April 2006

"Quality Focus" - monthly Genesis column

Blind Faith in Measurement

Don't trust dimensional data until the process of measurement has been removed as a source of variability...

So far in this column this year we began to grapple with the concept of “quality” and started to look at some quality definitions – or more appropriately – quantifications. 

What we have not yet considered is the subject of “measurement,” which is of HUGE concern when considering “quality.”  Why?  Do not we simply measure our process, manipulate the data, and discover the extent of the variability?  Well, it is not as straightforward as that. 

Actually, measurement systems themselves are RESPONSIBLE for a portion of the variation under consideration.  Always.  No exceptions.  Only the magnitude of such error differs for various measurement systems under consideration. 

Why is this?  Are not calibrated gauges always accurate?  Precise? 

In a measurement system, calibration is expected (we will not delve into this subject here).  However what we are considering here is repeatability and reproducibility.  But is this not the same as calibration?  Absolutely not. 

Repeatability is concerned with the ability of one operator – or in the case of on-line measurement, one system – to repeat the same set of measurements on the same parts - or in a broader sense, the same cans.  Of course if we present an operator with say ten cans and ask they be measured 3 times each, repeatable measurements are inevitable if the cans are measured with any pattern. 

However, if the cans are randomly measured, without operator knowledge of any identification whatsoever, the measuring results will most often take on a totally different dataset.  Depending upon the discrimination of the gauge (how “small” a dimension it can measure), one can rest assured the measurements will not repeat.  It is only a matter of the magnitude of this measurement variation – and, once again, gauge discrimination. 

Conversely, reproducibility is concerned with multiple operators and multiple sets of parts – or cans.  One way to tell if there is a reproducibility issue in a plant is when only a certain individual is “allowed” to make a particular measurement.  All trained operators -  or in-line systems – should return nearly the same measurements.  It is important to note that we are considering not only the gauge, but personnel and the environment.

We then can elevate from a simple (and crude) “calibration” assurance to a Repeatability and Reproducibility study; or a “Gauge R&R.”  This is the critical factor in a measurement system, as it considers all that affects that system. 

The following is a very rudimentary repeatability test for an on-line system, in this case for cans.  Please note this is not a statistically valid experiment, but merely an illustration to drive home a point. 

Choose 30 cans from the line, mixing “good” cans versus what the on-line system has rejected as “bad.”  To make the test more robust, do not choose “really bad” cans, but a few that are “marginal.”  Run the cans 3 times through the inspection system at 3 different line speeds each (9 runs).  You may be surprised at the results.  Did the same cans get blown off each time?  How many were different? 

Is Gauge R&R really so important?  Without its application – on almost any measurement system – it is not possible to know the true value of the measurements being collected.  Many problems are actually exacerbated by measurement systems, especially with very tight tolerances. 

Given the high precision involved in can manufacturing, it is little wonder why so many production problems linger for years without resolution.  Conversely, the machinery and supplies used in canmaking are designed to very tight specifications, as well.  Without a proper Gauge R&R – as part of a Measurement Systems Analysis - one is essentially “measuring in the dark.”