FAA Prepares AI Route-Advice Test Around Washington Airports
SMART will enter existing FAA systems as route advice rather than new instructions, limiting its first operational role while leaving its underlying AI models undisclosed.
Listen to this story
The audio brief
Story brief
3 key pointsSMART is moving from procurement into a constrained operational test around Washington’s three major airports, with a possible September 21 start. The FAA will use it only for route advice through existing systems, not new instructions for controllers or airlines. The pilot puts part of Air Space Intelligence’s $875 million, 12-year award under real operating pressure while targeting a jump from roughly 15 minutes of conflict planning to two hours.
- 01
The Washington test could start September 21, but SMART’s forecasts will not direct operations.
- 02
SMART targets two hours of conflict planning versus roughly 15 minutes today; this remains a design goal, not live evidence.
- 03
The award also replaces the FAA’s Flow Management Data and Services system at its Virginia command center.
The Federal Aviation Administration is preparing to deploy an AI system that will advise on congestion and possible conflicts around Washington’s three major airports, potentially beginning September 21. The limited rollout brings the FAA’s $875 million, 12-year contract with Air Space Intelligence into an operational setting.
The system, called SMART, is designed to predict traffic flows and flag potential conflicts using airline schedules, weather, airport capacity and airspace conditions. Its forecasts are meant to give the FAA, airlines and aircraft operators a shared view when weighing routes and departure times.
At first, SMART will provide alternative route information through existing FAA systems. The FAA says the deployment will not change procedures for air traffic controllers or airlines, placing the system in an advisory role rather than making it a new source of operating instructions.
SMART is only one part of the award. The contract also covers Flow Management Data and Services, intended to replace the current system at the FAA’s Air Traffic Control System Command Center in Virginia. Together, the projects pair a replacement for a core flow-management system with a predictive tool that draws on operating data.
The FAA expects the Washington-area deployment to be the first step toward a nationwide rollout across the 29 million square miles of US airspace it oversees. It is introducing the system while managing an ongoing controller shortage and a broader effort to replace aging air-traffic-control equipment.
The FAA has not disclosed what types of AI models SMART uses. Philip Mann, an aviation safety and AI governance consultant who previously worked at the FAA, supported the narrower launch but said he will be watching for evidence of performance under real workload and degraded data, as well as clarity on responsibility when a prediction is wrong.
Editorial analysis
Our Read
The important design choice is not simply that the FAA is adding AI. It is separating prediction from operational authority in SMART’s first use. Route recommendations can shape decisions without changing the formal procedures of controllers or airlines. That gives the agency a narrower starting point as it modernizes aging equipment amid a controller shortage. The next meaningful public signal is whether the FAA defines the operational evidence and accountability arrangements that would support a wider rollout, especially when its model types remain undisclosed.
Sources
- arstechnica.comFAA tees up $875M AI tool to help manage air traffic congestion
- startupfortune.comThe FAA Bet $875 Million on an AI Startup to Fix Air Traffic Control - Startup Fortune
Reader comments
Newest comments first. Replies stay oldest first.