FAA Expands AI Air-Traffic Advice Nationwide as Controllers Raise Concerns

The Federal Aviation Administration is expanding the use of an artificial-intelligence system designed to advise air-traffic managers across the United States. SMART, previously deployed in airspace around Washington, DC, will now be used across all 22 of the country’s air-route traffic control centers.

The rollout is not an attempt to hand flight control to an algorithm. Instead, SMART is intended to help people compare routing and scheduling options when the national airspace system becomes congested. The FAA says the system cannot replace controllers or take control of an aircraft, and local controllers can choose whether to accept its recommendations.

What SMART is designed to do

According to the FAA, SMART is a cloud-based platform that combines data on weather patterns, flight paths, traffic flows, and controller staffing. It uses those inputs to forecast traffic conditions and suggest scheduling changes, preferred routes, and arrival sequences.

The recommendations also pass through a human review step. A team of FAA specialists in Washington, DC, validates the system’s output before passing relevant advice to facilities around the country. That arrangement makes SMART a decision-support system rather than an autonomous air-traffic controller. It also keeps operational authority with trained personnel on the ground.

The case for a national rollout

The FAA says SMART has helped reroute 180 flights since its September 21 deployment in the National Capital Region. On average, the agency reports, those flights saved about 10 minutes each and nearly 80 gallons of fuel per flight. The system was later expanded to air-route centers serving Atlanta, Jacksonville, and Miami.

The agency also gave The Dallas Morning News an example involving a cargo flight arriving at Dallas-Fort Worth International Airport. SMART reportedly recommended a shorter and faster route, although the flight was not identified and the FAA has not provided comparable information about its effect on passenger traffic.

Airline executives have welcomed the expansion. For carriers, even modest reductions in airborne time can improve schedule reliability and reduce operating costs. But results from a limited regional deployment do not automatically establish that the system will perform consistently across different weather conditions, traffic patterns, staffing levels, and local procedures.

The missing perspective

The National Air Traffic Controllers Association, which represents more than 15,000 members, told Flying that it was not consulted or involved in SMART’s design, testing, or implementation. As a result, the union said it could not assess the tool’s effects on the air-traffic-control workforce.

That omission matters because AI advice can change how controllers prioritize tasks, document decisions, and respond when a recommendation conflicts with operational reality. The FAA is deploying SMART while also dealing with controller shortages and other long-running infrastructure challenges. A forecasting tool may help specialists evaluate options more quickly, but it cannot remove uncertainty from the underlying data or replace frontline judgment during severe weather, communications problems, or unusual traffic conflicts.

A credible nationwide deployment will require more than headline savings. The FAA should make clear how often recommendations are accepted or rejected, how the system is evaluated for safety, and how responsibility is assigned when its advice is wrong. Controller participation will also be important for testing whether the tool improves operations without creating new burdens.

SMART illustrates a cautious model for AI in critical infrastructure: algorithms generate options, while human experts validate them and retain authority. Whether that model earns lasting trust will depend not only on minutes and gallons saved, but also on transparency, accountability, and meaningful involvement from the people expected to use it.

Source: Ars Technica AI