Trust and Training: The Key to Successful Human-AI Collaborative Combat Aircraft Systems

Last week, General Atomics Aeronautical Systems (GA-ASI), an American military aeronautics contractor, successfully executed tactical flying maneuvers with its MQ-20 Avenger drone using artificial intelligence (AI) and Low Earth Orbit (LEO) satellite communication (SATCOM).

In a press release statement, GA-ASI said that it conducted the flight test using its “live, virtual, constructive collaborative combat aircraft ecosystem over a [LEO SATCOM] provider’s IP-based Mission Beyond Line of Sight (BLOS) datalink” on April 6, 2023.

The company demonstrated joint capability between AI and human pilots in controlling the MQ-20 unmanned aircraft system (UAS). Likewise, it highlighted the connection that allowed machine learning to rapidly “retain and redeploy” as it executes combat maneuvers in real time.

BREAKING: GA-ASI's AI-piloted #AvengerUAS conducts live air combat maneuvers via Low Earth Orbit satellite communications.

​This marks the first use of a LEO satellite provider‘s connection on a tactical unmanned aircraft. pic.twitter.com/iLRzmk7JFB

— GA-ASI (@GenAtomics_ASI) April 11, 2023

The trial also marked the first LEO SATCOM provider link deployment on an operationally relevant combat UAS platform.

“The flight demonstrated GA-ASI’s unmatched ability to fly autonomy on real, tactically relevant, unmanned combat aerial vehicles,” GA-ASI Senior Director of Advanced Programs Michael Atwood stated. It displayed effective BLOS Command and Control through the collaboration between three defense primes. This showcases our rapidly maturing [Collaborative Combat Aircraft] mission system suite and moves us one step closer to providing this revolutionary capability to the warfighter.”

During the flight test, the human pilot would input its command through LEO SATCOM, in which the autonomous AI pilot within the MQ-20 system would then adapt through “reinforcement learning algorithms.”

In addition, the programmed AI pilot can autonomously track and maneuver around its set route, which instantly changes once updated via operator commands on hands-on throttle-and-stick.

Earlier this year, GA-ASI performed a preliminary flight test to validate the aircraft’s Collaborative Combat Aircraft mission system, pairing it with an MQ-20 “digital twin” to assess the system’s AI and machine learning capability in live, virtual, and constructive multi-objective missions.

Driven by Artificial Intelligence

The MQ-20 Avenger is the US armed forces’ next-generation, medium-to-high altitude long-endurance (M/HALE) unmanned combat aerial vehicle (UCAV) platform built by GA-ASI in the late 2000s.

Also called Predator C, the tactical drone is basically an upgrade to the renowned MQ-1 Predator and MQ-9 Reaper (Predator B) powered by a turbofan engine instead of the usual propeller system.

Its design is also unique from its predecessors, featuring stealth components—such as internal weapons storage and an S-shaped exhaust that allows the aircraft to have reduced infrared and radar signatures—and better speed performance thanks to its upgraded engine. Besides that, the MQ-20 uses the same weapons and ground support infrastructures as the MQ-1 and MQ-9. Armaments aboard Predator C include Hellfire missiles, bombs, guidance tail kits, precision-guided glide bombs, and laser-guided explosives.

Moreover, it carries the “Lynx” Synthetic Aperture Radar (SAR) system and a GA-ASI-built Advanced Low-observable Embedded Reconnaissance Targeting (ALERT) system, a version similar to the fitted F-35’s electro-optical targeting systems.

By the 2010s, GA-ASI developed two more variants: one sea-based for the Navy (Sea Avenger) and an Extended Range version, all potentially able to fit with AI pilot features—possibly with the human-machine collaborative system through LEO SATCOM in the future.

The MQ-20, while it is not inducted in service, continues to be operationally ready, with progress on autonomous human-machine maneuvers as a promising feature.

Potential Future Wingmen

The Pentagon previously mentioned its plans to integrate advanced AI-controlled UAV platforms to support crewed fighter jets.

Drones are loyal machines that willingly do tasks that human pilots usually wouldn’t; they are very replaceable if lost in combat and could lessen workforce casualties. When done right, they can be programmed to be faster, more precise, and autonomously do their own tasks and decision-making via machine learning algorithms while Airmen focus on top-priority missions.

Like any other partnership, trust should be first and foremost established between the AI and human pilots, thus the need for extensive training.

While AI systems offer many advantages, human judgment remains something to consider. This quality cannot be simply processed by powerful machine calculations and algorithms with thousands of stored scenarios.

Dr. Mark J. Lewis, the Acting Deputy Under Secretary of Defense for Research and Engineering, explained during a virtual defense conference in 2020 that AI drones will not necessarily replace all human fighter pilots. Instead, he saw these high-tech machines as an effective support system.

“The key takeaway was the artificial intelligence system did so well because it wasn’t so concerned about self-preservation. It was willing to do things that a human pilot wouldn’t do and that’s the advantage of artificial intelligence,” said Lewis. Adding, “… I think the real answer is teaming AI with the human for the combination of both. I’m pretty confident we’re going to have human pilots into the future.”

To learn more about the future of air combat using AI UAVs, you’ll want to check out the video below.