Is SPC (Statistical Process Control) useful?

williams22 said:

Statistical Process Control is useful as you are quality engineer because it is used to check whether the product is meeting the specifications or not!!

Click to expand...

ummm. actually it isn't . Statistical Acceptance Sampling is used to determine if product meets specifications.

SPC - or more specifically control charts which are the intended focus of this thread - is used to determine process stability and can serve as a basis for capability studies (which may be at the heart of the confusion).

Capability studies are intended to determine if the process is capable of producing product that meets the specifications adn just how capable that process is relative to the specifications.

jasonb067 said:

Those most knowledgeable in SPC are doing very little good if they do not understand the process to which it is applied, (unless they are teaching or showing others the concepts).

The most knowledgeable about the process tend to not have the most in depth of SPC tools and all that they are capable of.

When you get both in one person then we can see the benefit fully.

Click to expand...

jasonb067 said:

Those most knowledgeable in SPC are doing very little good if they do not understand the process to which it is applied, (unless they are teaching or showing others the concepts).

Click to expand...

Yes! I agree, due to the fact is, the Spc technique is quite "abstract"' so those people likely to be teacher, consultant, auditor ... Who is quite "far" to the process and can input to improve it. So, the benefit from those people is just on "theory" only. In my deepth thinking, the society had spent a lot of expenses for those lectures, printings, time ... And money - but get back very little!

jasonb067 said:

The most knowledgeable about the process tend to not have the most in depth of SPC tools and all that they are capable of.

Click to expand...

Yes too! As u can see, those people are engineers on the process, they are quite busy with those daily workload, that could bring them more "real" pressure & benefit. Why they have to spend their valuable time into those bundle of theory and the payback = unpredictable & too small. That is the reason, most of them are not interested in SPC. The fact is also that, they are cleaver enough to understand the SPC - but don't want to.

Your ideal situation "when those two come in one person" will never happen due to above reason: the clever guy will not spend too much time on too little return!

If you don't think so, just to do a test: who's well explaining here and there in SPC - i believe that, he/she is a teachers, consuntants, expert ... But not process engineer, that is why no one can give me the sound example of real benefit for ceirtain period of time. Only some example for a certain moment.

Sorry to state so, but it is the fact! So i would think that - who, as statistical expert, need to find out more practical and useful tool to serve the production than those control charts!

Effocus said:

Yes! I agree, due to the fact is, the Spc technique is quite "abstract"' so those people likely to be teacher, consultant, auditor ... Who is quite "far" to the process and can input to improve it. So, the benefit from those people is just on "theory" only. In my deepth thinking, the society had spent a lot of expenses for those lectures, printings, time ... And money - but get back very little!


If you don't think so, just to do a test: who's well explaining here and there in SPC - i believe that, he/she is a teachers, consuntants, expert ... But not process engineer, that is why no one can give me the sound example of real benefit for ceirtain period of time. Only some example for a certain moment.

Sorry to state so, but it is the fact! So i would think that - who, as statistical expert, need to find out more practical and useful tool to serve the production than those control charts!

Click to expand...

In my experience, the workers intimately involved with the process "get" SPC quite easily, and already had an intuitive feel for what the chart quantifies. They see the adverse results of tampering by management to react to random noise - and just make things worse.

Granted, there are some folks who are not good teachers. I would just suggest take a look at the following series of youtube videos and see what you think:

(broken link removed)

Steve Prevette said:

Normality tests (at least if you follow Dr. Shewhart's original work, and as expanded upon by Dr. Wheeler and Dr. Deming) are NOT required for use of SPC.

Click to expand...

Yes, what is far more critical is if the data are from a process whose stable state is non-random, continuous or random, discrete distribution. Shewhart's charts are based on random variation as the stable state. Not all processes fall into that category. But...just because Shewhart's charts do not always fit that use, does not mean there is no SPC technique that does.

Just wondering..is there any statistical method to perform on one piece part measured few times?..sorry I might post in a wrong thread.I have not really figure out how to open a new thread.Thanks

Tobuakar said:

Just wondering..is there any statistical method to perform on one piece part measured few times?.. sorry I might post in a wrong thread. I have not really figure out how to open a new thread.Thanks

Click to expand...

(broken link removed)

Effocus said:

Yes too! As u can see, those people are engineers on the process, they are quite busy with those daily workload, that could bring them more "real" pressure & benefit. Why they have to spend their valuable time into those bundle of theory and the payback = unpredictable & too small. That is the reason, most of them are not interested in SPC. The fact is also that, they are cleaver enough to understand the SPC - but don't want to.

Your ideal situation "when those two come in one person" will never happen due to above reason: the clever guy will not spend too much time on too little return!

Click to expand...

The most profitable compaines I have worked for have process engineers that understand and use SPC to improve their processes. They are not fighting to keep the line running they are using SPC to identify waste and variation and then drive them to their lowest controlable point. If you want specific measured examples send me a message through my profile.

To the process engineer that is too busy. If you don't understand or want the benefit of the SPC tool then don't use it. You can continue to be too busy doing real work "fixing your process".

The truth is the process engineers that can not use SPC as one tool in their tool box get into a position where they know everything about their process except how to make it better. They spend every day fighting fire but no real improvement

I hope that you too have the ability to fully understand and apply SPC to actually improve your process.

Again, don't become one of many that knows everything about your process execpt how to make it better.

Good Luck!

Steve Prevette said:

Granted, there are some folks who are not good teachers. I would just suggest take a look at the following series of youtube videos and see what you think:

(broken link removed)

Click to expand...

Hi Steve, thanks for sharing the video. However, this is same as E. Deming done in last centery for p - np chart. Did u notice that the level of NC product is too high for the current level now of production and it is not so now a day condition!

Effocus said:

Hi Steve, thanks for sharing the video. However, this is same as E. Deming done in last centery for p - np chart. Did u notice that the level of NC product is too high for the current level now of production and it is not so now a day condition!

Click to expand...

Most definitely - that is my hope that I have maintained the standard of Dr. W Edwards Deming!. Yes, he used the np chart, I have simplified it to the c-chart (which makes very little difference - if we must be exact - it should be the hypergeometric, but then again, as Dr. Deming pointed out, this is NOT a pure random situation, it is random mechanical sampling).

A point of the Red Beads is that the proces itself is Non Conforming (I assume that is what you meant by NC). No amount of goal setting, rewards to the willing workers, chastising the willing workers, etc etc etc will change that. This is the lesson of the Red Beads.

Effocus said:

Hi Steve, thanks for sharing the video. However, this is same as E. Deming done in last centery for p - np chart. Did u notice that the level of NC product is too high for the current level now of production and it is not so now a day condition!

Click to expand...

In addition to Steve's comments, I would mention two things about the defect rate issue you raise.

The first of course is that red beads or defects are used because it's easy for the demonstration and it makes Deming's point about blaming the operators for defects that are an inherent part of the system. The defect rate itself is irrelevant.

And there were many processes - even in the 'old' days - that had very low defect rates and they either got that way in part thru the use of SPC OR they stayed that way due to the use of SPC to provide early warning - using continuous data - to prevent high defect rates.