AI for Business

AI in Aviation: Promise, Pilots, and a Groundstop on Friday the 13th

On September 27, US transportation secretary Sean Duffy posted a video on X with the words "GAME CHANGER." "Introducing 'SMART' - the new air traffic management tool that saves YOU time, cuts back...

Share:
AI in Aviation: Promise, Pilots, and a Groundstop on Friday the 13th

On September 27, US transportation secretary Sean Duffy posted a video on X with the words "GAME CHANGER." "Introducing 'SMART' - the new air traffic management tool that saves YOU time, cuts back DELAYS, and makes your flying experience BETTER," he wrote. The AI-infused air traffic control system was unveiled with fanfare. Less than three weeks later, that post took on a darker life as the FAA issued a national groundstop on Friday the 13th that threw the world's airline schedules into chaos.

The administration quickly announced there was no evidence of a cyberattack and no sign that SMART had anything to do with the communications failure behind the decision to suspend flights. But when asked that Friday night whether he could definitively rule out AI failure, Homeland Security secretary Markwayne Mullin could only say, "There's nothing certain right now."

SMART is only the tip of the iceberg in an AI aviation revolution. Consider the Cessna Caravan that took off from Buchanan Field Airport in Concord, California, and flew south to Phoenix on a calm September morning. The single-engine turboprop then turned east to Fort Worth, stopped at Shaw Air Force Base in South Carolina, and finally arrived at Dare County Regional Airport off the coast of North Carolina. The zigzag, coast-to-coast trip covered 3,200 miles, navigating busy airspace and maneuvering around storms. None of it would be remarkable except for one fact: the entire journey took place without a single human hand touching the plane's controls.

According to Joby Aviation, which carried out the project, a safety pilot was aboard at all times but never needed to intervene. The pilotless trip marked a major milestone. Air travel has long been technologically conservative, and with good reason: when human lives are at stake, it's better to be time-tested and a little dated than cutting-edge and unpredictable. But AI is making inroads with unprecedented speed, promising to remake the economics of air travel and transform how people and things move through the sky. A big question remains: can we trust this technology to be reliable and safe?

"We're introducing all this new technology to improve the system, and it's going to be a disaster," says a recently retired FAA employee, who asked to remain anonymous. "This is literally a Pandora's box that we are going to regret."

Aviation's embrace of AI traces to the early 2000s, when advances in batteries and microelectronics made drones far more capable and inexpensive. The surge convinced officials that robots and humans would eventually need to share the airways. In 2007, the federal government rolled out NextGen, a sweeping project mandating broad changes to air travel, including a requirement that by 2020 aircraft in the United States be equipped with ADS-B, which constantly sends radio signals telling everyone in the vicinity where you are and where you're headed.

With a road map to robot-friendly skies in place, Amazon and Walmart began working on drone delivery. In 2014, Google unveiled Project Wing, which by 2018 was ferrying food, coffee, and medication to 164 homes near Canberra, Australia; the next year the pilot program expanded to Christiansburg, Virginia. Around the same time, startups began work on electric vertical takeoff and landing aircraft, or eVTOLs - essentially oversized drones designed to carry a small number of people a short distance. By 2022 there were 600 models under development. Joby unveiled its S4 design in 2020, saying it could carry a pilot and four passengers 100 miles or more at up to 200 mph; its maiden flight took place last year.

A commercial launch seemed just around the corner, but deadlines kept slipping. Engineering a flying autonomous robot was only part of the problem. You also needed an airspace they could operate in, trained humans to supervise them, and an economic framework. "It's a very important, safe, careful dance," says Bowles.

"Aviation regulation has been conservative and prioritizes reliability, because the consequences of failure are severe," says Mary Schiavo, an aviation lawyer and former inspector general in the Department of Transportation. "Certification standards and airworthiness require proof and demonstrated safety, and that usually means incremental rather than radical change."

AI faced another kind of resistance in Europe. In 2017, Airbus and Dassault asked EASA, the European Union's version of the FAA, to consider allowing airliners to operate with a single pilot in the cockpit, with automated systems shouldering the copilot's workload. The European Cockpit Association, a union of airline pilots, lodged a strong protest, saying such a move "would remove an independent safety professional from the flight deck and give operators greater influence over the safety judgments that a pilot would otherwise make." Last year, EASA shelved the plan.

Nevertheless, AI pressed onward. In 2023, the FAA began issuing waivers to allow delivery-drone flights beyond the visual range of operators; by the end of this year it could make the rule change permanent. In August, Amazon announced plans to expand drone delivery to 500 locations across the country by the end of the year, and Walmart is planning drone delivery from 270 store locations by the end of 2027. "In five to 10 years, it's going to be completely ubiquitous," Anthony Vittone, chief operating officer of the drone technology company DroneUp, told Sherwood News last year. "People are going to be, like, 'I don't know how I lived without it.'"

Joby has moved ahead too. In April, it flew the first-ever point-to-point eVTOL flight within New York City, from John F. Kennedy Airport to three heliports in Manhattan, and has carried out similar flights in California and Texas under the eVTOL Integration Pilot Program, a federal initiative rolled out by the Trump administration this past March. Bowles says these demo flights are just the "front edge" of the S4's operational rollout, to be followed by flights with nonpaying passengers, then paying passengers, then point-to-point service. "All of that is permitted under eIPP, with increasing levels of oversight," he says.

To start, these flights will have human pilots, but paying one person to fly a handful of passengers is expensive; for the model to make economic sense, the aircraft will have to fly themselves. And if self-driving is hard for cars, it's harder for aircraft, where an unexpected control system failure isn't just dangerous but likely deadly.

Given the scale of the challenge, AI's entry into human passenger flight won't be as flashy as piloting eVTOLs; it will occupy a more modest, behind-the-scenes role. Data companies like Palantir and aviation stalwarts like Jeppesen and Thales already sell AI-enabled software that lets regulators, airlines, and flight dispatchers sift through masses of data to identify risks and plan scheduling.

In September, the FAA announced it had deployed software built by the Boston-based data company Air Space Intelligence. The tool, "Strategic Management of Airspace, Routes and Trajectories," or SMART, is aimed at "fundamentally reshaping how we manage our airspace." The program isn't intended to replace controllers, the agency says, but to make their jobs easier by using predictive AI to head off scheduling snags. "Combining SMART with our ongoing hardware upgrades gives us the tools to cut flight delays, optimize traffic, and keep airspace capacity across the country open - all while maintaining our gold standard of safety," FAA administrator Bryan Bedford said in a press release.

The system is supposed to help controllers make decisions, not make them for them. But if the program feeds nonsensical information into the system, will human operators catch the mistake? "This is the Trump/DOGE/Musk effect," says the retired FAA employee. "They're pushing through all this new technology without any understanding of the complexity of the national airspace system."

Some lawmakers share those concerns. On October 13, almost two dozen Democratic lawmakers reportedly sent a letter to the Government Accountability Office asking it to review the program, citing fears that it might "hallucinate safety-critical information."

Then there's the question of whether SMART could worsen the chronic shortage of air traffic controllers. The FAA hopes the technology eases the shortage by requiring fewer workers, though other industries have yet to see AI-generated efficiencies actually materialize. Even if the software doesn't directly replace humans, the efficiency could be used to justify workforce reductions. In May, the FAA reduced its staffing target for active air traffic controllers from the 2024 goal of 14,633 down to 12,563, citing "modern staffing models and scheduling tools."

Some are more optimistic. "I think the SMART program is an important improvement," Schiavo says. "No one wants the current aviation environment with midairs, runway incursions, myriad near misses, delays, groundings, loss of separation, and long taxi or clearance wait times." Michael McCormick, a professor of air traffic management at Embry-Riddle Aeronautical University, sees advantages to a predictive tool. "When you have the computing capability to look at the entire trajectory of the aircraft, you can see where the conflicts are going to be," he says. "And the general axiom in air traffic control is that the sooner you know about a potential conflict, the less effort is required to resolve it."

That sounds great, but building machines that do a good job most of the time is one thing; machines that can function at the level of reliability aviation requires is another. Researchers are focusing on so-called deterministic neural networks - models trained offline whose weights are frozen before deployment. These can't hallucinate, but they may harbor other risky failure modes. In a 2021 paper, Darren Cofer, a researcher at Collins Aerospace, pointed out that when such systems encounter situations they haven't seen in training, they can generate "unexpected (and potentially dangerous) outputs."

The stakes will be higher still when AI software starts making real-time air traffic control decisions, which will begin in airspace devoted to autonomous aircraft - most likely select corridors near drone hubs and eVTOL vertiports. With swarms of robot aircraft zipping back and forth, things will be too hairy for human pilots and controllers. "The objective is not to use traditional air traffic control," McCormick says, "because it would overwhelm the national airspace system if it tried to handle that volume and complexity of airframes moving around."

With two systems working side by side - one with human decisionmakers, the other with AI - we'll have a real-time basis to compare both approaches. Any computer-based system can potentially be hacked; any AI system creative enough to ideate novel solutions could also hallucinate crucial details. Without a human in the loop, seemingly obvious mistakes might go unnoticed until it's too late.

Source: Wired

Ready to Modernize Your Business?

Get your AI automation roadmap in minutes, not months.

Analyze Your Workflows →