FAA SMART Program: What Airline Ops teams need to know

What is the SMART program?

Airlines in the United States are preparing for SMART – the Strategic Management of Airspace, Routes, and Trajectories, a new initiative rolled out by the FAA. SMART is an AI-enabled predictive air traffic management system that “pulls together airline schedules, weather, airport capacity, and airspace restrictions to predict where congestion and conflicts will build, then flags them 2-2 ½ hours out.” Its objective is to manage congested airspace with small “nudges” before take-off. Instead of untangling congestion during flight, nudge a route early and the congestion never happens.

The first operational demonstration is slated for September 2026, so it’s coming fast. The demo will be narrow in its initial scope, restricted to airspace at 24,000ft and above.

‍

How is airspace congestion handled now?

Currently, if there is a deviation from the scheduled flight plans, the United States ATC uses Collaborative Decision Making (CDM), an internal framework that lets airlines swap, substitute, and cancel slots. It isn’t a perfect system, but airlines have equal positioning in the schedule-based order decisions. SMART moves the decision making into an algorithm.

‍

Who will control the algorithm?

In other words, who holds the keys? This is a question of concern both here and in Europe. Europe is working on a similar predictive-AI congestion tool called ASTRA through its SESAR research program. Europe’s airspace is controlled through EuroControl, joining individual systems across the EU, the rest of Europe, and dozens of air navigation service providers. SESAR was founded as a large public-private partnership to create and implement the roadmap for European airspace. It is not controlled by one national entity, and they have designed their research to be as transparent as possible. For example, ASTRA’s designers state that the algorithm must “explain its recommendations and predicted impacts” to airflow managers. Europe has treated explainability as a design requirement up front to any of the SESAR projects. Their approach is slow and controlled and any AI program is treated as advisory as there is no central or singular governing body enforcing it. ASTRA finished the exploratory phase in late 2025 and is currently working on research and validation.

The U.S. approach is betting on speed and central authority with one single AI system. This all-in on a single vendor solution is certainly a bold way to fast-track fixing an overwhelmed system and comes with speculation, and anxiety specifically for smaller carriers. The question, put most directly by OAG, the global aviation and travel data company, is “when demand exceeds capacity, who decides which flights absorb the pain?” Will a system that is optimized for the total network favor carriers with more flights or bigger hubs, or even a financial partnership?

‍

What about the why? Transparency is key.

If SMART moves, slows down, or reroutes a flight, how does an airline find out why? And does the airline have a say in the decision-making? How will the recommendations be tested? OAG also notes that “validation, governance, and decision rights” are all still undefined. While we won’t learn the answers until later this year, Flight Science is already working on the building blocks to enable an airline to participate in SMART while protecting their interests.

Slot decisions have real revenue and competitive consequences. SMART works on behalf of the national airspace, while Flight Science works on behalf of the airline. Given a set of potential trajectories, Flight Science can highlight them to dispatchers and recommend a solution that balances an airline’s operational priorities, while giving airspace managers the flexibility they need. And our models continuously predict airspace and airport conditions in flight, giving airlines additional cost saving and safety benefits. In each case, we clearly communicate the why behind our recommendations, empowering dispatchers and pilots with the decision-making data they need.

‍

How do these changes affect passengers?

All flight changes have ripple effects, and passenger disruption is the most visible. SMART’s predictive system is being built to manage flight decisions about 1-2 hours before the flight, meaning the adjustments have been made before the passengers board. But SMART trajectories are designed to create the cleanest airspace, without visibility into passengers’ itineraries. If the predictive system pe-emptively slows down or reroutes a flight, the downstream effects can be costly.

Flight Science’s True Cost Index reflects downstream cost factors like missed passenger connections, crew timeouts, and overflight fees, supporting recommendations that balance the actual costs of a flight and allowing dispatchers and pilots to react in real time to changing conditions.

‍

Flight Science is ready to work with the FAA to make SMART successful while improving the operational experience for our airline partners. Click here to talk with Flight Science.