircraft (RPA) systems, radars, electro-optic and mission systems. Notable products include the Predator® RPA series and the Lynx® Multi-mode Radar. With over seven million flight hours, GA-ASI offers long-endurance aircraft capable of executing missions with integrated sensor and data link systems, providing persistent flight capabilities for enhanced situational awareness and rapid strike capabilities. The company also delivers various ground control stations, sensor control/image analysis software, pilot training and support services, as well as develops meta-material antennas.