In a scene reminiscent of Star Wars fame, the US Air Force just flew an artificial intelligence system acting as a pilot’s sidekick on the U2 Dragonlady reconnaissance aircraft for the first time in history, Military.com reported.
The AI algorithm, call sign ARTUµ (pronounced R-2, in homage to beloved “Star Wars” droid R2-D2), flew alongside a pilot, identified only as Major “Vudu,” from the 9th Reconnaissance Wing, Beale Air Force Base, California, on Dec. 15, the service said in an announcement Wednesday.
ARTUµ took charge of the sensor controls right after takeoff; the pilot and the AI shared data to execute their operation, Military.com reported.
During an exercise that simulated a missile strike, ARTUµ controlled various sensors and navigation while the pilot flew the aircraft, according to a release. While Vudu watched for hostile aircraft, the AI scanned for enemy launch weapons.
“ARTUµ’s groundbreaking flight culminates our three-year journey to becoming a digital force,” Dr. Will Roper, assistant secretary of the Air Force for acquisition, technology and logistics, said in the release.
“Putting AI safely in command of a US military system for the first time ushers in a new age of human-machine teaming and algorithmic competition. Failing to realize AI’s full potential will mean ceding decision advantage to our adversaries.”
ARTUµ’s training included a culmination of “half-million computer-simulated training iterations,” the release said. Researchers from Air Combat Command’s U-2 Federal Laboratory created and trained the AI, Military.com reported.
“We know that in order to fight and win in a future conflict with a peer adversary, we must have a decisive digital advantage,” said Chief of Staff Gen. Charles “CQ” Brown.
“AI will play a critical role in achieving that edge, so I’m incredibly proud of what the team accomplished. We must accelerate change and that only happens when our airmen push the limits of what we thought was possible,” he said.
The Air Force has been experimenting with AI systems that can train with its pilots.
Another leap-ahead initiative is the Air Force’s Skyborg program, which will develop AI-enabled drones to fly ahead of fighter jets and respond to incoming threats before they reach the human pilot, Military.com reported.
Use of AI within the Cold War-era surveillance plane — which has been flying since the early 1960s — was enabled by another software management system called Kubernetes, developed by Google, to bridge the systems.
In October, the U2 Federal Lab team updated the aircraft’s software for the first time while in flight using Kubernetes — another milestone for the Air Force, Roper said at the time.
Officials on Wednesday said the AI was made to be “easily transferable to other systems” with its open-source architecture design.
In a tweet Wednesday, Roper said that the Air Force modified the “world-leading μZero gaming algorithms to operate the U2’s radar,” Military.com reported.
The service adapted the μZero gaming program, an online algorithm that has outsmarted human players in games like Chess and Go, into the AI’s operation.
“Like any pilot, Artuμ … has strengths and weaknesses,” Roper said in the tweet. “Understanding them to prep both humans and AI for a new era of algorithmic warfare is our next imperative step. We either become sci-fi or become history.”
“I think it’s fantastic,” ‘Hawk’ Carlisle, head of the National Defense Industrial Association and a retired general who ran Air Combat Command from 2014-2017, told Breaking Defense.
“I think most everybody sees it as a continuation the evolution that’s occurred so far” in fourth and fifth generation fighters — an evolution “that makes that person in the loop, that young woman, or young man, that much more effective, that much more capable.”
“I think there’s a lot of work still to do,” added Carlisle, who has been following the work at Beale. “I think technologically we’ll get there … When you think about that, there’s things that are uniquely qualified for AI. For flying an instrument approach, having an AI copilot makes sense.”
That said, he noted that the real question will be “really understanding where a person in the loop needs to be. There’s things machines can do faster than the human mind,” but “you have to have the appropriate intersection where the people are part of the decision process.”
Teal Group’s Richard Aboulafia said in an email to Breaking Defense that “AI has always had potential to improve targeting and force application in full-up shooting wars. Full-up shooting wars, thankfully, are quite rare.”
For at least the past two decades, ground-centric counterinsurgency fights, no-fly zone enforcement, maritime patrol, and other proxy conflicts have defined the American way of war, all missions where “AI is of very limited utility,” Aboulafia said.