FAA launches AI traffic tool at DC-area airports, with nationwide rollout to follow

The Federal Aviation Administration is deploying an artificial intelligence tool to help manage air traffic at the three major airports serving Washington, D.C., as soon as Monday, September 21 — the first operational use of a system the agency eventually plans to extend across all 29 million square miles of U.S. national airspace.

Covering Ronald Reagan Washington National (DCA), Washington Dulles International (IAD), and Baltimore/Washington International Thurgood Marshall (BWI), the tool bears the name SMART — Strategic Management of Airspace, Routes, and Trajectories. Rather than reacting to congestion after it forms, the system works ahead of it: drawing on airline schedules, weather data, airport capacity, and airspace conditions, it projects traffic flows, spots potential conflicts early, and delivers alternative routing and planning options through the same channels air traffic managers and airline dispatchers already rely on.

Crucially, SMART does not issue clearances, change controller procedures, or automate separation decisions. The FAA told the airline industry that the system would produce 'alternative route information' made available through existing FAA systems — a framing that, according to Politico, eased weeks of confusion and concern among airline officials about what the tool would actually require of them.

Air Space Intelligence, a Boston-based company, won a 12-year, $875 million contract in June that encompasses SMART. The contract also covers a broader system called Flow Management Data and Services (FMDS), intended to replace the FAA's legacy Traffic Flow Management System and serve as the new backbone for the Air Traffic Control System Command Center in Warrenton, Virginia. Philip Mann, a former FAA official who spent 17 years at the agency and now advises on aviation safety and AI governance, has described FMDS as the 'backbone' and SMART as the 'predictive layer above it.'

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, Vector Strategic Consulting LLC

Mann called the decision to launch in limited scope before going nationwide the 'right call.' But he said he will be watching for what evidence gates the next expansion — specifically, whether the system performs under real workload and degraded data rather than demonstration conditions, and whether there is a written answer to who is accountable when a prediction is wrong.

What I will be watching is what evidence gates the next expansion: performance under real workload and degraded data rather than demonstration conditions, and a written answer to who owns the outcome when a prediction is wrong.— Philip Mann, Principal Consultant, Vector Strategic Consulting LLC

Washington's airspace is a deliberately demanding proving ground. The region combines dense commercial schedules, complicated coordination between civilian and government helicopter operations, and corridors that have been under heightened scrutiny since a midair collision near Reagan National on January 29, 2025. Separately, the FAA has said that in February 2025, following that collision, it turned to AI tools to pinpoint hotspots where high volumes of helicopter and airplane traffic intermingle.

For travelers, the practical goal of SMART is to reduce the kind of cascading disruptions — triggered by thunderstorms, low visibility, or runway configuration changes — that turn a localized constraint into a regionwide backlog. By spotting bottlenecks earlier, the system is designed to help traffic managers and airline operations centers coordinate reroutes and revised departure timing before delays compound.

As the FAA pursues a multi-billion-dollar push to retire aging infrastructure — among it hundreds of radar systems that date to the 1980s, plus radios, telecommunication lines, and the voice switches linking controllers and pilots — the SMART rollout is taking shape. Compounding that challenge is a shortage of air traffic controllers stretching back decades, a situation Mann characterizes as a 'permanent crisis' that traces its origins to President Ronald Reagan's 1981 dismissal of more than 11,000 controllers. According to a U.S. Government Accountability Office report released in December 2025, the controller workforce shrank by 6 percent over the prior ten years, a decline the report attributed in part to government shutdowns in 2013 and 2018–2019 and to pandemic-era limits on training. In early 2025, hundreds of probationary FAA hires were laid off as part of cuts by Elon Musk's Department of Government Efficiency, though air traffic controllers themselves were spared.

Through partnerships with carriers including Alaska Airlines, Air Space Intelligence says its separate Flyways AI platform — built around a four-dimensional digital twin of national airspace — already plays a role in managing more than 40 percent of all U.S. air traffic by predicting traffic and weather conditions. It is not yet clear what types of AI models underpin SMART itself — whether they are older deterministic models that follow predefined rules, machine-learning models that identify statistical patterns, or something else. Mann has also flagged uncertainty about whether the initial rollout remains limited to aircraft flying at 24,000 feet and above, as the FAA described in June.

How Washington performs will likely shape the pace of expansion to other metro areas. With the FAA positioning FMDS as the future backbone for the Command Center and targeting a nationwide rollout by 2028, the technology's lasting effect will hinge on how smoothly it functions across a national network in which weather disruptions and congestion routinely spill beyond any single airport.

Why it matters — SMART is the first AI system the FAA has deployed to actively advise on national air traffic management, and how it performs at Washington's congested airports will determine how quickly — and whether — the technology reshapes flight scheduling and delay management across the entire country.

Reported by nytimes.com, washingtontimes.com, arstechnica.com, thetraveler.org