AI & Transportation · September 22, 2026 · 2 min read
The FAA’s SMART AI Tool Will Help Controllers Anticipate Congestion, Not Replace Them
The FAA has begun deploying SMART in Washington, D.C., centralizing roughly 200 data streams to help aviation specialists plan around weather, traffic, and staffing. Human controllers remain responsible for operational decisions.
The FAA’s SMART AI Tool Will Help Controllers Anticipate Congestion, Not Replace Them
September 22, 2026
The Federal Aviation Administration has begun using an AI-supported planning platform called SMART—Strategic Management of Airspace, Routes and Trajectories—in Washington, D.C. The U.S. Department of Transportation says the system brings together about 200 data streams, including weather, flight paths, traffic flow, and controller staffing, to help aviation specialists see where congestion may develop.
The system is meant to improve planning and reduce delays, not to hand flight separation decisions to an autonomous model. Specialists use the platform’s forecasts and visualizations to decide whether adjustments are needed. The distinction is important in a safety-critical environment where inaccurate recommendations can affect thousands of passengers.
A data coordination challenge
Air traffic decisions rely on information spread across airlines, airports, weather services, and FAA systems. Bringing these feeds into one view may help staff identify conflicts sooner and plan routes more efficiently. But data integration also creates risks: delayed feeds, inconsistent definitions, missing records, or bad forecasts can make a dashboard misleading.
The FAA says SMART uses forecasting to show future conditions and support recommendations. A useful evaluation should disclose how far ahead it predicts, how accurate those predictions are, and how the system performs during severe weather or unusual demand. Users should know when data are stale and how to override a recommendation.
Human control and operational safety
The FAA should make clear which actions the software can take automatically and which require a human decision. For aviation, the safest approach is a well-defined division of responsibility, trained operators, audit logs, fallback procedures, and regular simulation before live changes.
AI-assisted planning can reduce workload if it filters noise and highlights relevant scenarios. It can increase workload if it creates too many alerts or requires controllers to double-check every recommendation. Human-factors research should measure whether operators understand the system and retain their ability to make decisions without it.
How to measure success
The system’s rollout should be assessed using delay minutes, cancellations, route efficiency, forecast accuracy, and safety events. Comparisons should account for weather, seasonal traffic, staffing levels, and other operational changes. Publishing performance data can help passengers and aviation workers judge whether the investment is improving service.
SMART represents a cautious use of AI in transportation: combine data and forecast capacity, while keeping specialists in the decision loop. Its value will be determined by real-world reliability, clear accountability, and whether it helps people manage airspace more safely and predictably.
Sources
Governance & Safety
Policy, evaluation, security and the control frameworks that make AI deployments defensible to auditors, customers and regulators.
Browse Governance & Safety