How did ‘wire replacements’ ground planes across the entire US?

On Friday, all flights across the United States were grounded for more than four hours due to what appeared to be a telecommunications issue. The Federal Aviation Administration (FAA) issued an alert citing “extraordinary circumstances” and called for all departing flights to remain ground stopped until further notice. According to media reports, about 3,500 flights were in the air when the alert was issued – all landed safely. With thousands of flights delayed and cancelled, millions of passengers were stranded in airports. The travel disruptions have been slowly resolving over the weekend. The ground stop was attributed to an update to air traffic control equipment during planned maintenance. So what exactly happened? What caused it? US Transportation Secretary Sean Duffy attributed the disruption to maintenance undertaken by government contractor L3Harris at a remote FAA facility. The work involved replacing copper telecommunications wires with fibre-optic links, as well as changing how the system sends signals from an older method known as time-division multiplexing (TDM) to the more recent internet protocol (IP). During this maintenance, a router (a networking device) reportedly generated more than 1,000 connection requests every four to five seconds. This affected other routers on the network, disrupting aircraft communications and some radar data feeds. Duffy described the upgrade as a transition from analogue to digital technology, but this conflates different aspects of telecommunications. Copper and fibre describe the physical medium carrying signals, while TDM and IP describe how information is transmitted and organised. Copper cables can carry digital data, and TDM is widely used in digital telecommunications. Fibre-optic systems also carry digital data, using light rather than electrical signals. Consequently, replacing copper with fibre doesn’t necessarily mean a change from analogue to digital. The more significant change is the method for sending signals. With TDM, the system sends information through a single channel at fixed time slots, a bit like letting through traffic across a single-lane bridge at set intervals. By contrast, an IP network can carry information in packets along many paths, which improves flexibility and how much data the network can handle. However, it also changes how the equipment interacts – and what can happen in case of a fault. If not isolated in time, a malfunctioning router can affect other equipment connected to the same network. On fibre, excessive connection requests will travel more freely. Why would a network disruption ground planes? Air traffic controllers use surveillance and radio communications to make sure planes travel smoothly and stay out of each other’s way. Radar and satellite-based surveillance provide aircraft positions, while communications allow controllers to issue instructions and coordinate movements between control centres. All this relies on telecommunications networks connecting facilities across the country. Aircraft can navigate independently during a network failure, but controllers can’t safely manage the usual volume of flights if the comms are not fully reliable. The precise cause of why the router sent so many requests remains under investigation. Incorrect installation, configuration errors, and defective equipment are all potential explanations, and we’ll have to wait and see. A nationwide ground stop following a technical failure is not unprecedented. In January 2023, a malfunction of the FAA’s Notice to Air Missions system, which supplies pilots with operational information, also suspended departures across the United States. Although these two failures involved different systems, both show that a national aviation network that shares services throughout can be broadly impacted in case of a fault. Was this related to Microsoft or AI? The incident prompted speculation about it being linked to the global Microsoft outage that occurred just the day before. Another suggestion was that the fault might lay with the FAA’s recently announced expansion of SMART, an artificial intelligence-assisted traffic management system. Both Microsoft and Air Space Intelligence, the company responsible for SMART, have denied involvement. Officials also noted that the maintenance of the network was completely separate from this air traffic control modernisation program. What can be done to prevent this happening again? In September 2026, a US government report classified 51 of the FAA’s 138 air traffic control systems as unsustainable, and another 54 as potentially unsustainable. The FAA is actively working on replacing ageing communications, radar, and control equipment. However, modernisation introduces additional risks. When you add a new technology to the system, not only do you need to fit it with existing infrastructure, but you also need to maintain operations. To prevent ground stops like the one this week, the FAA will need to do more than replacing cables or purchasing equipment. If they segment the network, faults won’t travel as far. With proper configuration, they can ensure communications don’t get overloaded if a similar router error were to happen. The ground stop was a good measure to ensure safety when the comms failed. But a resilient aviation system must not only respond safely to equipment failures, but also prevent localised faults from disrupting national operations.