Several members of Congress are calling on the Federal Aviation Administration (FAA) to provide detailed information about recent outages of the Notice to Airmen (NOTAM) system.
I wonder if this could have anything to do with staffing. I know the trouble started before they began randomly firing people, but it looks like maybe it’s gotten worse since then. Clearly, no great improvement. Maybe people do matter after all.
Back in the 1960s, NOTAMs came off a teletype machine or were read to you by a briefer at the FSS. It was slow and sometimes a little vague, but it got the job done. You’d listen, take notes, and check the bulletin board before heading out. It was part of the rhythm of flying.
Then came January 11, 2023. The NOTAM system went down nationwide. For a couple of hours, flights were grounded across the country. The cause turned out to be a corrupted file during routine maintenance. That incident made it clear—this system needed more than patches and workarounds.
Later that year, Congress passed the NOTAM Improvement Act, calling for a full upgrade and a reliable backup by the end of September 2024. The FAA got the backup online just in time. The main upgrade, though, is still not finished.
Then in 2025, we saw two more outages—on February 1 and March 22. NOTAM updates froze for several hours. Pilots could still access existing notices, but nothing new came through. The backup helped, but it’s still unclear how well it handled the load. In complex, high-traffic airspace like LAX, DCA, and ORD, even short-term gaps can matter.
Now we’re hearing it may take $354 million to modernize the system. That sounds like a lot, but considering the scope and what’s at stake, it may be completely justified.
I’ll admit, I like what we have today, EFBs, digital briefings, TFRs shown in real time. It’s a huge leap from the old bulletin board days. But none of that matters if the source data isn’t coming through.
I appreciate the committee for following up. They’re asking the right questions. Proficiency depends on having timely and complete information. That’s what keeps everyone sharp and safe. Here’s hoping this upgrade finally lands.
Completely agree, Raf. More than once a timely NOTAM has saved my rear-end from either embarrassment or damage. Whatever it takes. Ditto for the digital information as well. Not all e-progress is a bad thing…
“$354 million to modernize…sounds like a lot, but considering the scope”
I’ve done global data system roll outs and modernization for 35 years.
Heck, just host a mirrored system on a cloud service and you achieve both uptime and redundancy. Only government math can make a $10 million project cost $350 million.
It’s not that simple, because government has special requirements for cloud services. Some of them are arguably unnecessary, but many of them are for privacy and security that we definitely do want.
Richard, programmers are not staff; they are a commodity item and are contracted. What’s bad (and has gotten worse) is the management and project oversight.
The NOTAM system now is designed to be WIDE OPEN for anyone in the world to read; it’s not supposed to have privacy. It “is that simple” to host a read only database on a cloud service.
(Honestly, one could argue that modern cloud services actually have better overall security than obsolete servers at the FAA).
Unfortunately it doesn’t matter whether something should be simple or not, there are special government requirements for cloud services. And while the end users are accessing it read-only, someone has to input the data and have write access.
And major cloud services already have those government requirements.
As said, this is so easy and so routine these days that (if Raf is correct and they are budgeting $354 million for this) then it should be considered inept or criminal.
Correct, they do, but they are a separate service on separate hardware, which means an added cost.
Check out page 6 of the document. I was also surprised.
ii. The NOTAM Crisis Harbinger of Future Disruptions
“Even before the FAA’s telecommunications crisis, the FAA was working to mitigate the risks associated with its faltering Notice to Airmen (NOTAM) system, which has been the source of significant disruptions throughout the NAS. The NOTAM system is vital for sharing and disseminating safety-critical flight information between both air traffic controllers and pilots. However, in early 2023, a complete failure of the NOTAM system caused nationwide ground stop resulting in significant flight delays. Despite the known vulnerabilities and risks associated with the current system, the FAA will struggle to fund this program without increased F&E funding. At minimum, the FAA will need $154 million just to conduct further research on a replacement NOTAM system, but will need $354 million to replace the broken NOTAM system. Much like the FAA’s looming telecommunications crisis, the NOTAM crisis was not at the top of any F&E priority lists until after the 2023 collapse resulted in cascading nationwide delays and ground stops. We need to learn the lessons from similar events in the past and chart another course, rather than repeat the same mistakes.”
$154 million just to conduct further research; $354 million to replace?
As far as hosted databases go, the NOTAM system is child’s play.
Going rates I see now for large turnkey business databases are $150,000 plus $40,000 yearly maintenance, Make 10 of them for insane redundancy and that’s less than $2 million total.
WHERE the hell is that extra $506 MILLION going?
It’s damn sure not going to to actually doing anything.
I’m skeptical too. $508 million is a lot to swallow.
But this isn’t just about hosting data. The NOTAM system ties into nearly everything. EFBs, dispatch centers, ATC, flight service, and more. It has to run nonstop, meet FAA and ICAO standards, and work across old and new systems.
I’d still like to see the itemized breakdown. That would help all of us judge it fairly. Ask Nick Daniels, President. NATCA
“The NOTAM system ties into nearly everything. EFBs, dispatch centers, ATC, flight service, and more. It has to run nonstop, meet FAA and ICAO standards, and work across old and new systems.”
Datasets streaming is just a simple broadcast, and all the standards for NOTAM’s are well developed and mature and known. $154 million to just RESEARCH something that is already well known and in the public domain is criminal.
As far as reliability, make 2, 3, 5 or 10 mirrored sites. As I said, this technology is well known, mature, and is now a commodity. $354 for sites that cost less than $200K each is even more outrageous and criminal. I’m beginning to think that NOTAM information is too important to be left up to incompetents and crooks. YMMV.
I’ve been trying to make sense of the Daniels testimony, and after reading through it, I think there’s more going on here than just a bloated price tag.
The $508 million figure, $154 million for research and $354 million for replacement, actually came from NATCA President Nick Daniels. It wasn’t pitched as a padded contractor number. It was his estimate for what it would take to modernize the NOTAM system after the 2023 failure.
Daniels pointed out that the system’s collapse in January 2023 forced a nationwide ground stop, and that only after that happened did the FAA start treating NOTAMs as a top funding priority. He framed it as part of a bigger issue. The FAA has been stuck in a cycle of underfunding until something breaks.
As for the $154 million in “research,” it’s not just reading documents, at least I don’t think so. It includes the engineering work needed to rebuild a system that integrates with ATC, dispatch, pilot tools, military operations, and international feeds, all while running nonstop. The $354 million replacement number includes rollout, hardware, software, testing, and the full update to this century’s technology.
Daniels also made the case that the FAA’s broader Facilities and Equipment funding has been running two to three billion dollars behind what’s actually needed for years. They’ve been stuck in a patch and pray model, and the NOTAM system is just one example.
So yes, I agree, $508 million is a big ask. But from Daniels’ side, it’s what it will take to avoid another full system failure. He wasn’t throwing around red flags about fraud or waste. He seemed more focused on finally getting the funding in place before the next breakdown.
That said, I still think a clear itemized breakdown would help everyone feel more confident about the numbers. But based on what Daniels laid out, I don’t think this is about crooks. I think it’s about long overdue investment in systems we all rely on.
There is no “research” to do; they already know the database size, it’s configuration, data types, relationships, update intervals, input and output formats, and expected growth. Everything is there and well documented.
If the stated goal actually is reliability, then it’s as simple as:
I am CIO by trade and I would never trust cloud service providers for the sole source of my critical data. There are so many outages from AWS, Azure, and Google Cloud where the underlying data and systems are all intact, but authentication failures or other activities like that stop people from accessing the data. The Cloud Service Provider’s response is, “your data and compute are intact and running”, but the failure of authentication services stops people from accessing the data. Should be setup with a local or co-lo data center copy that can be made read-write in the event of cloud provider failure. GovCloud (AWS), Azure Government, and Assured Workloads for Government (Google Cloud) all have additional costs beyond traditional cloud provider’s general availability products because of more stringent government requirements (and my personal belief is that the cloud service providers intentionally make these systems more costly because they know the government has no problem paying a premium).
I’ve been in global system & server deployments for over 30 years. That’s why I suggested it as a mirrored backup ported to at least 2 different sites. Chances are that “if” the FAA, Azure, and AWS are all down at the same time… then we’re at DEFCON 1 and any outage in that case is, for the most part, irrelevant.
I generally do not like of approve of cloud because basically it’s very high rent for using other people’s servers. However, it’s one hell of a lot less expensive for ensuring global uptime than dropping $500+ million. A hell of a lot less!