FAA to test $875M AI traffic tool SMART over Washington DC

FAA to test $875M AI traffic tool SMART over Washington DC

The Federal Aviation Administration is preparing to roll out SMART, an $875 million artificial intelligence system built to advise air traffic controllers on managing congestion in the airspace above Washington, DC. US government and industry officials told The Wall Street Journal the tool could debut for the region's three major airports as soon as Monday, September 21. The Washington rollout is meant as the first step toward a nationwide deployment of SMART across the 29 million square miles of airspace the FAA oversees.

The FAA describes SMART as a system that uses AI models to predict air traffic flows and flag potential conflicts, drawing on operational factors such as airline schedules, weather, airport capacity, and airspace conditions. The aim is to give the FAA, airlines, and aircraft operators a shared, AI-generated view of conditions so they can agree on more efficient routes and departure times. The agency says the expected benefits include cutting aircraft fuel burn, improving on-time performance, and speeding recovery after weather disruptions or congestion events.

Philip Mann, a principal consultant at Vector Strategic Consulting LLC who advises on aviation safety and AI governance and who spent 17 years working at the FAA in multiple roles, called the decision to start with a limited scope before going nationwide the "right call." He argued that SMART's risk was never about any single prediction: as a national-scale system with AI components, it carries more unknowns than anything the FAA has fielded before, and narrowing the initial scope shrinks those unknowns.

Politico reported that the airline industry was confused for weeks about the FAA's plans to deploy SMART. Anonymous airline officials told Politico their concerns eased only after the FAA clarified that the tool would not change any existing procedures for air traffic controllers or airlines. Instead, the FAA said SMART would produce "alternative route information" delivered through the agency's existing systems, rather than a new interface or process that controllers and airlines would need to adopt.

Key facts

  • SMART, an $875 million FAA AI system, could begin advising air traffic controllers over Washington, DC's three major airports as soon as Monday, September 21.
  • The Washington rollout is the first phase before a planned nationwide deployment across the 29 million square miles of US airspace the FAA oversees.
  • SMART predicts air traffic flows and flags potential conflicts using data on airline schedules, weather, airport capacity, and airspace conditions.
  • Philip Mann, who spent 17 years at the FAA and now advises on aviation AI governance, called the phased rollout the 'right call' because a national-scale AI system carries unknowns that a smaller scope helps contain.
  • Politico reported weeks of airline-industry confusion over the plan, which eased once the FAA said SMART would not change controller or airline procedures and would only surface 'alternative route information' through existing systems.

Why it matters

This is a federal safety agency putting an AI system into an operational role over some of the country's busiest airspace, not a company shipping a product. The FAA is treating AI-generated traffic forecasts as a working input for how controllers, airlines, and aircraft operators coordinate routes and departure times, and framing the Washington launch explicitly as a controlled first phase of a much larger, nationwide deployment.

Who it affects

Air traffic controllers, airlines, and aircraft operators in the Washington, DC area are first in line, since SMART is set to cover the region's three major airports. If the nationwide rollout follows, the same advisory system would eventually touch flight planning and routing across all US airspace the FAA oversees.

How to use it

SMART is advisory rather than a new workflow: the FAA says it will not change any procedures for controllers or airlines. Instead, it surfaces 'alternative route information' through the FAA's existing systems, so airlines and controllers see AI-generated forecasts and conflict warnings without adopting a new interface or process.

How solid is it

The account rests on named and unnamed sourcing of different strength: the $875 million figure comes from the source's own headline and is not broken down or explained in the article body, and the September 21 debut date comes from unnamed 'government and industry officials' cited by The Wall Street Journal rather than an FAA announcement. No contractor or developer of SMART is named, and no year is given for the September 21 date.

Risks and caveats

Philip Mann's own framing is a caveat as much as an endorsement: he describes SMART as carrying more unknowns than anything the FAA has fielded, precisely because it is a national-scale system with AI components, and treats the narrow initial scope as a way to manage that risk rather than eliminate it. Politico's reporting adds a separate risk already visible before launch: weeks of airline-industry confusion over the FAA's plans, which the agency had to defuse by publicly ruling out procedural changes.

“SMART's risk was never any single prediction, it is a national-scale system with AI components carrying more unknowns than anything the FAA has fielded. Every cut in scope shrinks the unknowns.”

— Philip Mann, principal consultant at Vector Strategic Consulting LLC, formerly of the FAA