Overview: On September 21, the U.S. Federal Aviation Administration (FAA) officially launched SMART, an AI-powered flight management tool, at three airports in the Washington, D.C. area. According to a statement released on the FAA’s website, U.S. Transportation Secretary Sean P. Duffy and FAA Administrator Bryan Bedford jointly unveiled the tool that day. Aerotime reported that the initial deployment covers Ronald Reagan Washington National Airport (DCA), Washington Dulles International Airport (IAD), and Baltimore/Washington International Thurgood Marshall Airport (BWI), with plans to expand to other regions nationwide.
According to the FAA website, SMART, which stands for "Strategic Management of Airspace, Routes and Trajectories," integrates 200 different data streams, including weather patterns, flight paths, traffic flow, and air traffic controller schedules, into a single platform. Its AI engine analyzes the data and presents information on the destination of aircraft, the scale of traffic the system can handle, and areas that may experience congestion or be affected by weather.
According to reports from WPDE and other media outlets, the data streams fed into the system also include airline flight plans, weather forecasts, flight paths, airport capacity, traffic flow, and airspace restrictions. In other words, data that was previously scattered across separate systems used by meteorological services, air traffic management, airports, and airlines has now been consolidated into a single, unified platform, allowing air traffic controllers and traffic management personnel to view all the information together, rather than relying on individual, isolated displays.
According to Aerotime, the core capability of this software is to analyze airline flight schedules, weather conditions, airport capacity, and airspace restrictions. It identifies locations where traffic demand may exceed available capacity and continuously updates predictions as conditions change. This provides FAA management and individual airlines with a unified "shared view" to facilitate proactive adjustments, predict delays, and reduce cancellations.
Unlike passively responding to traffic congestion, these tools aim to shift the focus to proactive management by addressing potential issues before they arise: first, identify "where congestion is likely to occur," and then decide whether to implement flow control, adjust flight routes, or stagger departures. According to Aerotime, the system will continuously update its predictions based on changing weather and traffic conditions, rather than providing a fixed plan once a day. For passengers, the primary goal is to reduce long waiting times before takeoff and minimize last-minute cancellations.
According to Aerotime, this deployment is taking place following the FAA's award of a 12-year, $875 million contract to Air Space Intelligence in June 2026. Additionally, according to aviation media reports, SMART will initially operate in the Washington D.C. area on a limited trial basis of approximately 90 days, rather than a full-scale takeover. According to an industry-circulated announcement, this is a "limited operational use," with the initial scope being the aforementioned three airports.
According to a report by Politico, this move integrates AI tools into the real-time operational processes of the U.S. air traffic control system. Transportation Secretary Buttigieg stated on the day of the announcement that the new system will not replace human air traffic controllers. This positioning implies that the predictions and recommendations generated by SMART will still need to be reviewed and implemented by air traffic controllers and traffic management personnel, and the responsibility will ultimately remain with humans.
The U.S. air traffic control system has long been plagued by increasing traffic and delays. SMART's choice of a "decision support, human-in-the-loop" approach is a relatively safe starting point. However, transitioning from limited trials at three airports to nationwide implementation requires overcoming several hurdles, including verifying effectiveness, gaining acceptance from airlines, and building public trust. As of this writing, the FAA has not yet announced a specific timeline for nationwide implementation. The 12-year, $875 million contract also serves as a reminder to the industry that AI in air traffic control is not a lightweight pilot project, but rather a long-term infrastructure investment. Its success will be measured by actual improvements in delay and cancellation data, not by immediate results showcased at a press conference.