Instrumenting an old lab rig

There's a reflex, when a department decides it wants a digital twin lab, to assume the old equipment has to go and shiny new instrumented rigs have to be bought. Usually, that reflex is wrong — and expensive. The machine you already own is often perfectly good; what it lacks isn't capability but instrumentation. And instrumentation you can add.

Retrofitting an existing rig — wiring sensors onto equipment you already have and connecting it to a live model — is frequently cheaper, faster, and less disruptive than replacement. It's not always the answer, but it deserves to be the first thing you consider, not the last. Here's how to think about it and, roughly, how it goes.

Why retrofit usually beats replacement

• Cost. Sensors and a data-acquisition setup cost a fraction of a whole new instrumented rig. The mechanical equipment — the expensive part — you already have.
• Speed. Instrumenting an existing rig is measured in weeks. Procuring, installing, and commissioning entirely new equipment is measured in months, sometimes across a budget cycle.
• Familiarity. Faculty and students already know the machine. You're adding a capability, not forcing everyone to relearn an unfamiliar system.
• Less waste. Perfectly functional equipment doesn't get scrapped to make room. That's easier to justify to whoever controls the budget, and simply the sensible thing to do.

The walkthrough: how a retrofit actually goes

Step 1 — Audit what you have

Start by understanding the existing rig honestly. What does it do? What quantities matter for the experiments you want — temperatures, flow rates, pressures, speed, vibration? Is the equipment mechanically sound? Very often the audit reveals that most of what you need is already there, and the gap is narrower than expected.

Step 2 — Decide what to measure

You don't sensor everything — you sensor what the physics and the experiment require. For a heat exchanger, that's inlet and outlet temperatures on both streams and the flow rates. For a rotating machine, speed and vibration. The principle: measure the quantities that appear in the governing relationships you want students to see, and resist the urge to over-instrument. More sensors is more cost, more wiring, more to maintain, and rarely more learning.

Step 3 — Choose and place sensors

Select sensors appropriate to the ranges and conditions — the right temperature probe for the expected range, a flow sensor suited to the medium, and so on. Placement matters as much as selection: a temperature probe in the wrong spot reads the wrong story. This is where a little engineering judgement pays off, and where getting it right the first time saves rework.

Step 4 — Data acquisition

The sensors feed a data-acquisition system that digitises the readings and passes them onward. This is the bridge between the physical rig and everything digital. For a teaching lab, the sample rates involved are gentle — a handful of readings per second is usually plenty — which keeps this stage simpler than industrial instrumentation would be.

Step 5 — Connect to the model

Live sensor data now flows to the simulation model, which predicts what the system should be doing and compares against what it actually is. This is the moment the old rig becomes a digital twin: real measurement and model prediction, side by side, updating as the experiment runs.

Step 6 — Dashboard and validation

A dashboard makes the whole thing legible to students — the live readings, the prediction, the comparison. And critically, you validate: run the real experiment against the model until the agreement is documented and defensible. An unvalidated model on a retrofitted rig is just a screen; a validated one is a genuine teaching instrument.

When retrofit is NOT the right call

Honesty requires the other side. Retrofit is the wrong choice when:

• The equipment is genuinely worn out. Instrumenting a machine that's mechanically failing just adds sensors to a problem. Fix or replace the machine first.
• The rig can't produce the physics you need. If the experiment you want requires capabilities the old equipment simply doesn't have, no amount of instrumentation invents them.
• Access for sensors is impractical. Occasionally a machine's construction makes placing the sensors you need genuinely difficult, and forcing it costs more than it's worth.

In those cases new equipment is the right spend. The point isn't that retrofit always wins — it's that it wins often enough to be the default question rather than an afterthought.

The takeaway

Before writing a purchase order for new instrumented rigs, look hard at what you already own. A mechanically sound machine plus the right sensors, data acquisition, a validated model, and a dashboard is a digital twin — at a fraction of the cost and time of replacement, and without scrapping equipment that had years left in it. Retrofit isn't the cheap compromise; frequently it's just the smarter engineering.