Skip to main content

MemoSuite: Calibrating Sensors at the Bench Instead of in the Plant

Calibrating a sensor in situ means taking your instrumentation people to the least convenient place on the plant, often in weather, sometimes in a hazardous area, and asking them to do careful work there. MemoSuite moves that job to a bench, and changes what happens at the measuring point into something far simpler: swapping a used sensor for one that is already calibrated.

Why field calibration is the wrong place to do careful work

A calibration is a precise procedure. It wants stable temperature, clean buffers in date, unhurried handling and somewhere flat to put things down. A measuring point in the middle of a process offers none of that. In a hazardous area it adds a permit, a gas test and often a second person before anybody has so much as touched a sensor.

The result is predictable and it is not a criticism of anybody's instrumentation team. Calibrations get rushed. They get postponed to the next shutdown. The interval quietly stretches from what the procedure says to what the day allows, and the first anybody knows about the drift is when a laboratory result and a process reading stop agreeing.

There is a second cost that is easy to miss. Every field calibration is time spent by a skilled person standing at a measuring point, and on a site with dozens of them that is a substantial part of a maintenance budget spent on travelling and setting up rather than on the work itself.

What makes bench calibration possible: Memosens

None of this works with a conventional analogue sensor, because the calibration lives in the transmitter rather than in the sensor. Memosens changes that, and it is worth understanding why before looking at the software.

A Memosens sensor couples to its cable inductively, with no electrical contacts at all. The coupling carries both the bi-directional signal and the power to the sensor. Two consequences follow. The connection is completely waterproof and unbothered by dirt, corrosion, salt bridges, overpressure and interference potentials, which removes the single most common cause of unreliable readings in wet and dirty installations. And the sensor is galvanically isolated, so measurement stays clean even where there are extreme potentials about.

The part that matters here is the memory. The sensor head stores its own calibration data, along with its operating time, designation and serial number. The sensor is no longer a passive electrode reporting a millivolt signal; it knows its own characteristics and its own history, and it tells the transmitter when it is plugged in.

Which means a sensor can be calibrated anywhere, at any time, and installed later.

What MemoSuite does

MemoSuite is Knick's sensor management tool for Memosens sensors. It calibrates them in the laboratory, and up to ten sensors can be calibrated at once, which is the difference between an afternoon's work and a fortnight of individual visits on a site running a bank of measuring points on a common interval.

The workflow at the measuring point then inverts. Rather than an engineer taking a calibration kit to the sensor, the engineer takes a pre-calibrated sensor out, swaps it for the used one, and brings the used one back to the laboratory for cleaning, calibration and assessment. Field time drops to the couple of minutes it takes to make the change, and everything requiring judgement happens at a bench.

A practical consequence worth planning for: this way of working needs a small pool of spare sensors in circulation, because at any moment some are installed and some are on the bench. That pool is not a cost so much as a redistribution, since the spares were previously bought anyway and sat uncalibrated in a cupboard until an emergency.

Basic or Advanced: the decision that actually matters

MemoSuite comes in two versions, and the difference is documentation and diagnosis rather than calibration accuracy.

  • MemoSuite Basic calibrates Memosens sensors.
  • MemoSuite Advanced calibrates, and additionally performs sensor diagnosis and database documentation. That database is built to meet GMP and FDA 21 CFR Part 11 requirements.

If you work in a regulated industry, that last sentence is the whole decision and there is not much to weigh. A calibration you cannot evidence is, for audit purposes, a calibration that did not happen. Reconstructing records from a paper log months later, in front of somebody asking precise questions, is an exercise worth paying to avoid.

If you are not regulated, Basic does the calibration job perfectly well, but do not dismiss the diagnosis side too quickly. Knowing the actual condition of a sensor is what separates replacing sensors on a fixed schedule from replacing them when they are worn. On a site with a lot of measuring points, that distinction usually pays for the software difference on its own, because the fixed schedule is always set conservatively enough to throw away serviceable sensors.

What changes operationally

  • Hazardous areas stop being a calibration problem. The calibration happens in the laboratory. The field job is a sensor change, with everything that implies for permits and manpower.
  • Calibration stops being weather-dependent, and stops being something that gets abandoned halfway through because it started raining.
  • Intervals actually hold. A two-minute job at the measuring point is a job that gets done when it is due.
  • Sensor history becomes visible. Because the sensor carries its own operating time and serial number, and Advanced keeps the database, you can see how a particular sensor has behaved across its whole life instead of treating each calibration as an isolated event.
  • Spares become genuine spares. A shelf of pre-calibrated sensors can be fitted immediately. A shelf of uncalibrated ones is a job waiting to happen at the worst possible moment.
  • The skilled work concentrates. One person doing ten careful calibrations at a bench will do them more consistently than five people doing two each in five different places.

Where it fits with the rest of the measuring point

MemoSuite handles the sensor. It sits alongside, not instead of, the other parts of a well-run measuring point.

The transmitter still logs what happens in service, and if you want calibration records flowing into your control system as well as into the MemoSuite database, that is the Protos II 4400 and CalLog, which we have covered separately. The automated cleaning and calibration systems still matter, and if a sensor is fouling badly enough that its life between changes is the real constraint, cleaning is the problem to solve first: bench calibration will not help a sensor that is caked after a fortnight.

The three answer different questions. How do I know this sensor is right? How do I keep it usable in the process? How do I prove what happened? MemoSuite answers the first, and answers it in the one place where careful work is actually possible.

At DP-Flow we supply the full Knick process analytics range and specify complete measuring points rather than parts of them. If you are reviewing how calibration is handled across a set of measuring points, and want to understand what moving that work to the bench would change in practice, that is a conversation we are well placed to have.