The Trump administration’s newest air-traffic upgrade is about to leave the lab and enter the skies over Washington.
The Federal Aviation Administration is preparing to roll out SMART, an artificial-intelligence system designed to identify congestion and flight conflicts before airplanes leave the gate.
The first limited operation could begin Monday across the Washington region, including Reagan National, Dulles and Baltimore-Washington International.
The company building the system said Saturday that the deployment is already moving ahead of schedule:
It's a new day in American aviation. With the roll out of FDMS and SMART on a record timeline and ahead of schedule, every American's flying experience is about to get better with less delays and more certainty. pic.twitter.com/ISE43jLNLX
— ASI (@AirSpaceIntel) September 19, 2026
Human controllers will continue making the decisions that govern passenger aircraft.
It is a decision-support tool that combines schedules, weather, runway capacity, airspace conditions and operating constraints so human air-traffic managers can see problems earlier and consider better routes.
That distinction is the center of the story.
Transportation Secretary Sean Duffy has said controllers will remain in charge. SMART is supposed to give them better information before congestion becomes a chain of delays, cancellations and missed connections.
The Federal Aviation Administration says SMART stands for Strategic Management of Airspace, Routes, and Trajectories. It will sit on top of a new Flow Management Data and Services platform that becomes the technological backbone of the FAA’s command center.
The current system often reacts after weather, runway closures or traffic volume have already created a bottleneck. Critical information is scattered across screens and spreadsheets, making it harder to develop one shared national picture.
SMART is meant to reverse that order.
It continuously analyzes airline schedules, weather, airport capacity, airspace conditions and operational limits to predict traffic flows and identify conflicts before they occur. The FAA says the combined system can help planners act days, weeks or even months ahead.
That is a major change from waiting until aircraft are stacked on taxiways and passengers are watching departure boards turn red.
The contract with Boston-based Air Space Intelligence is worth $875 million over 12 years. It covers SMART and the larger data platform that will replace aging tools at the Air Traffic Control System Command Center in Virginia.
Ars Technica summarized the scale of the first deployment:
FAA tees up $875M AI tool to help manage air traffic congestion https://t.co/d6vBOf0V5O
— Ars Technica (@arstechnica) September 18, 2026
Ars Technica reports that the Washington trial is the first step toward a nationwide rollout across the 29 million square miles of airspace overseen by the FAA. The software will provide forecasts and suggested alternatives through existing agency systems rather than changing controller procedures.
The report says the first phase could cover the region’s three major airports and begin as soon as September 21. The system is expected to suggest alternate-route information while controllers and airline operators retain their current responsibilities.
That cautious rollout is sensible.
Air traffic is no place for Silicon Valley’s old habit of shipping an unfinished product and fixing it after users complain. The FAA should test the system under real workloads, degraded data and ugly weather before expanding its reach.
There must also be a clear record of who makes each operational decision and who is accountable when a recommendation is wrong.
But caution is not an argument for keeping 20th-century tools forever.
America’s controllers manage one of the safest and most complicated airspaces in the world while dealing with aging radar, old communications hardware and persistent staffing shortages. Giving them an earlier, unified view of traffic is exactly the kind of modernization the federal government should pursue.
Quartz reports that SMART may flag conflicts as far as 45 days in advance and suggest small adjustments to individual flights. The broader Department of Transportation goal is to finish the national air-traffic-system overhaul by the end of 2028.
The system will ingest schedules, weather, runway capacity and other operating limits, then give the FAA and airlines a shared picture for planning routes before disruptions spread. The report also places SMART inside a larger modernization program that includes $12.5 billion already allocated for air-traffic infrastructure and Duffy’s request for another $6 billion to $10 billion for AI integration.
Duffy has been blunt about the human role. AI may help process patterns and surface options, but it will not replace the controller responsible for managing the airspace.
The National Air Traffic Controllers Association has made the same point, describing the technology as support for the workforce rather than competition with it.
The possible Monday start has now reached national coverage:
The Federal Aviation Administration could start using an AI tool to help with air traffic control as soon as Monday, according to a report from The Wall Street Journal. https://t.co/Qut4a92bey
— Scripps News (@scrippsnews) September 19, 2026
Scripps News reports that the FAA could begin using the tool in the Washington area as soon as Monday. Its report emphasizes that SMART will assist air-traffic management rather than take over control of aircraft.
The initial Washington deployment gives the agency a concentrated, high-traffic environment for testing the recommendations before a national expansion. That limited start also gives controllers and airlines time to compare SMART’s suggested routes against real weather, runway limits and traffic patterns without changing the chain of command.
The potential payoff reaches beyond shorter delays.
Better routing can reduce fuel burn, speed recovery after storms and make available airspace easier to use. Airlines, controllers and FAA managers could work from the same picture instead of negotiating around disconnected datasets.
The technology also creates a test of government competence.
An $875 million contract should produce measurable improvements, not a flashy dashboard, a consulting empire or another federal software project that misses its deadline.
The FAA should publish performance standards for forecast accuracy, delay reduction, system reliability and controller workload. The public should know what must happen during the Washington trial before SMART expands nationwide.
President Trump and Duffy have promised a modern air-traffic system by 2028. This week gives that promise a practical test.
If SMART helps controllers see trouble earlier without weakening their authority, it could turn one of air travel’s most frustrating problems from a daily reaction into a manageable forecast.
Monday’s launch is small compared with the national system.
But it is the moment the administration’s $875 million bet stops being a contract and starts becoming a flight plan.
What are your thoughts?
