Hi I am struggling with finding the best way to carry out an attribute MSA.
The problem is that for this product we only produce approximatly 10 items a week with 2 operators. RFT's are low and are out the door as soon as possible. Im am only at green belt level and any guidance would be helpful.:o

Hi I am struggling with finding the best way to carry out an attribute MSA.
The problem is that for this product we only produce approximatly 10 items a week with 2 operators. RFT's are low and are out the door as soon as possible. Im am only at green belt level and any guidance would be helpful.:o

Hello White Stork,

Welcome to The Cove! Glad you're here! :bigwave:

Not an easy question because of the low volume you manufacture. By the way, does your customer require MSA? For starters, you may want to explore the links at the bottom of this page. "Low volume" seems to be the key term here. I hope that my Fellow Covers will join soon.

Stijloor.

Hello Stork
Why you want to run the study, what do you expect from it ?
What are your parts, what is their coplexity, are these 10 parts aiplanes, or pencils ? Give us some more details, to be able to give you some inputs.
In one word, what is your problem ?
Thanks and best regards
György

Hi All
Thank you for the replies.
I am taking part in a Six Sigma training program and as part of the program I have to do an MSA. These products are automotive rear lenses (very exspensive luxury cars). The problem is that lens inspection is completely visual. :bonk:

Hello Stork
Intresting situation and task.
Was the person aware of all the above when he gave you the task ?
On the other hand:
- is this process running for a while, if yes did you have quality problems ?
- do you have scrap parts out of those 10 pcs produced weekly ?
- if yes can you collect them or they will go to rework
- if the parts go to rework, is this done contiuous or you collect them, and rework all at same time
- is any failure reason specific for this product, or the failures are similar to other product you run in shop
Sorry that's all at the moment, but maybe we can figure out something.
BR
György

Hi All
Thank you for the replies.
I am taking part in a Six Sigma training program and as part of the program I have to do an MSA. These products are automotive rear lenses (very exspensive luxury cars). The problem is that lens inspection is completely visual. :bonk:

Hi White Stork. Welcome to the Cove,
I believe one of the keys for you to do that is to shift your paradigm from studying the process itself that makes the lens to that of studying the capability of your inspectors to be consistent in thier visual judgment. With that i suggest you run an attribute MSA.
You can search the forum on this for starters.

>>rey

Hi Thanks for the mails.

Gyorgy:-
further to you mail. Yes the process has been running for a couple of years. Yes there are scrap however these are continually reworked and not collected together. there are many reasons for failures, these include scratches particals buffing marks actually there are a lot of reasons that these items can fail. We do have some similar product however they do not have the same customer requirements and the lenses are hard coated so they are less problematic.
Im not convinced that an attribute MSA can be done.

Rey:-
This is what I am trying to do but with such low volumes and the number of potential problems im not sure if an MSA can be done.:bonk:

Hello Stork
If the process has been running for years:
- what is your internal scrap or rework which is "supposed" to be due to inspection inconsistencies of operatoprs, if you have such a data
- what is your process variation history in terms of yields, first pass and after rework
- what is your external DPM, and do you have customer complaints
Some of the answers to above questions may support the idea that there is no need for an MSA study(just my opinion, in some cases the performance data history may be used as "process validation"
If you or your boss wants to run an MSA, the he/she should help in organizing this.
You may collect those 10 part production once, run on them a standard inspection by the most skilled person, this will be the standard, then take the operators for each shift together for one day, and run the study.
In my opinion, you can run the study on a different product, it will still give a good reference let say 70-80% about the operatrs skill and performance.
Just my opinion.
Good luck
György