The Federal Aviation Administration’s ambitious Next Generation Air Transportation System, known as NextGen, was once billed as a sweeping transformation of U.S. air traffic control. Launched in 2003, the program set out to replace the nation’s radar-based system with satellite navigation, automation, and digital communications.
More than two decades and roughly $36 billion later, a growing body of oversight reports paints a different picture: a modernization effort that ran over budget, fell behind schedule, and achieved only a small share of its expected gains.
That isn’t to downplay the benefits the program has brought so far; it does have measurable benefits we enjoy today that were not always a given. Digital communications generally allow for faster and more accurate communication. Satellite-based surveillance adds more clarity and precision to determining aircraft locations. Performance Based Navigation helps pilots and operators take shorter flight paths, in turn cutting their fuel burn and emissions. Automation tools help controllers simplify the complexity of their jobs, to a certain extent.
Agency officials say these improvements represent meaningful progress toward the long-term vision of a Trajectory Based Operations system nationwide.
That is unfortunately not, of course, the full picture.
The Department of Transportation’s Office of Inspector General (OIG) concluded that the FAA has delivered a delayed, over budget, and less transformational modernization effort than originally planned. Through the end of 2024, the agency had achieved only about 16 percent of the total expected benefits from the initiative that began in 2003. Many of the system’s key technologies—intended to cut delays, lower fuel burn, and increase airspace capacity—are now postponed until the next decade.
Lofty Goals, Limited Results
NextGen began life as a multi-decade project to overhaul much of the existing air traffic control system. The basic premise was to move air traffic control from ground-based radar to a satellite-driven network capable of handling growing volumes of air travel. Through fiscal 2022, the FAA said it spent more than $14 billion on the project, with total public and private investments projected to exceed $35 billion by 2030, according to the Government Accountability Office (GAO).
GAO auditors found that while the agency met some milestones—such as deploying new digital communications in towers—it missed others by years. Key automation and sequencing systems, including Time Based Flow Management and the Terminal Flight Data Manager (TFDM), have been repeatedly delayed. The TFDM program, an automation tool introducing electronic flight strips to replace paper ones, is now behind schedule and will not be added to a series of airports until 2030, Reuters reported.
Reuters also reported that the FAA reduced the number of airport deployments by nearly half, scaled back planned capabilities, and saw program costs rise by more than 20 percent.
The FAA attributed a lot of those roll backs to the pandemic and ensuing workforce shortages. Fair enough, to a point.
But GAO investigators also found a fairly significant lack of planning and documentation. The agency has not updated its life-cycle cost estimates for NextGen since 2017, for example. It also still lacks a key risk mitigation plan to identify and prioritize major program risks. Both shortcomings make it a vastly greater challenge for the FAA to assess its own performance, refine its annual budget requests, or manage long-term goals effectively.
Accounting for Outside Forces
Auditors have also pointed to external factors—including airline equipage, weather, and fuel prices—that continue to affect NextGen’s implementation. In its July 2025 report, the OIG said the FAA’s limited analysis of those factors “impacts the overall reliability of FAA’s projections of NextGen’s benefits and costs.”
The office recommended that the agency develop a formal process to identify and refine key external influences, a step the FAA agreed to complete by September 2026.
The same report noted that FAA’s benefit projections have fallen sharply over time. In 2013, the agency maintained that NextGen would produce as much as $199 billion in benefits by 2030. By 2024, the projection had dropped to $63 billion by 2040. The OIG attributed that eye watering plummet to deployment delays, economic shifts, and uneven airline adoption of required avionics.
Structural and Cultural Challenges
The modernization effort has also been constrained by workforce shortages and institutional strain. Reuters reported that the FAA is short of its staffing requirements by around 3,500 controllers. Lasting gaps have forced controllers into situations requiring them to work mandatory overtime and six-day weeks as a result. Meanwhile, overtime costs have risen more than 300 percent since 2013, totaling $200 million last year.
Transportation Secretary Sean Duffy has urged lawmakers to continue funding modernization despite the setbacks. After Congress approved $12.5 billion in July to overhaul the air traffic system, Duffy called on lawmakers to award an additional $19 billion for the project.
GAO officials have said that updating NextGen’s life-cycle cost estimate and developing a detailed risk mitigation plan would help the agency manage its modernization efforts more effectively. The FAA concurred with both recommendations.
A Program in Transition
Under the FAA Reauthorization Act of 2024, the agency’s NextGen offices are set to close in 2025, with responsibilities shifting to a new Airspace Modernization Office. The OIG said that as the FAA begins new modernization efforts, developing realistic and achievable long-term plans—including comprehensive risk assessments—will be critical to success.
In a written response to auditors, FAA officials said the agency remains committed to “developing and implementing a cost-effective system” and intends to apply lessons from NextGen to future modernization initiatives.
After more than 20 years of investment, NextGen is certainly still a foundation for what comes next—but also a cautionary tale. The FAA insists that its lessons will guide a new era of air traffic modernization. Whether that next chapter delivers the transformation envisioned two decades ago remains to be seen.
I was a hardware/software program and project manager for many years in the federal and private sectors in charge of major modernization projects as well as a PMI certified project manager. My experience was that most of the problems with inaccurate cost and schedule are introduced in the initial estimates. After that as reality sinks in, the cost and schedule estimates have to be dramatically increased. That’s not too surprising since less is known about the multitude of variables and goals (project scope) in the early stages of a project. NextGen program is a huge set of complex projects so it’s even less surprising that the initial scope, cost and schedule estimates were way off. Sometimes that’s also due to having inexperienced project managers and users involved in creating the initial estimates. I don’t know if that applies to NextGen but it’s prevalent throughout the federal agencies that I encountered. Requirements management is essential to defining the scope of a project and it drives cost and schedule estimates. Engaging users in the initial requirements phase is difficult and expensive and many agencies can’t afford to allocate their most experienced users to the effort so they allocate less experienced or less knowledgeable users to it instead. That results in poor identification of the requirements for the projects and thus poor cost and schedule estimates. When external pressures such as audits and budget fluctuations force an agency to revisit original inaccurate estimates, the result is the dramatic jumps in cost and schedule that NextGen is experiencing. Another part of the problem faced by federal agencies like the FAA is that there are many isolated systems that were developed over time due to budget constraints and lack of a coherent systems architecture that would integrate them. By the time the pressure to integrate them rises to a sufficient trigger point, it is very difficult to achieve effective integration of old hardware, software and databases with modern technology.
There are various solutions to these problems. Step one is to create a systems architecture that defines the anticipated projects in enough detail to avoid massive changes as time goes on and technology improves. Step two is to engage a small number of knowledgeable, experienced project managers and users in defining project scope (requirements). These people get pulled out of their normal career fields to work on those projects, which can hurt their chances of advancing in their core organizations, so the new project work has to become part of their career advancement opportunities. It is essential to have certified, experienced project managers develop and manage the project costs and schedules using robust project management tools with resource-loaded critical path networks. Many federal agencies don’t place enough importance on this to train their project managers properly. Project managers don’t need to be technical subject matter experts, but they need to have a good understanding of the technical aspects of their projects. I’ve seen many projects where the project managers have no clue as to the technical requirements so they are unaware of the true scope, cost and schedule.
Lastly, there needs to be continuity in the personnel involved in long term projects and continued resource allocation to update the scope, cost and schedules on at least an annual basis. Too many agencies get exhausted by the continued effort over time and suffer significant personnel turnover. That’s because the people involved need to stay in touch with their normal careers and have normal opportunities for advancement in those careers. Stable, committed agency leadership is extremely important to make this happen. What we usually see in federal agencies instead is chaotic changes in leadership due to politics at least every four years with a new administration. In any large bureaucracy changes at the top trickle downward as new leaders replace existing staff with their own choices. Often the agendas of the new leadership are radically different from the old and introduce massive changes and instability in project priorities, scope, cost and schedule. We’re seeing that now with this administration. That happens when the new leaders are more focused on achieving their own agendas than understanding and preserving current projects. Obviously, the solution is to select the right leaders and preserve efforts like NextGen that span multiple administrations.
That’s a very nice descriptor, and it is appreciated.
However, one has to return to first principles.
What problem are you attempting to solve?
Next Gen is attempting to use technology to streamline the buggy whip manufacturing process.
Example:
“The TFDM program, an automation tool introducing electronic flight strips to replace paper ones”
Why are you still using flight strips?
METAR KDCA 101152Z 03006KT 10SM FEW040 11/07 A3047 RMK AO2 SLP319 T01110067 53006 $
What information are you attempting to convey here, and is this the best way to present the information?
Yes- What problem are you trying to solve AND- Is it a problem or are we just looking for another way to do the same thing.
If you can find an original copy of the NexGen report that Congress supposedly read before they voted on the appropriation (that’s a laugh - none of them actually read it OR if they did, they had no idea what they were reading and asked ZERO questions) there are multiple scenarios presented in order to support the expenditure of a MASSIVE amount of taxpayer dollars that were, for the most part, hypothetical- very little if any supporting data. Yet it sailed through Congress as most of us in the FAA at that time shook our heads and asked- What the heck??? We’re going to go to control “suites” but we can’t get the basic radar repaired??? We will have money for training scopes for new hires but we can’t fix the real scopes that we are actually running live traffic on???
It was a total scam designed and supported by the industrial complex and the Senators and Congressmen who were in their pocket that, in the beginning, gave out one estimate and then ran the bill up 100 fold and NEVER completed the job.
I hope someone can find a copy of the original report. Critical thinking and critical questioning of the supporters combined with a price-locked contract is sorely needed.
Sometimes projects get low balled because, if the true cost were presented to taxpayers, it would never fly. We have a saying in DoD - if the politics don’t fly, neither will the system
I think in these situations you have to sell the project in phases with some gaps in between not a monolithic “NextGen” system.
You’re correct about getting a flexible architecture in place. You’ll need it to do the project in phases and do upgrades over time.
As for certified PMs, PMP just means you passed a test and were able to produce documentation that you actually ran a project. Forget all that. I need someone with a track record or who was mentored by someone with a track record and comes highly recommended from that individual - I don’t care what their certifications say. Heck, I’ve successfully transition technology from DoD labs to the warfighter without PMP.
As for the shifting winds every four years - I often say 2 years because that’s how long the congressional cycle is - I often tell PMs (I teach a class locally in my org) that you need to plan your project in 2-3 year chunks. Why? That’s the attention span of the system. All the more reason to plan the project in phases. But I don’t think you can say “phases”. Give each phase a different name so it looks like a different program. Then you aren’t answering questions like why it’s costing so much and you don’t have to remind people that if the true cost were stated up front it would never have gotten funded.
If you haven’t already, go read a book names “Why Big Things Fail”. It basically boils down to lack of sufficient planning and bad requirements.
So to what degree does the inadequacy of NextGen affect ADS-B if at all?