Business

FAA ground stop: officials tie outage to new equipment overloading old switches

The FAA blames worn-out telecom gear for halting every departure in the country. Its own officials also named the trigger: new equipment going into the network, the seam a federal audit had already flagged.
Victor Maslow
Add us on Google

The official explanation for why the United States stopped every departing flight on a Friday afternoon arrived in two parts, and only one of them made it into the funding pitch. Transportation Secretary Sean Duffy blamed old switches and routers in the Federal Aviation Administration’s telecom network, which choked on data and began feeding inconsistent radar and communications information to controllers. That account points to age, and it ends with a request for money.

The second part came from Homeland Security Secretary Markwayne Mullin, standing beside him. The overload, he said, appeared to have been triggered by the new equipment the agency was installing. That points somewhere less comfortable: to the way the FAA is replacing its network, not only to the network it is replacing. A federal audit published weeks earlier had already described that weak seam.

What failed, according to the FAA

Duffy said instability in radar and communications data began shortly after 3:30 p.m. Eastern at the Air Route Traffic Control Centers, which handle aircraft in the cruise phase, and at terminal radar approach facilities. The FAA ordered the ground stop at 3:54 p.m. and reopened the airspace at 8:15 p.m., after more than four hours. It was only the second nationwide stop of all departures since the September 11 attacks. The previous one, in January 2023, followed a failure of the system that distributes safety notices to pilots and lasted about 90 minutes, according to Flightradar24.

“Our switches, our routers, actually got overloaded with data … and started to give us inconsistency in the data from those routers. And because of that, we shut the system down,” Duffy said at a White House briefing, as reported by ABC News. He said the FAA still relies on “1980s little copper lines” and drew a firm line around the failure: “Our new IP switches did not go down. It’s the old hardware that went down.” The agency restored service after deploying fixes to reduce the load on those switches, GeekWire reported.

Mullin’s version added the trigger. As the agency put “patches into the system” and added “new equipment and new technology,” he said, it “was overwhelming the older system.” Neither official said what that equipment was. Al Jazeera summarised the briefing as officials admitting that new technology added to the system had overloaded antiquated parts of it, causing intermittent communications and radar gaps. No cause has been formally determined. The FBI does not believe a cyberattack or criminal activity was behind the failure, and investigators were examining whether a software update played a role, Bloomberg reported, citing an official familiar with the matter.

The audit that described the risk

In September the Government Accountability Office reviewed the FAA’s accelerated rebuild of air traffic control and found that the agency had never built an integrated master schedule for the first phase. It runs 11,389 individual project schedules instead, and they are not connected. “As a result, individual projects are scheduled for installations at the same sites, generally without optimizing the schedules to reduce impacts to controller operations,” the auditors wrote. FAA officials said they planned to fix this but could not say when.

The same report records that FAA officials consider the telecommunications program “the most critical as it provides the backbone for ATC communications.” GAO also found no reliable lifecycle cost estimate for the overhaul and warned that the two gaps “introduce heightened risk.” The Department of Transportation only partially agreed with both recommendations.

The audit did not predict a nationwide switch overload, and nobody should claim it did. What it describes is a program installing new systems on top of old ones, site by site, with no single schedule showing how those installations interact. Mullin’s account of the outage describes new equipment overwhelming old equipment. Tony Stanton, a former senior executive at Australia’s Civil Aviation Safety Authority, told Al Jazeera he found it “incredibly surprising” that officials had not anticipated an outage “when there was a report published only in September that absolutely talked to this risk at hand.”

The money question

Duffy spent most of the briefing pressing Congress for funds. Lawmakers appropriated $12.5 billion for the overhaul in July 2025, the administration has asked for another $30.1 billion, and Duffy said he wants $50 billion. “We didn’t think we would have this kind of outage,” he said. “You will have more of them unless the American people and the Congress tell us to build.”

Nobody disputes the case for ripping out copper lines; GAO makes it too. The weak point in the pitch is sequencing. Money speeds up installations, and on the officials’ own account this failure began where installation met legacy hardware. A bigger budget without the integrated schedule and cost estimate the auditors asked for buys more of the same parallel work. AT&T took over the lead role on the telecom replacement in August with an initial $74.3 million award, and the FAA aims to finish that network by the end of 2028, according to GeekWire. Every month until then, old and new hardware run side by side.

Where was the backup?

Redundancy is the question oversight committees are now asking. “The whole aviation community is based on redundancy,” Ron Bartsch, a professor of international aviation law at the University of the South Pacific, told Al Jazeera. “So the fact that a redundant system hasn’t kicked in immediately just raises more questions.” Senator Tammy Duckworth, ranking member of the Senate aviation subcommittee, said whatever caused the stop “will warrant a thorough and complete investigation,” and Senator Maria Cantwell said the Commerce Committee would review how planes communicate “through primary frequencies, backup frequencies, and other issues.”

The AI angle that dominated early coverage has been set aside. Air Space Intelligence, whose SMART traffic-planning tool the FAA said days earlier it was expanding to all 22 en route centers under a 12-year, $875 million contract, said the ground stop “is not based on or a result of the SMART system.” Duffy said there was “no possible way” it was involved, because the tool is not connected to the core systems controllers use to track flights.

The bill lands on the airlines

Airlines carry the cost of the unwinding. Flightradar24 counted nearly 10,000 cancellations to and from US airports through Saturday as of late Friday night, with Delta and American above 2,200 each, United at 1,800 and Southwest near 1,400. Cirium said at least 827 flights had been diverted by early evening, 670 of them back to their departure airports. JetBlue cancelled everything except Europe-bound departures from JFK and Boston, and some aircraft that landed during the stop waited more than five hours for a gate.

Duffy expects the system back at capacity for the Thanksgiving rush, which starts in less than two weeks. That rush will run on the same network, with old and new hardware still wired together, and the FAA has not yet said which piece of new equipment was going in when the old one gave way.

Tags: , , , ,

Add us on Google

Discussion

There are 0 comments.