Transportation Secretary Sean Duffy unveiled multi-billion dollar plans to overhaul the nation’s air traffic control system during a press conference on Thursday.
“…once-in-a-generation opportunity to build a brand new, state-of-the-art air traffic control system,” said Duffy.
State of the art - So, they’re getting 3D visualizations with predictive trend vectors and AI assistance, and replacing simplex AM radio with something that can’t get stepped on, integrated with on-board autopilot traffic avoidance? Not.
Sounds like installing the same old sh** on some new hardware. “Thanks for the money,” said some “contractors” and one Mr. Musk.
If you’re an aircraft owner … BOHICA! I predict the requirement for dual-diversity transponders is coming once the transition to satellites starts up. The single diversity box you installed won’t suffice in all locations. There may even be the requirement to install equipment capable of digital communications like texting.
Oh well … the Cirrus owners will love it … more buttons to push.
The ATC Modernization Plan is heavy on private contractors and public-private partnerships — a clear indicator that privatization is on the horizon. This structure isn’t just about technical upgrades — it’s a slow transfer of control over the nation’s air traffic infrastructure to private interests.
Once critical systems are managed by companies like Leidos, Raytheon, Thales, SpaceX, and Palantir, the government’s ability to fully reclaim oversight will be severely limited. This is the same model that shifted NAV CANADA from a government-run operation to a private, nonprofit corporation, and the signs are all there that the U.S. is heading down a similar path. Once this is recognized— pushback will follow.
AI is the VERY LAST THING you want. ATC software should NOT “learn” and make decisions on its own. I use AI for my work and it frequently is VERY wrong.
I understand and agree, Joe. AI has its issues — I’ve seen it draw from outdated data and make basic errors that would never pass in a critical setting like ATC. It’s not there yet for full autonomy, and I’m with you that AI shouldn’t be making life-or-death decisions on its own.
But this isn’t the first time a new technology has faced pushback. Remember when PCs started showing up in offices? Folks worried they’d replace human jobs and take over the world. Instead, they ended up augmenting human work, not replacing it. AI in ATC can do the same — handling routine tasks, spotting patterns in real-time data, and flagging potential conflicts faster than any human could. With the right safeguards and human-in-the-loop oversight, it can become a force multiplier rather than a loose cannon.
| Function | With Human Interface | Without Human Interface |
|---|---|---|
| Traffic Flow Management | High (98-99%) | Moderate (85-90%) |
| Conflict Detection | High (97-99%) | Moderate (80-90%) |
| Surface Operations | High (96-98%) | Moderate (80-85%) |
| Communication Management | Moderate (90-95%) | Low (70-80%) |
| Decision-Making in Emergencies | High (99% with human oversight) | Low (50-70%, high risk of failure) |
So, yes – proceed with caution, AI has potential.
Privatization is just around the corner.
Combine that with 2030 and its time to start working on getting scrap aluminum prices to go up.
The ATC Modernization Privatization Plan…
There, I fixed it…
I thought they were going to modernize the entire NAS by the end of this year. This plan talks about a 3-year timeline, but then elsewhere mentions some things not being competed until 2040. In fact, most big items won’t be competed until after this administration is gone, “based on historic funding levels”.
The outline isn’t wrong about what needs to be done, but the implementation of it will require doing the exact opposite of what this administration claims is its goal of reducing expenses.
I did a brief systems review of NextGen back in 2006 I believe. The current hardware is custom designed and over 50 years old. Even back then there were concerns about the longevity of the 1990s hardware. The current system must be modernized and should converted to OTS (off-the-shelf) systems.
NextGen has not gone well as it got mired in bureaucratic red tape. With these types of programs, a fundamental decision is OTS (off-the-shelf) or custom hardware. The FAA chose custom for reliability and security reasons. The rationale is that a single resistor can be replaced on a circuit board instead of replacing the whole board. The problem is the FAA needs to have a person that can replace the resistor and there are fewer and fewer competent technicians.
No doubt ATC could be completely automated except that weather and pilots cause chaos. AI has difficultly with chaos as most chaotic systems have not been seen before and are not available for training the AI.
The big question in my mind is how a new ATC will interface with pilots. Today, we rely on voice communication. In the future, will most communication be done via datalinks? Will flight plans and updates be sent to our GPS units via data transmission? In one NextGen review, there was a discussion that pilots would communicate with each other via datalinks and ATC would be unnecessary. If so, how will we read and interact with this data when we are bouncing round in the soup?
Dream on! All of that including building 6 new Centers, how many radar systems, etc., in the next three years. The initial studies won’t even be completed by then not to mention the number of contracts that would be needed for anything to start. No way. This administration will be long gone before this happens and then will be forgotten. And staffing…not mentioned. Also more runways needed. Traffic volume already exceeds what are available at many airports.
If the goal is for a successful and safe ATC system, it will need to remain a primarily voice-based communication system. It’s not much of a stretch to say that on every flight I’ve had, my SA has been enhanced by hearing what others on the frequency are saying. Move that all to text-based data links and you lose that.
But it doesn’t have to be either-or: it can be both. It would enhance safety to have clearances also sent via text. IFR routing clearances is one example, but so would taxi clearances. We could even stick to using voice and just back it up with text.
“State-of-the-art” means state-of-the-art at the time the contract is signed. As rogeranderson60 points out, it will be years, if not decades, before any of this is rolled out. Then folks will bemoan the delapidated, antiquated system that’s been delivered.
This will be a rerun of the FAA’s Advanced Automation Program that shoveled billions of dollars into IBM’s coffers before it was cancelled. Same thing will happen except now Elon’s new company, “ATC-ModernizationX”, will magically win the contract (through a rigorous, DOGE-approved bidding process).
Not enough important people have died spectacularly yet. This “once-in-a-generation opportunity”, like all the others before it, will be dragged down by the tender ministrations of our political and bureaucratic boat-anchors. Once everyone’s oars are in the water, we’ll be lucky if the NAS stays afloat, much less makes progress toward that goal.
Remember, Ike gave us the Interstate Highway System in 1956, and now, over a half-century later, it’s pretty good. It’s still not complete, but that’s mission-creep for you. And that project didn’t require that every driver upgrade their vehicle to take advantage of it.
.
None of us here are going to be flying Jetson-mobiles through metropolitan airspace, though your great-grandchildren might. But remember, ol’ George didn’t pilot his commuter-flivver, he rode in it like a personal taxi. The only way such a system can work is to remove those error-prone “bags of salt-water” from the equation.
Living in a semi-rural area, I figure I’ve got another decade of local flying in me, if I’m lucky. Whether I’ll still be winging the all-day cross-country to KOSH every year by then is far from a given.
To put some reality to Secretary Duffy’s plan, the House Transportation Committee allocated $12.5 billion to be spent in FY2025 for modernization of the ATC system. Interestingly, the final numbers approved by the Committee deleted $2 billion “for air route traffic control center replacement.”
So while they will be training thousands of new hire controllers on the old equipment, they will be designing, testing, training and implementing an entirely new system, and it will all be done by 2028? Try 2038, maybe.
This was nothing more than smoke and mirrors. A robust cloak thrown over the faces of Americans while telecom and aviation conglomerates salivate over a huge contract that only they will get. FWIW, supercenters have been in the talks of ATC for 15 years or more with some land already acquired for said supercenters. Short of a trillion dollar investment this project is over before it started. Nothing to see here. We dont now, nor will we in 3-4 years have enough controllers to staff our current centers let alone split that force to occupy and takeover the new ones. Oh and lets not forget that those supercenters will likely be in Red States in rural locations with low chances for hazardous weather. Then the cost to relocate staff. The gargantuan mountain in front of them to accomplish this impossible climb for 1 center let alone 6 in 3-4 years is laughable at best. Cant wait to retire or resign in 26 months.
Indeed, if you read the actual “plan”, the farthest-out date it mentions is 2040.
No Details for Downtime and Disruptions in the 2025 Modernization Plan
The FAA’s 2025 Modernization Plan doesn’t include a mention or consideration for handling the downtime and disruptions that are bound to come with ripping out old gear and installing new systems at thousands of sites. This is a major oversight.
Without a solid game plan for keeping things running during the upgrades, the risk is high for unplanned outages, delayed flights, and miscommunications that could ripple across the entire airspace system. That’s exactly the kind of chaos the plan is supposed to prevent. If the goal is a safer, more efficient airspace, a plan for handling disruptions has to be part of the deal.