8.1.1 Identification of Statistical Tools

If you determine that statistical tools are not applicable to your manufacturing process, how is the best way to documenting this to be protected during a third party audit.
Please provide your feedback.
Thanks

 

Statistical tools may reach beyond just SPC. You may do sampling at incoming, you may do sampling if you don't do 100% inspection, there may be some statistical basis for determining when problems are systemic or trigger field action, etc.

 

Anerolc
Why do you say that dont apply statistics?
Whatever business you are in, have lot of data to analyze.
Rejects, nonconformities,complaints,production rates,etc.
Well on this, its possible to use statistics at least at its minimal expression.
Pareto, histograms, std deviation,etc.
Regards

 

Statistical process control (SPC) may not be utilized in every application, that is a specialized tool suited for certain conditions, however data analysis does utilize a statistical approach whether you make a product or perform a service. Qualmx hit it correctly, there is much data there, but to actually make sense of it some statistical approach is necessary.

 

Hi Anerolc,
As our dear forum members pointed out that getting away from some kind of statistical approach or not applying statistical approach for your manufacturing process is not practical possible,you may be applying statistical approach in some form or the other failing to acknowledge, kindly review your process with statistics in mind . Please note that if you are a medical device company manufacturing you have to comply with 21 CFR 820.50 Statistical Technique, I don't believe that you can apply exclusion in your QMS for that particular Reg.

 

Do you never make rejects / rework?
Do you ever measure efficiency and effectiveness?
Do you ever do customer surveys or VOC analyses?

I am trying to understand how one cannot use any SPC per se in their process.

As they say 80% of people dont understand stats the other 31% do.

 

Good point, Tony! Even if all you have is attribute issues, there are attribute methodologies to determine if you are heading off track or not!

 

Thank you for your feedback, I was a little bit biased about to see only control charts in manufacturing floor as spc, however there are 7 tools to be applied right?, I think the issue Is that personnel is not really convinced to use them because they are custom the old way (sampling- 100%inspection) , also I have realized that we have variable data were control charts can be applied in the manufacturing process. How can I convince them to use it because is it is a shall of ISO TS 16949?

 

Depends on the variable. I can convince people to use charting in machining because it can tell them when to make adjustments and when to change tools (in some cases). It actually saves them work and scrap. It doesn't work with Shewhart charts, but their are statistical charting methodologies that do work. So, the key is to use the correct chart with the correct measurement system (adequate resolution to detect change) to provide useful information. The point of charting is to detect when the process needs attention. Shewhart charts detect when the process sends a signal that it is changing abnormally. It is like a gas gage in a car. Sure, you can drive a car without a gas gage - but isn't it clearly better to have that information?

 

Thanks Bob. Hopefully they realize of the importance of using tools correctly and don't see them just like a burden.