FAA Begins Using AI to Predict Airspace Congestion and Reduce Flight Delays

The US Federal Aviation Administration has begun using an AI-supported system in day-to-day air-traffic management, initially deploying the technology in Washington, D.C.-area airspace. The system, called SMART, is designed to predict congestion and weather-related disruption before problems develop, giving air-traffic managers more time to respond.

SMART combines around 200 separate data streams, including weather conditions, flight paths, traffic flows and controller staffing. Its predictive engine analyses that information to produce forecasts, visualisations and recommendations that can help air-traffic professionals anticipate pressure on the network rather than reacting only after congestion has developed.

The FAA is drawing a clear boundary around the AI’s role. SMART does not fly aircraft, replace air-traffic controllers or autonomously issue instructions to pilots. Human air-traffic professionals retain decision-making authority, making the system an advisory and planning tool rather than an autonomous controller.

The initial deployment is intended to be followed by a gradual nationwide rollout. Reuters reports that SMART is being supplied by Air Space Intelligence under a 12-year contract worth approximately $875 million, potentially putting the technology into one of the world’s largest and most complex aviation networks.

The deployment represents a significant practical step for AI beyond chatbots and consumer software. Predictive AI is now being introduced into safety-critical US infrastructure where decisions affect aircraft movements and passenger delays, but with humans retaining operational control. The test will be whether analysing those 200 data streams allows the FAA to identify disruption early enough to make the aviation network run more efficiently.

Sources

• FAA: State-of-the-art air traffic control announcement
• Reuters: FAA begins use of AI tool in bid to reduce flight delays