The VDA-5 standard is a big step forward for the evaluation of industrial measurements. It combines the practicality of gage studies with a rigorous uncertainty evaluation approach. This is definitely heading in the right direction, although the methods it uses can still miss some important sources of uncertainty, as I will show towards the end of this article. Read the full article, published on engineering.com…

*Related*

The hybrid MSA Analysis interests me and your paper brings it to life, thank you for that. It’s always healthy to debate these points, some observations if I may:

It’s true that the AIAG MSA manual seems to be the reference in Industry, certainly for Automotive but not necessarily Aerospace for example. Another technique called EMP was introduced in the late 1980’s which is certainly used in parts of Airbus, I know this because I introduced it and have trained many companies in this approach.

There is a lack of respect for the concept of data homogeneity in both the Uncertainty approach and standard AIAG MSA. I have seen measurement process precision improved by as much as 60% just by being aware of and fixing simple aspects of the process, this is what metrologists and engineers should and are doing every day.

Combining what is essentially a modelling approach with an empirical approach needs to take into consideration the essential practicality of an industrial measurement process. For example, we take a measurement, ensure that the data is homogenous this in turn will represent predictability and therefore we can begin to characterise. Once we have homogenous data and a fair representation of our experimental conditions considered by the metrologist the repeatability represents the best that we can achieve WITHOUT changing the process in a way which would be expensive and time consuming.

We are left with the accuracy which yes will include the uncertainty, but the real question of interest to operations is can I live with this value or do i have to compensate my measurements?

Lastly, both these approaches use the standard deviation as the estimate of measurement variation. In AIAG MSA, there are ratios constructed which implicitly assume that there is a linear relationship between the measurement and the part variation yet how can this be true? They are actually implying that if the measurement variation changes by 1 unit the same is true for the product variation, clearly this is nonsensical.

I fully support that careful consideration of all sources of variation when studying any variable is important. I also fully understand that the methods tend to overestimate the effects of measurement variation on production parts, the key to using them is to be aware of this also aware of what else is out there to take this into consideration.

If the outcome of an approach or method is constantly forcing you to implement expensive changes which don’t even make sense to experienced engineers it should lead you to question the accepted status quo. This is certainly true of the approaches outlined in this paper.

Thanks

Geraint Jones