How to Convert an Existing Autoclave to IoT

Automatic cycle records can be added to the sterilizer you already run — and moved onto the next one you buy. What it takes is a way to measure what the machine is doing, and a way to send that measurement somewhere it will still be readable in five years.

Technical guide · IOT.AUTOCLAVEXPRESS.COM

The problem with the paper logbook

In most practices and small hospitals, the record of a sterilization cycle is a printed strip, a handwritten line in a register, or nothing at all. The strip fades. The register is filled in at the end of the day, from memory. When someone asks, two years later, what happened during a specific cycle on a specific morning, the honest answer is usually that nobody knows.

This is not negligence. It is the predictable outcome of asking busy people to be a data acquisition system. The equipment ran correctly in the overwhelming majority of cases, but the evidence that it ran correctly was never captured in a form that survives.

Connecting the autoclave changes one thing only, and it is the right thing: the record is created by the machine, not by a person, and it is created every single time.

What "IoT" actually means here — and what it does not

The phrase is used loosely, so it is worth being precise. In this context, connecting an autoclave means:

• measuring the temperature of each cycle continuously;
• transmitting those measurements off the machine as they happen;
• storing them in a form that can be searched, exported and audited later;
• raising an alert when a cycle deviates from the expected profile.

It does not mean that the system takes over the sterilizer. Nothing is written back to the autoclave. No valve is actuated, no cycle is started or stopped remotely, and no software decides that a load is safe. That boundary is deliberate, and the rest of this guide will keep returning to it, because it is what keeps the installation simple, non-intrusive and free of the regulatory weight that a control system would carry.

The measurement chain, end to end

A retrofit is not a single device. It is a short chain, and each link does one job:

1. A PT100 probe, class AA, 4-wire (IEC 60751) — the sensing element. Platinum resistance thermometry is the reference method for this temperature range, and the class AA tolerance is what makes readings around 121 °C and 134 °C meaningful rather than indicative.
2. A stainless steel 316L thermowell, 1/2" BSP thread — protects the probe and provides a defined contact surface.
3. A stainless steel clamp, thermal paste and optional thermal insulation — holds the assembly firmly against the drain line and improves the thermal coupling.
4. A PT100 to RS-485 converter speaking Modbus RTU — turns a resistance into a digital value on a robust industrial bus.
5. A gateway, RS-485 to Ethernet, WiFi or 4G — carries the readings to the platform over an encrypted link.
6. The platform — stores the cycles, draws the curves, sends the alerts and produces the reports.

Two power supplies are needed, 9–30 V DC, one for the converter and one for the gateway. That is the whole bill of materials.

Why the drain line and not the chamber

This is the single design decision that makes a retrofit realistic, so it deserves an explanation rather than an assertion.

Putting a sensor inside the chamber means entering the pressure envelope. That means a penetration, a seal, a qualification of that seal, and a serious conversation with whoever holds the warranty on the machine. It also means that every installation is different, because every manufacturer builds the envelope differently.

The drain line is outside the envelope. During a cycle it carries the condensate and follows the thermal profile of the chamber closely enough to characterise the cycle: the rise, the plateau, and the descent. Clamping a probe onto it externally requires no drilling, no seal, no modification of the pressure vessel and no electrical connection to the autoclave's control board.

It also means compatibility stops being a matter of brand. Every Class B autoclave has a drain line. The practical question is not "which model is this?" but "can the drain line be reached to fit an external clamp?" — a question a technician answers by looking, in a few minutes, before anything is ordered. For the placement details themselves, see our dedicated page on PT100 sensor placement for autoclave sterilization.

What you can measure, and what you cannot

Be careful here, because this is where connected-equipment marketing tends to overreach.

Temperature: always. That is what a PT100 does, and it does it well.

Cycle phase: yes, by deduction. The temperature curve makes the rise, the plateau and the descent plainly visible. Plateau duration and plateau stability are read from the curve, not from a separate sensor.

Pressure: only where the autoclave exposes it. A PT100 does not measure pressure, and no amount of software turns a resistance into a bar reading. Some autoclaves publish pressure on their own digital output; where they do, a gateway acting as a Modbus master can read it alongside the temperature. Where they do not, there is no pressure channel, and any vendor claiming otherwise is describing something that cannot happen.

Choosing how the data leaves the room

Three connection paths cover essentially every site:

Ethernet

The most stable option and the simplest to defend to an IT department: a wired link, a single outbound connection, no radio emission at all. Where a network socket is available near the sterilization room, this is the default choice. It is also the only option that involves no radio type-approval consideration whatsoever, since nothing transmits.

WiFi

Useful where running a cable is impractical. The trade-off is a dependency on the site's wireless coverage and on whoever administers it: a password rotation or a new VLAN policy can silently interrupt the recording.

4G

An independent link that does not touch the site network at all. This removes the IT dependency entirely, which is often the fastest way to get an installation approved in a hospital. Note that the SIM card is supplied and paid for by the customer, and that a cellular gateway must be type-approved for the country it operates in — a point to settle locally before ordering, since approval regimes differ from one market to the next.

Wiring: the details that decide whether it works

RS-485 is forgiving, but not infinitely so. The recurring field problems come from a short list:

• Use 4-wire shielded twisted cable for the probe, and a twisted pair for the A(D+) / B(D−) bus.
• Ground the shield at one end only, on the cabinet side. Grounding both ends creates a loop and injects exactly the noise the shield was meant to keep out.
• Fit a 120 Ω termination if the device sits at the end of the line.
• Respect the bus length limit — up to 1200 m, which is far beyond anything a sterilization room requires, but the figure matters when several machines share one run.
• Use thermally conductive paste at the contact point, not insulating paste. They look alike on a shelf and do opposite things.

A 4-wire connection is not a formality either: it cancels the resistance of the leads themselves. With a 2-wire arrangement, the cable adds its own resistance to the measurement, and the error grows with cable length — which is precisely how a class AA probe ends up delivering class B accuracy.

What the records are actually good for

Once cycles are recorded automatically, several things become possible that were not before:

• Retrieving a specific cycle on demand. A date and a time produce a curve, not a shrug.
• Noticing drift early. A plateau that is consistently a fraction of a degree lower than it was six months ago is invisible on a paper strip and obvious on a chart.
• Alerting during the cycle rather than discovering the problem when the load is already in use.
• Producing reports without anyone transcribing anything.
• Keeping an audit trail that shows what was recorded and that it was not altered afterwards.

None of that certifies anything. It documents. The distinction is not a legal nicety: documentation is what you can produce when someone asks a question, and it is worth a great deal precisely because it does not pretend to be a verdict.

The three objections you will hear

"It will void my warranty"

The installation adds nothing inside the machine. No drilling, no seal, no connection to the control electronics. The clamp on the drain line is mechanically comparable to insulation or a pipe support. Where a manufacturer contract is unusually restrictive, the question is worth asking in writing before installation — but the physical intervention is about as minimal as an intervention can be.

"We already validate our sterilizer"

Periodic validation and continuous recording answer different questions. Validation establishes that the equipment can perform correctly under defined conditions, on the day it is tested. Recording shows what happened on every other day. Neither replaces the other, and the second one is what you are missing when a question arrives about a specific load.

"Our IT department will never approve it"

This is a real objection and it deserves a real answer. The system opens a single outbound connection, receives no inbound traffic, transmits machine data rather than patient data, and can run on a 4G link that never touches the hospital network at all. That last option turns a months-long approval discussion into a decision the biomedical department can take on its own.

What a retrofit looks like in practice

1. Check the machine. Confirm it is genuinely a Class B unit and that the drain line is reachable for an external clamp. This is a visual inspection, and it occasionally reveals that a sterilizer sold as Class B is not one.
2. Choose the connection. Ethernet where a socket exists, WiFi where coverage is good and stable, 4G where independence from the site network is worth more than the SIM cost.
3. Fit the chain. Thermowell and probe, clamp, thermal paste, optional insulation, converter, gateway, two power supplies.
4. Verify against a known cycle. Run a standard cycle and compare the recorded curve with what the machine's own display reports.
5. Hand over the account. Users, alert thresholds, report recipients.

The technical work is genuinely modest. What takes time is the decision, and the decision is usually easier once it is clear that nothing about the sterilizer itself is being changed.

Frequently asked questions

Considering a partnership? Here is exactly what it involves.

Talk to our team about your installation

IOT.AUTOCLAVEXPRESS.COM is a division of SCMDSD LLC (USA). The platform records and documents sterilization cycles. It does not determine the conformity of a cycle and does not authorise the release of a sterilized load; that decision remains with the qualified personnel responsible for it. All technical resources