Before carrying out measurement uncertainties, there is the process of consideration! Whether uncertainty according to VDA5, according to the GUM or whatever: uncertainties are not mass MSA, not just different maths for the same measured values.
The author has received many enquiries along the lines of "If you install 500 type 1s and type 2s, does it take that long?" and the answer has been yes. Of course, but why do you do that?".
Measurement uncertainties start with a piece of paper and a pen. What do you want to achieve? What components do you need to take into account? Is the time of year a problem if temperature fluctuations have an effect? (Not every measuring device is located in an air-conditioned environment). Is the time of day a problem because incident daylight affects the optical measuring device? How are the components cleaned before the measurement process, and is it always identical? Type-2 Study compares the X repeat measurements of X operators against each other, but should one take 30 consecutively produced components from production and give them to 3 operators, 10 pieces each, each cleaning their components themselves beforehand?
In his time as an engineer, the author always started to determine measurement uncertainties by sitting on a chair, observing the operators as they acted, and then analysing the data measured so far. Are systematics already visible there? Can effects already be observed through recorded "boundary conditions" (in Q-DAS parlance: additional data)?
And only then do the technical details come into play: Repeat measurement, comparative measurements, Linearity....
This is an effort! And can be a greater effort per characteristic, per study. Therefore, the characteristics that are then analysed in a Measurement uncertainty study should be precisely defined. Those with the smallest specification limits, worst measuring point (because, for example, all 3 axes have to be moved on a coordinate measuring machine to measure a feature), and so on. Not simply everything. Because users who, after importing hundreds of MSA characteristics (which is unfortunately widespread), are faced with hundreds of Measurement uncertainty studies, will then stop.