Donald, AI Wont Make the Milk Flow or the Airplanes Fly

And it won’t keep aircraft flying when the FAA core capability - its communications backbone - fails
Donald, you know the line... Before we get to it, let me tell you what cows and airplanes have in common.
Some years ago, I worked on renewing 3,000 switches, 200 firewalls and 2.000 km copper wiring across 85 dairies. All operating 24/7. All dependent on their networks to keep production running.
Without the network, production could not access the recipes. Without the recipes, it could not process the milk. And the cows were not waiting for IT.
That is what came to mind when the FAA’s telecommunications disruption brought departures across the United States to a halt on 13 November 2026.
Aircraft waiting to fly. Milk waiting to be processed.
Different operations, different stakes. The same dependency on infrastructure beneath the service everyone takes for granted.

The capability beneath the promise
For a country as large as the United States, moving people by air is a core national capability. Businesses, families and entire communities depend on it.
For a dairy business, processing milk is the core capability. Production turns a perishable raw material into storable products customers can buy.
Neither capability exists independently of its infrastructure. You can have aircraft, airports, pilots and passengers. You still need dependable air traffic communications.
You can have milk, production equipment, employees and orders. You still need access to the information that tells production what to do.
The network belongs in the same management conversation as production capacity, delivery commitments and revenue. Its availability helps determine whether those commitments can be kept.
What the FAA case tells us
The FAA reported that excessive network traffic destabilised communications and radar feeds, prompting the nationwide ground stop. Safety procedures worked, and aircraft landed safely. The disruption nevertheless interrupted a service on which the country depends. FAA account
According to the Department of Transportation, the incident occurred during work transitioning an old copper connection to fibre. As of 17 November, the precise root cause remained under investigation. Initial DOT account, investigation update
That matters. An incident during renewal does not prove that ageing cabling caused the failure. It does, however, put a difficult operational question in front of every leader responsible for critical infrastructure:
How do you renew the backbone while keeping the operation running?
The wider problem was already documented. In September 2024, the US Government Accountability Office reported that the FAA had classified 51 of its 138 air traffic control systems as unsustainable and another 54 as potentially unsustainable. GAO report
Those findings should command attention long before passengers encounter a departure board full of cancellations.
Back at the dairy, the milk keeps arriving
Across 85 dairies, the equipment numbers were substantial. But counting switches and firewalls was only the beginning.
My work covered documentation, approach, design, implementation, reliability calculations and lifecycle management, including renewal of copper-based backbones with fibre. The purpose was to protect production through the change and maintain the infrastructure it would continue to depend on.
Documentation makes that work possible. You need to understand what exists, how it connects and which operational capabilities depend on it. An inventory tells you that a switch is installed. Useful documentation helps you understand what stops if it becomes unavailable.
That understanding informs the design. Where is resilience required? Which dependencies could undermine it? How will the operation recover when something fails?
It also informs the implementation. A dairy operating around the clock cannot make its dependencies disappear while the network is being renewed.
The change sequence, validation and recovery arrangements must reflect that reality. Replacing a connection is only complete when the operation depending on it can function as intended.
Fibre still needs a plan
Moving from copper-based backbones to fibre was part of renewing the foundation. But fibre alone cannot guarantee availability. The surrounding equipment, configuration, connections and operational procedures still matter.
A modern component enters an existing system. The transition must account for how the old and new parts interact, what could fail and how the consequences will be contained.
The FAA case brings that issue into sharp focus. Renewal is necessary, and the transition itself requires disciplined risk management. Leadership must fund both the future infrastructure and the work needed to reach it safely.
A reliability figure is the beginning of a decision
Reliability calculations were also part of my approach. For repairable equipment, the usual term is mean time between failures—MTBF. It provides a statistical estimate, not an expiry date or a guarantee for an individual device.
The business questions extend beyond that figure.
Which failure could stop production? How quickly could service be restored? Are replacement parts available? Could two supposedly independent connections fail together? A component can have an impressive reliability figure while the operation remains exposed to a single point of failure or a lengthy recovery.
Management needs to understand that exposure in operational terms: what could stop, for how long, with what consequences, and what investment would reduce the risk.
Lifecycle management carries those decisions forward. Equipment ages, support ends and requirements change. Renewal needs continuing ownership and funding after the project team has finished.
The cost outlives the outage
When connectivity returns, the consequences may still be unfolding. A dairy may have interrupted production, delayed deliveries and raw material with less usable life remaining. An airline may have aircraft, crews and passengers in the wrong places.
Recovery consumes time, money and resources. Missed production can mean lost revenue. Missed commitments can damage trust and goodwill. Some losses can be recovered. Others cannot. A perishable resource does not regain its useful life because the incident has been closed.
That is why a deferred maintenance budget cannot automatically be presented as a saving. The organisation may simply be accepting a larger, less controllable expense later.
Now add AI
AI makes the investment conversation more exciting. It also makes the dependency question more urgent. An AI service using live production data depends on access to that data. An agent promising delivery dates depends on functioning order, inventory and production systems. Automated decisions depend on information arriving reliably enough for the task.
Each new capability places another expectation on the infrastructure beneath it. You can build a compelling demonstration while those dependencies remain unresolved. Running a dependable operation is a longer commitment.
Before approving the next AI initiative, leadership should ask:
Can the backbone support the promise we are about to make—and have we funded its continued reliability?
Ageing equipment, fragmented architecture and unclear ownership will not be repaired by a more ambitious presentation. AI will not make the milk fly.
Donald, this standard applies to you too
The FAA’s infrastructure problems span administrations. Renewal investment is underway, including funding approved during the Trump administration. It would be inaccurate to claim that one president created the entire problem or that nothing has been invested. DOT funding and progress update
But inherited problems become responsibilities when you take office.
The same principle applies when you become CEO, join a board or take ownership of a critical operation.
An outage alone does not establish leadership failure. The test is whether leaders understood material risks, assigned responsibility, funded a credible response and followed through.
My position is that leaders who knowingly leave those risks unmanaged while financing new ambitions should lose their jobs. The backbone is the spine of the organisation. Keeping it healthy preserves the ability to operate, adapt and grow.
So dream about AI. Build new products. Develop new services. And maintain the capabilities that allow you to deliver any of them.
Because if you knowingly neglect the backbone until aircraft cannot depart or milk cannot be processed, the next performance review should be yours.
Donald, you know the line allready: You’re fired.
References
Federal Aviation Administration — Nationwide Ground Stop: Questions and AnswersThe FAA’s account of the 13 November 2026 disruption, its operational impact, safety procedures and investigation.Read the FAA account
US Department of Transportation — Update on Restoration of Airspace Following Nationwide Ground Stop · 14 November 2026The initial explanation of the incident during copper-to-fibre transition work and the restoration of operations.Read the initial update
US Department of Transportation — Update on Nationwide Ground Stop Investigation and Air Traffic Control Modernization · 17 November 2026The investigation’s status, including uncertainty about the precise root cause, alongside information on modernisation funding and progress.Read the investigation and modernisation update
US Government Accountability Office — Air Traffic Control: FAA Actions Are Urgently Needed to Modernize Aging Systems · 23 September 2024 · GAO-24-107001Independent oversight documenting that 51 of the FAA’s 138 systems were classified as unsustainable and another 54 as potentially unsustainable.Read the GAO report
Author’s professional experience — Network renewal across 85 dairiesYour account of renewing 3,000 switches and 200 firewalls in 24/7 production environments, including documentation, design, implementation, reliability calculations, lifecycle management and copper-to-fibre backbone renewal. An anonymised firsthand case; no public reference linked.
Tags:





Comments