The hard part of rebuilding air traffic control is the sequence

FAA Administrator Bryan Bedford asked Congress on Tuesday to support a $22.4 billion fiscal 2027 budget, while reporting progress on the agency’s air traffic control overhaul. By September 1, he said, more than 64 percent of its old copper-based telecommunications network had been replaced and 21 towers had moved from paper flight strips to electronic ones.[1]

A Government Accountability Office report released a day earlier identifies a different measure of progress: whether the work fits together. GAO found 11,389 individual project schedules but no integrated master schedule, and no comprehensive, well-documented estimate covering the program’s full life-cycle costs.[2]

Those findings address complementary questions. Equipment can be arriving quickly while the order in which it becomes operational remains insufficiently defined. For the National Airspace System, a replacement program has to function as a transition: traffic continues to move while installations, training and changes to operating procedures compete for time at the same facilities.

The date has a boundary

The first phase, covering communications, surveillance, weather and training systems, targets December 2028. The second concerns new automation for tracking aircraft and optimizing traffic; it has no estimated start or completion date. FAA’s approximately $10.2 billion estimate for that phase excludes facilities. DOT partially concurred with GAO’s recommendations, pointing to existing estimates and schedules and work on an AI-assisted scheduling tool. GAO maintained that integration and full cost coverage were still needed.[2]

The distinction matters whenever the overhaul is reduced to a single completion date. An equipment deadline can be a meaningful commitment within its defined scope. It cannot also serve as evidence that every later automation capability will be ready. The useful question is which operational functions will have changed by that date, and which will still depend on another investment and transition.

Consider a hypothetical facility scheduled for a communications replacement and a separate equipment installation. Each project may have a plausible timetable on its own. Together, they could depend on the same access window or the same personnel. Moving one activity may then change the feasible date for the other. Counting two schedules does not expose that relationship unless the dependencies are represented.

GAO’s longstanding schedule-assessment guidance connects this problem directly to money. A credible cost estimate must account for the effects of slippage. Delays can produce added staffing and overtime costs, while a reliable integrated schedule lets managers examine how one change affects later work. The guidance treats the schedule as a model of the program, capable of testing whether its planned sequence is achievable.[3]

That model is particularly valuable when acceleration is the objective. It can show where additional resources would advance completion and where they would merely move a queue to the next constrained activity. A faster installation rate is useful evidence; the critical path determines how much of that speed reaches the operational result.

Urgency has a documented history

An earlier GAO review, published in September 2024, described why replacement could not comfortably wait. FAA’s 2023 assessment had classified 51 of 138 systems as unsustainable and another 54 as potentially unsustainable. Among those 105 systems, 58 had critical operational impacts. GAO identified 17 as especially concerning because of their condition, importance and modernization timing; four critical systems had no ongoing modernization investment as of May 2024.[4]

That is a dated baseline, not a current inventory of failures. It records the condition and investment picture against which the urgency of the overhaul developed. Some replacements then lay six to thirteen years away. The new program’s speed should therefore be evaluated against a real maintenance problem, while the old counts should not be carried forward as if nothing had changed.[4]

The policy tradeoff is more demanding than choosing between haste and caution. Continuing to rely on difficult-to-sustain equipment carries costs of its own. A transition plan has to reduce that exposure while managing the temporary demands created by replacement. Evidence of the old system’s weakness strengthens the case for action; it also raises the value of knowing when each dependency can safely be retired.

People and interfaces have their own timetable

Bedford’s testimony also put the workforce in context. In August, FAA had about 11,000 certified professional controllers across 313 federal facilities, against a target of 12,563. About 4,000 controllers were in training, including roughly 900 who had previously certified elsewhere and transferred. The fiscal 2027 request includes $95.4 million to hire and train 2,300 trainees.[1]

FAA’s May hiring plan explains why intake cannot immediately close a certification gap: a new hire can take more than two years to certify, depending on the facility. Its planned hiring rises from 2,200 in fiscal 2026 to 2,300 in 2027 and 2,400 in 2028. The plan also includes changes to training capacity and methods.[5]

Modernization and staffing are consequently linked without being interchangeable. More trainees are a future contribution to capacity, and new equipment is a future contribution to capability. Neither number alone describes the resources available for a particular transition. Planning has to account for the work required to make both useful at the facility where they are needed.

FAA’s description of Flow Management Data and Services, or FMDS, illustrates the intended operational change. The program replaces the Traffic Flow Management System and combines flight plans, schedules and current aircraft positions to anticipate congestion. It aims to consolidate information, reduce repeated manual entry and improve the exchange of routing information between FAA and operators. Those are stated program objectives, rather than demonstrated nationwide gains.[6]

A 2024 interface-design study by Sinan Abdulhak and colleagues examined this problem through interviews and user testing. Its proposed interface sought to unify more than 50 traffic-management services and streamline the evaluation and monitoring of traffic-management initiatives. The researchers reported workflow benefits in their design study and identified development of a live prototype as a next step. That finding does not establish the performance of a deployed FAA product.[7]

It does show why the operational unit of improvement is a task, not simply a replacement screen. A useful change should make it easier to obtain the relevant information, understand the available choices and carry a decision through the system. The installation count measures delivery. Facility acceptance and demonstrated workflow performance will determine what that delivery accomplishes.

The next persuasive evidence would connect those levels: a timetable with visible dependencies, costs that cover the full operating life, and milestones tied to usable capabilities. Tuesday’s equipment figures establish momentum. The audit asks whether that momentum can be converted into a coherent transition across the national system.