Linking continous monitoring to calibration

[Hendi]

Hello,

we are a company at MDR/ISO 13485 landscape and as such, we monitor our measuring equipment as we have to.
In the past this was done periodically, depending on risk-based requirements, by accredited calibration labs or for lower-tier stuff in-house with documented and recorded processes, against standards or against reference meters. Intervals are derived from useage, stability between periodic calibrations and generally how well stuff behaves (and somewhat price) also according to a documented process.

To gain confidence in a bunch of meters and loggers I started to introduce online monitoring campaigns, mostly guided by "there is a computer readout for it" and hierarchy. This includes e.g.

• a display and logging program for room climate loggers spread in the building that periodically records deviation of air pressure between loggers,
• another program logging atmospheric pressure reading from one of the low-tier bunch above against a highest-tier pressure gauge (recognizing the effect of height-dependent barometric pressure)
• The same highest-tier pressure scale is periodically compared against a medium-tier device at low, high and midpoint.
• Other example are placing transport loggers on top of the room logger and record temperature deviation after a equalisation time, or
• sticking a higher-tier Pt100 Thermometer in a heating cabinet,
• and so on.

The main reason to do this to have a early warning system, triggering process evaluation and inspection of tools/gauges.

How do I link the results of those inline/online/continuous (of course you can consider it that after trials this are all very well documented and validated processes) data to the periodic calibration?

Are there guidance or standards about

• deriving e.g.calibration interval from monitoring?
• Is "inheritance" possible if specified?
• Replacing a inspection+measurement session by "just evaluate last values from the log and call it a day if it fits requirement X" ?

I'm clear that it is our requirements and needs to be risk-based, but is there anything I should read before thing about those things?

BR
Hendi

 

[dwperron]

Where I see problems with your plan is the Traceability requirement in 13485 section 7.6 :
"As necessary to ensure valid results, measuring equipment shall: be calibrated or verified, or both, at specified intervals, or prior to use, against measurement standards traceable to international or national measurement standards"
You are comparing high tier instruments against lower tier instruments, that does not support measurement traceability. In fact, I don't see anything about using equipment that was calibrated outside of your organization, thus there is no traceability to a national lab.
Yes, what you have here is fine for a check on the status of your instruments to catch potential problems, and if you have set up such a system I would continue it. But this is no substitute for a calibration program.

 

[Hendi]

The clarity might have been lost in translation.

In the past this was done periodically, depending on risk-based requirements, by accredited calibration labs or for lower-tier stuff in-house with documented and recorded processes, against standards or against reference meters. Intervals are derived from useage, stability between periodic calibrations and generally how well stuff behaves (and somewhat price) also according to a documented process.

"In the past" might be misleading, the calibration program didn't end.
"High tier": done by accredited labs outside, like standards and reference instruments.
Lower-tier stuff: it depends.
It's all based on our idea of required accuracy, importance, useage, treatment, calibration history, if we have necessary standards and competence to do in-house and behaviour compared to the rest of the instrument family (monitored by the "early warning system"), and documented in the test plans. The intervals of external monitoring might be long. There are our documented decisions why. (Having a published or standardized basis for this would be nice, tho.)
Actually, we look not just for presence of a calibration certificate when the device comes back but evaluate it (and boy, did find things worth asking, even with DAkks and ILAC marks.) - this is another place where we should catch if we messed up family members of the device.

Ohhh... maybe this "linking" I asked for should not be done in the first place. Calibration intervals are the one thing, and everybody not blindly following the 12 month rule but caring for it has to suffer for it. Tight monitoring for bad gauges between calibrations is a independent good thing, but can contribute as one input for the first.

While not pursuing accreditation or claiming to be fully equivalent, we are somewhat competent in measurement. And willing to improve. Even asking for this gave philosophic insight.