More news on this day
The Federal Aviation Administration is beginning limited testing of a new artificial intelligence-driven system designed to spot air-traffic bottlenecks earlier and reduce the flight delays that ripple across the country during busy travel days and disruptive weather.
Get the latest news straight to your inbox!
What the FAA is rolling out now
Published coverage indicates the FAA has started using an AI-enabled capability known as SMART in a limited mode in airspace around Washington, D.C., with a plan to expand gradually to other parts of the U.S. system. The initial deployment is focused on identifying potential congestion and schedule conflicts before they snowball into large-scale delays and cancellations.
The Washington-area focus puts early attention on one of the most operationally complex regions in the National Airspace System. Reports tied to the rollout describe SMART as a decision-support layer that brings multiple operational inputs into a single picture for traffic managers, rather than a system that replaces controllers or issues tactical separation instructions.
Several outlets have linked the timing of the pilot to broader reliability concerns in U.S. aviation infrastructure, where technical failures and capacity constraints can quickly cascade. Recent reporting about disruptions in the Northeast has also noted the FAA testing an AI-based system to anticipate schedule conflicts and weather issues, with early activity centered in the Washington region.
Inside SMART: data fusion first, then earlier decisions
Publicly available information from the FAA describes SMART as an enhancement within a broader platform called Flow Management Data and Services, or FMDS. The concept is to consolidate critical data such as airline schedules, flight plans, weather, airspace conditions, airport capacity and other constraints so traffic managers can act earlier in the day and, in some cases, earlier than the day of travel.
In the FAA’s description, FMDS is intended to become a technological backbone for the Air Traffic Control System Command Center, which oversees national traffic flow initiatives like reroutes, ground delays, and other tools used to balance demand with available capacity. SMART is positioned as the part that applies AI-driven prediction and conflict detection on top of that data foundation to help prevent bottlenecks before aircraft depart.
Recent coverage has characterized SMART as able to synthesize a large number of data feeds, including weather and staffing-related inputs, to help anticipate where demand will exceed capacity. The central promise is not just better predictions, but faster consensus around realistic plans when schedules, thunderstorms, runway configurations, or staffing constraints collide.
How this fits with NextGen and the FAA’s modernization push
The FAA’s AI move is arriving after years of incremental modernization under the NextGen umbrella, which has focused on digital communications, satellite-based navigation and surveillance, and system-wide information sharing. FAA summaries of NextGen emphasize the long-term shift from reactive air traffic control toward more strategic traffic management built around time and trajectory planning.
FAA materials describe foundational systems such as En Route Automation Modernization (ERAM) and a networked data backbone as prerequisites for more advanced traffic management concepts. They also highlight that weather remains a leading cause of delays, and that modern traffic flow tools depend heavily on consistent, shared weather and operational data.
Separately, the FAA’s own reports show the agency is transitioning from the NextGen program structure after its formal wrap-up at the end of 2025, with modernization continuing under a new framework. The SMART rollout, along with FMDS, is being framed as part of that next chapter: adding computing power and unified data views so the system can anticipate constraints and coordinate earlier, rather than reacting after queues form.
Why travelers should care: fewer surprise holds, but no instant fix
For travelers, the most visible impact of better flow management is often indirect: fewer last-minute gate holds, more predictable departure slots during constrained periods, and a reduced chance that a delay at one hub snowballs into rolling disruptions across a route network. The FAA’s rationale centers on reducing conflict-driven slowdowns that are rooted in mismatches between planned schedules and real-world capacity.
Even with better forecasting, delays are not expected to disappear. Weather, equipment outages, runway changes, and other real-time disruptions still require human decision-making and coordination across FAA facilities and airline operations centers. Recent reporting on Northeast disruptions underscores how quickly technical failures can disrupt operations, and that software improvements and infrastructure upgrades are separate, parallel efforts.
The key near-term limitation is scope: early operations are described as limited and regional, with expansion expected only after testing and iterative refinement. Travelers may not notice a clear difference immediately outside the initial deployment area, and even within it, improvements may be most visible during high-complexity days when a better plan prevents a queue from forming in the first place.
What happens next: scaling from a pilot to national operations
The FAA’s June 2026 announcement tied FMDS and SMART to a contract award to Air Space Intelligence, describing the systems as complementary technologies intended to improve how flights are scheduled and managed throughout the National Airspace System. The agency’s stated intent has been to begin initial operations in fall 2026, with broader scaling steps dependent on performance during early use.
Published coverage in late September 2026 indicates the limited-mode rollout has begun around Washington, with a gradual expansion plan. That timeline puts the system’s first real-world test directly into the fall travel season, when shifting weather patterns and dense schedules often stress the network.
For the travel industry, the practical question is whether a unified, predictive view of constraints can reduce the need for blunt instruments like broad ground stops and sweeping reroutes, replacing them with earlier, more targeted adjustments. The FAA’s public framing suggests the technology will be judged on measurable operational outcomes such as fewer delays and cancellations, alongside the non-negotiable requirement that humans remain responsible for controlling the airspace.