It is a in Bristol. The rain is thin and persistent, the kind of weather that feels like it’s trying to solve a puzzle by seeping into the very brickwork of the hardware shop on Gloucester Road.
Upstairs, in a small office that smells of ozone and slightly burnt coffee, Owen is holding a single, scuffed proximity badge. He has a cheese and pickle sandwich in his left hand and the badge in his right. It was sent to him in a padded envelope by the practice manager of a local veterinary clinic-a client he inherited when a local installer named Dave finally retired to the coast.
The clinic’s doors work. They have worked since . The readers are yellowed by a decade of fluorescent light, but they click with a satisfying, mechanical finality every time a staff member presents a card. The problem is not the system. The problem is that the clinic has hired four new veterinary nurses, two interns, and a part-time receptionist, and they are out of cards.
Full-site “Rip-and-Replace”
Legacy Proximity Cards
The gap between “ecosystem migration” and operational necessity.
Owen has been receiving emails from a manufacturer’s representative for weeks. The emails are glossy. They use words like “ecosystem,” “seamless migration,” and “cloud-native.” They want Owen to tell the clinic that their legacy system is a security risk. They want him to quote for a full-site rip-and-replace that would cost the clinic nine thousand pounds.
Owen just needs eighty cards.
The Maintenance Economy
We are taught to look at access control as a series of monumental upgrades-the “rollout” that transforms a campus or the “deployment” that secures a skyscraper. In reality, the industry survives on the maintenance economy. It is a sprawling, informal web of people like Owen who keep the world’s doors swinging by finding eighty cards that will talk to a reader installed when the Blackberry was still a dominant piece of technology.
I was wrong about this for a long time. Early in my career, I worked as a junior consultant on a project for a mid-sized textile warehouse. I was convinced that their reliance on 125KHz proximity cards was a ticking clock. I told the owner that the lack of encryption was a liability. I pushed for a “unified” digital platform that promised real-time analytics and remote management.
We ripped out sixty-two working readers. We replaced them with high-end, networked smart readers. , the main server suffered a corrupted firmware update during a thunderstorm. The warehouse was locked down for . The owner lost four thousand pounds in shipping delays. The old system, for all its technical simplicity, had never once failed to recognize a valid bit-stream. I had replaced a reliable physical reality with a fragile digital promise because I was enamored with the “new.”
Technological Reliability in Environment
Cathedrals of the “Next”
This bias toward the new is built into the way products are marketed. Trade shows are cathedrals of the “next.” You will see massive booths dedicated to facial recognition, biometric palm scanners, and mobile credentials that live in a digital wallet. You will almost never see a booth celebrating the 125KHz low-frequency (LF) card.
Yet, in the damp clinics, the dusty warehouses, and the suburban office parks of the world, those LF cards are the workhorse. They are chosen for reliability in environments where high-frequency (HF) signals might struggle.
When you get down into the chip architecture, the world becomes much more specific than the marketing brochures suggest. A card isn’t just a “card.” It is an ATA5577 or an EM4305. It is a programmable bit of silicon that has to match the internal logic of a reader that hasn’t seen a technician in five years.
If Owen orders the wrong frequency-if he accidentally buys 13.56MHz MIFARE cards when the vet clinic is running on a 125KHz proximity loop-the cards are expensive coasters. The practice manager won’t care about the “longer read range” or the “higher security” of the HF chip. She will only care that the nurse standing in the rain at 6:00 AM can’t get in to feed the cats.
This is where the friction lives. The manufacturers want to close the door on the past. They want “mandatory upgrade paths.” They want to move everyone toward a subscription model where the “cloud” manages the permissions. But the people who actually do the work-the locksmiths, the small integrators, the facility managers-know that the past is where their clients live.
To bridge that gap, you need a supplier that understands the maintenance economy. You need someone who doesn’t just sell you the “latest” but helps you identify the “current.”
I spent yesterday updating the firmware on a suite of acoustic analysis software I haven’t opened in .
“A promise is a tension. When a brand says limited 16 times, the thread loses its memory.”
– Sofia, Thread Tension Calibrator
The Threat of the “Unsupported”
The update promised “enhanced UI responsiveness” and “integrated social sharing.” I don’t want to share my frequency response graphs on social media. I just want the software to open when I need to calibrate a room. The industry-wide obsession with the update is often a mask for a lack of utility. We update because we are told the old version is “unsupported,” which is a polite way of saying the manufacturer has decided to stop helping us use the things we already paid for.
In the access control world, “unsupported” is a threat. It’s a way to force a capital expenditure on a business that was perfectly happy with its operational status quo. The integrators who survive are the ones who refuse to be bullied by the “legacy” label. They are the ones who know that you can still find quality, factory-direct manufacturing for chips that the big players want to forget.
When a client doesn’t know what they have, the process has to be clinical. You take the scuffed badge. You send it to a lab. You test the frequency. You identify the chip architecture-whether it’s a standard proximity card, a MIFARE 1K paper ticket for a one-off event, or a high-security DESFire 4K for a laboratory.
Only after the chip is identified can you talk about the material. Maybe the hotel wants wooden NFC keys to look sustainable, or maybe the construction site needs epoxy tags that can survive being dropped in a puddle of wet concrete. The technical trade-offs are real. Low-frequency cards are champions in damp conditions because the longer wavelength isn’t absorbed by water as easily as high-frequency signals.
High-frequency cards offer a longer read range and the ability to hold more data, but they require more power and are more sensitive to metal interference. Ultra-high frequency (UHF) cards can be read from across a parking lot, but they are finicky and power-hungry. A good integrator doesn’t pick the “best” technology; they pick the technology that matches the environment.
The veterinary clinic doesn’t need a parking lot read range. They need a card that works through the pocket of a white coat while the nurse is carrying a nervous terrier. They need the 125KHz proximity card that Owen is currently pricing out.
It doesn’t have a keynote speech at a tech conference. It happens in emails between integrators and manufacturers like
where the conversation is about chip compatibility and sample testing rather than “disruption.”
There is a specific kind of arrogance in assuming that every system should be “modern.” My friend Nina N.S., an acoustic engineer, often talks about “noise” in a system. She isn’t just talking about sound; she’s talking about unnecessary complexity that obscures the signal.
“A door that needs a cloud-server heartbeat to stay locked is a high-noise system. A door that uses a local 125KHz reader to trigger a physical relay is a low-noise system.”
– Nina N.S., Acoustic Engineer
For a local vet clinic, low noise is a virtue. When we push for the “single platform” or the “unified dashboard,” we are often just pushing for a single point of failure. We are trading the resilience of local, legacy systems for the convenience of centralized control.
Resilience in the Informal Web
For the manufacturer, this is great. It creates a “sticky” customer. For the integrator like Owen, it’s a trap. If he moves all his clients to a single proprietary platform, he loses his ability to solve their problems independently. He becomes a middleman for a software company’s billing department.
The sprawling, informal web of patchers is what makes the world’s infrastructure resilient. It’s the ability to find a replacement part for a machine that hasn’t been manufactured since the late nineties. It’s the knowledge that a specific magnetic stripe card can still be produced with a custom logo to match a chain of budget hotels that isn’t ready to spend a quarter-million dollars on an RFID upgrade.
Owen finishes his sandwich. He pulls up a spreadsheet and enters the chip specifications he identified from the sample. He isn’t going to email the manufacturer’s rep back. He’s going to order the eighty cards from a factory-direct source that doesn’t care if his client’s system is “legacy.” He’s going to charge the clinic a fair price for the cards and an hour of his time to program them into the local controller.
The clinic will keep the nurses they just hired. The nurses will be able to get into the building. The cats will get fed.
The eighty cards that arrive in a plain white box are the only reason the clinic’s doors remember how to stay shut.
The Obsolete Fallacy
We talk about the “digital transformation” as if it’s an inevitable tide. We act as if the old ways of doing things are simply waiting to be washed away. But the tide has a way of going out, leaving the sturdy, well-maintained structures standing while the flimsy, over-engineered ones rust. A system that works is never “obsolete” to the person using it. It only becomes obsolete to the person trying to sell them a replacement.
The real work of the world is done in the margins of the brochures. It is done by people who understand that a fourteen-year-old reader is a triumph of engineering, not a failure of procurement. It is done by identifying the chip, testing the sample, and making sure the signal gets through the white coat and the rain.
Owen closes his laptop. The rain is still falling on Gloucester Road. He has a few more calls to make-a membership club that needs a thousand barcodes printed on PVC cards and a hotel that wants to try wooden key cards for their executive floor. Each of them represents a small piece of the maintenance economy. None of them are looking for a revolution. They are just looking for a way to keep their doors open.
As I sit here, looking at the notification for yet another software update that I will likely ignore, I think about that vet clinic. I think about the satisfying click of a relay that doesn’t need the cloud to tell it what to do. There is a quiet authority in the person who can look at a legacy system and say, “I can keep this going.”
We don’t need a single platform to rule everything. We need more people who know how to find the eighty cards that work. We need the manufacturers who are willing to keep making them, and the distributors who are willing to test the scuffed badges sent in padded envelopes. We need to respect the persistence of things that already work.