An artificial intelligence project launched within the U.S. aviation security agency aims to ease the burden on the thousands of air traffic controllers who guide planes through the air, companies involved in the fledgling effort told POLITICO.
The plan, led by Federal Aviation Administration chief Bryan Bedford, envisions a major overhaul of the nation’s increasingly complex airspace. But it will not want to replace human controllers in making the second-by-second decisions needed to keep aviation safe, two of the project’s three vendors said.
Instead, the goal of the project is to reduce flight delays and simplify the tasks of controllers by making better use of information such as airline scheduling data to reduce congestion before it happens, according to aerospace technology company Thales and software company Air Space Intelligence. Palantir, a third-party technology organization involved in the effort, declined to comment.
The AI-driven program is called Strategic Airway Management, or SMART.
“To be absolutely clear, SMART is not intended to separate aircraft or perform any type of safety-critical function,” said Todd Donovan, Thales’ vice president for space mobility solutions for the Americas.
“It’s really about preparing the demand in the air, the demand at the airport, so that we don’t cause congestion unexpectedly. We’re trying to deal with it proactively,” Donovan said. However, he also said SMART can prevent “two planes from being in conflict.”
What that looks like in practice remains unclear, including how the technology would fit into the FAA’s existing mesh of computer systems. It could be months before an answer is available as Thales, Air Space Intelligence and Palantir, which the FAA invited to participate in the program, compete to lead the project.
The agency said it plans to award the contract “soon,” adding that SMART will “predict flight flow and adjust departure times to resolve conflicts.” The National Association of Air Traffic Controllers, which represents about 11,000 fully certified FAA controllers, accepted a request for comment but did not release a statement.
The AI effort comes as Transportation Secretary Sean Duffy ramps up a multibillion-dollar effort to improve the aging technology that the administration uses before President Donald Trump’s second term ends. NATCA has supported this main goal.
Efforts continue one year aftera fatal helicopter crashnear Ronald Reagan National Airport — the worst air disaster in nearly a quarter century — revealedincreased stresson the air security system burdened by the United States.
In the recent pastinterview with CBS NewsDuffy rejected the idea of replacing controls with automated technology.
“Should I replace the controller and have an AI manage the space?” He said. “The answer to that is hell no, that’s not going to happen.”
Big picture
The main task of the controllers is to separate the aircraft physically, while the FAAcommand centerout of Washington overseesbalance of space requirements and capacity to a greater extent. This last strategy aims to reduce the issues that canripple in the atmospheric systemsuch as storms, staff shortages or too many planes arriving at the airport at the same time.
Take, for example, the severe weather that is unfolding across the Southeast. Controllers in the FAA’s indoor building can direct pilots to wait in formation away from the thunderstorm until it passes over the airport.
But hours before that, with possible disruptions across the region, the FAA could limit the number of planes that can fly across the region. This type of planning seems to be what SMART focuses on – versus the kind of spur-of-the-moment, quick-thinking decisions that controllers often make, such as, for example, redirecting an approaching plane from an airport when a helicopter is also flying nearby.
“The idea is that you start months in advance, looking at the (carrier’s) schedule data, and as you get closer, you start having a weather forecast. … ‘Tomorrow, there’s supposed to be a storm coming into this area, so you’re going to have less visibility at (Ronald Reagan National Airport), and you might as a result reduce your capacity,'” Donovan said, referring to the number of planes. “How do we think about it a day in advance and work with the airlines to say, ‘OK, we can vary things a little bit, we can anticipate some of it’ so that … together we’re not just reacting to something that happens, we’re actually planning for it and trying to work things out.”
Phillip Buckendorf, CEO of Air Space Intelligence, said that “weeks and months” in advance, “you want to predict the flight paths based on the schedules that are there,” then more information when it comes to the day of travel, “how can you adjust everything basically, flight through AI,” to “improve” the air. Alaska Airlineshe has entered into a contract with his companyfor an AI platform to help shippers and “improve the forecasting and flow” of traffic.
Making a change ‘up’
But Donovan said SMART could also prevent at least some of the cases where regulators need to intervene to make sure planes stay safely away.
The project, he added, is an invitation-only, “challenge-based” competition where the FAA evaluates what Thales, Air Space Intelligence and Palantir come up with.
“What happens is strategically before the flight, or before the plane gets to the air traffic controller, small adjustments are made upstream. And because of that, instead of two planes being in conflict,” Donovan said, they will pass each other at a reasonable distance.
“‘What if we slow down the flight 30 minutes earlier for a little while?’ … So the controller now sees the traffic, looks at it,” and there is “no problem,” he said. “The work does not change, but the idea is that the load should be less.”
Buckendorf said SMART is focused on traffic flow, but there will be “a lot less stress on the tactical side” – resulting in increased efficiency and safety.
Weatherfirst reported the deal.
Each of the companies has a lab at FAA headquarters in Washington, and Bedford stops by to check on the work, Donovan said.
During a press event in Washington last month, Bedford compared the nation’s airspace to “Los Angeles gridlock,” adding that every morning is “fraught with conflict and delays and potential cancellations.”
Duffy in a separate meeting in April, without mentioning SMART by name, said the three companies are working with federal officials on “developing a program to look at … how flights are managed.”
“Now we can use AI,” he said.
The initial “proof of concept” phase is “nearing completion,” Donovan said, and Bedford is targeting September for the start of operational demonstrations, with “validation and confidence-building” to follow throughout 2026.
Whichever company wins, how the FAA will award the project contract is uncertain.
“It’s not a plan that, as far as I know, has a budget item,” Donovan said, with the FAA looking for money “to pay for what they’re doing now.”
