The FAA Tests $875 Million AI for Air Traffic Management

The FAA and the Department of Transportation are testing an AI-based air traffic management assistance system in Washington, without replacing controllers. Valued at $875 million, the initiative aims to shift from a reactive to a predictive approach. The controllers' union claims it was not involved in the development, and experts are calling for caution. A gradual rollout to other airports is planned, with support from major airlines.

The union not involved and experts call for cautious implementation

The organization representing air traffic controllers states that it did not participate in the development of the program and emphasizes that any technology must complement, not replace, the expertise and judgment of controllers responsible for the safety of the national system. Air traffic specialists advocate for a cautious approach during deployment. Margaret Wallace prioritizes the training of controllers and the integration of human factors, noting that this is a challenge inherent to any major technology in a critical environment. She believes that FAA leadership must conduct deliberate testing and implementation, ensure adequate training, and maintain safety safeguards during the transition. At this stage, the actual effectiveness of the system remains undetermined.

Stated goal: from reaction to prediction, without sidelining humans

The Department of Transportation claims it wants to evolve the FAA from a reactive stance to predictive management. Transportation Secretary Sean Duffy asserts that the system will help prevent problems, reduce delays, lighten the load on controllers, and lower travel costs. The system is presented as a tool for assistance, not intended to replace controllers, and is meant to provide them with recommendations. Margaret Wallace stresses the importance of maintaining the central role of humans at every stage and sees this initiative, if conducted carefully, as a first step towards modernizing a system that has been challenged by years of staffing shortages.

Announced functions: 200 consolidated data streams for anticipation

The FAA orchestrates nearly 45,000 flights daily in the United States. The SMART system is designed to aggregate 200 streams of information, from flight paths to weather parameters and personnel indicators, in order to provide a unified view of ground and in-flight operations. Based on this data, controllers could anticipate demand, weather, and airspace constraints, planning hours, days, or even weeks in advance, according to authorities. The tool employs artificial intelligence to assist in this planning.

Initial deployment in Washington, $875M, and support from airlines

The trial has started at three airports in the Washington area: Reagan National, Dulles International, and BWI Thurgood Marshall. The initiative, amounting to $875 million, was developed by Air Space Intelligence, an American AI company active in aerospace and defense, which has previously worked with the US Air Force, Alaska Airlines, Delta Air Lines, United Airlines, and the Department of Defense. Leaders from several airlines, including United, Delta, American, and Southwest, have expressed support for the rollout. The FAA plans a gradual expansion of the system to other airports across the country.