Ravi's Blogs Measurement Systems Analysis- A management perspective

Ravi Khare submitted a new resource:

Measurement Systems Analysis- A management perspective - thoughts on MSA

Read more about this resource...

 

Thanks for the review Andy.

You are right; the measurement process needs to be examined in the perspective of the measurement being made, not just tolerances.

To expand upon a few thoughts in this context…

Every measurement process has a character. How complex is the measurement process? Is it governed by external noise factors that will impact correct evaluation? There can be simple evaluations like measurement of distances between two features, while others may be fine measurements obtained by complex electronic circuits. Such complex measurements may get seriously impacted by the proximity of high voltages and electromagnetic spikes. In such cases robustness (ability to sustain external variations, shocks, vibrations) of the measurement system would become a driving factor for its selection.

The criticality of application would be an important aspect to consider. For instance a medical measurement system (or method) that detects the presence of a life threatening condition in humans would need to have a much lower miss rate than the more routine ones. There cannot be a blanket beta error that is mandated by a standard for all applications across the board. Each type of measurement system in such cases would deserve a careful individual in-depth evaluation.

I like to look upon Measurement Systems as satisfying one of the following two purposes:

• Measurement systems which are used for acceptance of good products. These classify good from not-good.
• Measurement Systems which are used for process centering.

The ‘acceptance’ kind of measurement systems would look at the design tolerance to make a decision. The measurement risk I would look at is the width of the gray zone around each measurement. This assumes importance when the measurement lies on the border of the tolerance zone.

The process centering measurement systems however need to however have a good resolution. It is important to determine the number of discrete steps the measurement system resolves the process variation into (number of distinct data categories, in technical terms). The better is the resolution capability the better control you have over the process. A good capability here will lead to a better process assurance, reduce inspection and prevent over-adjustment of the process.

ISO 9001 requires suitable equipment. What better way to demonstrate and document suitability than the scientific method of MSA? I second you on this.

Yes, it’s thus not just about measurement system variation ‘eating into the tolerance zone’. There are governing considerations that stem out of the criticality of application, complexity measurement process, demands of the manufacturing process that is measured and the purpose of measurement that you have in mind.

Thanks again Andy, for the nudge. Without it, these thoughts would not have come out and the discussion on MSA would have been inadequate.

I’m sure there is more on MSA that we have not thought about, and a lot of MSA that we don’t know of. Further discussions by participants of the forum, and finer points brought out by experts will help us learn.

I request Atul to move this thread to the MSA forums if he feels it appropriate.

 

I don't know how to reply to Ravi's reply...so I'll post it here...

My words were misleading...my manager reports to me...he is not "my manager".
All the same, he too thought it was a good overview to "GET IT" about the role of MSA.
Short on the details, but nice coverage of the purpose.

 

This is a well written article on MSA. The figure under Why MSA? tells the story!