How to reduce the process SPC monitoring

Alba Godinez

Registered
Hi to all
We are in process to reduce the dimensional monitoring through SPC process to all the part number that show a stability in the manufacturing process according to some conditions defined in my site. But one of the condition that is necessary to define is:

What is the optimum CPk/PPk value needed to can decrease the frequency of characteristics monitoring?
 

Jim Wynne

Leader
Admin
We are in process to reduce the dimensional monitoring through SPC process...
Can you clarify this statement, please? Are you saying that you're using SPC to justify a reduction in dimensional inspection?

What is the optimum CPk/PPk value needed to can decrease the frequency of characteristics monitoring?
There is no such thing as an optimum value, and Cpk/Ppk are not good indices to use in a case like this. You should be able to do what you're seeking to do just by examining the charts.
 

Alba Godinez

Registered
Can you clarify this statement, please? Are you saying that you're using SPC to justify a reduction in dimensional inspection?

We will continue using SPC (and we use SPC for IAFT requirement), but we want to reduce the frequency in which we monitor the process, can you advise how I can do it o what we need to consider before implementing?
 

Ninja

Looking for Reality
Trusted Information Resource
I've done this...here's what I did (no guarantee it was the best way).

I had years of data, and considered cutting process monitoring measurements in half...so I clipped out every other data point in the historic data and checked to see if the "warning flags that mattered" would have been tripped.

For those flags that did not trip, I considered what the cost would have been for the quality excursion...then compared that to the cost savings from decreasing monitoring.

You want to cut down monitoring to save money...so money is the basis of any decision made moving forward.
The one line I did NOT consider easing up on was final testing...the cost of shipping NC product to the customer was far too high.

Then we cut monitoring in half for a month...verified whether the anticipated cost savings was real, and verified process stability in final check.
Then did this route again.
Then cut monitoring frequency in half for good since it seemed to be safe.

Start with the data you already have...and play make believe (what if I only had some of this)...and see what the data says.
HTH
 
Top Bottom