as quality engineere, could you please tell me whether statistical process control is useful or not?

as quality engineere, could you please tell me whether statistical process control is useful or not?

I have moved this post to this forum since it is a question regarding SPC.

as quality engineere, could you please tell me whether statistical process control is useful or not?

Usually it is. What is the specific situation you want to use it on?

I have found that as a tool to monitor high volume processes, yes, it is useful. The problem I have with SPC is that companies use it only for data collection and do nothing with the data itself. Often, SPC is a requirement from the customer (H-D is notorious for such requests) to meet Tier 1 responsibilities. I've also seen where the people doing the exercise have no understanding whatsoever of what they are doing. They do it because they were told to. IMHO, I think this is the norm and often a result of some overzealous manager who really knows nothing of the tool.

as quality engineere, could you please tell me whether statistical process control is useful or not?
There is no straight answer to this question without knowing what you are trying to achieve and under what circumstances where the question has arrisen.

Statistical Process Control is generally used "live" to control a process where a fit, form, function and loss function satisfactory distribution is needed or so that there is no need to subsequently inspect 100%.

It can also be used retrospectivley (as Ppk) to study final inspection batch release or even good inwards release with the usual reservations if mixed process batches are assessed.

There are a few circumstances where it might not be appropriate but if it is specified as a customer requirement, you are going to have to really understand and explain your objections.

In my opinion, you have to know what your aims are, and if SPC is the way you must fully commit to do it right and understand the theory.

Generally I have found SPC, and the way of thinking that goes with it, to be useful to most anything. The idea is we are trying to separate the random noise that occurs on a day to day basis (common cause variation) from a signal that something is changing (special cause variation). SPC (in my opinion) is the easiest statistical tool that allows that separation to be made reliably. You may wish to see some of the materials on the Hanford website about SPC at http://www.hanford.gov/rl/?page=1144&parent=169

SPC is a great tool. Great tools imply two things: applicability and knowledge of use. There are great applications for SPC, and some that are not.

Too, there have to be people knowledgeable/trained in applying and interpreting SPC. Otherwise, one may create a worse situation than not having it in the first place.

:agree1: We are a manufacturer of automated assembly and welding equipment and we find SPC a requirement to sell our equipment. Most of the companies we supply to set the CPK limits to runoff to. A very useful and convinsing tool.

In my opinion, you have to know what your aims are, and if SPC is the way you must fully commit to do it right and understand the theory. -martin elliott

I'm with Martin. I have seen SPC used diligently to drive process improvement with excellent results and I have also seen an SPC program result in nothing but an utter, tragic waste of time, effort, and perfectly innocent trees.

Regards -John

In order to fully understand your process, SPC is the way to go. It provides so much information provided you use the data to your benefit. There is no point collecting data because your customer says they want it, use the data (you have it anyway) and see how you can improve the process.

I find a lot of people fire fight, that is they address non conforming products one at a time, yet they could have hundreds of different parts coming through the manufacturing facility and would never find the time to understand and improve the performance of each part. However, looking at the processes used to manufacture each part, and grouping all the hundreds of parts into smaller processes, would enable the user, using SPC data, to address many parts within a process thus reducing the amount of time fire fighting.

Combating the worst performing processes will enable you to hit the majority of parts. But in order to do this, SPC is required to isolated the actual performance of each process.

Well that’s my first ‘post’ finished with, longer then I had envisaged but there you go!

Regards

It's absolutely useful,nevertheless as I know most companies don't implement spc really,they use spc in order to handle customer's audits or requirements.

Generally I have found SPC, and the way of thinking that goes with it, to be useful to most anything. The idea is we are trying to separate the random noise that occurs on a day to day basis (common cause variation) from a signal that something is changing (special cause variation). SPC (in my opinion) is the easiest statistical tool that allows that separation to be made reliably. You may wish to see some of the materials on the Hanford website about SPC at http://www.hanford.gov/rl/?page=1144&parent=169

Good answer, and a helpful link. It might help for the OP to recast the question from "Is SPC useful?" to "Is understanding the variation in my processes useful?" The answer becomes a bit more obvious, I think.

It's absolutely useful,nevertheless as I know most companies don't implement spc really,they use spc in order to handle customer's audits or requirements.

This is true. We have been doing SPC for years and are just now starting to use it to improve processes. In order to do this the operators must understand thew concept. Most think it is just to collect data, which has been the case in the past, so they don't even use the real-time charting in our software. One customer we have is pushing for improvement using SPC which has given us the drive we need to possibly implement Real SPC! Only time will tell if it works out.

When properly applied, SPC provides you a gage by which you can tell if your process is yielding the results you expect - or needs an adjustment. You can live without it - just like you can live without a gas gage or speedometer in your car. But, why would you want to? What harm is there in knowing what is going on with your process??

If we remove ineffective application issues, SPC is without question useful.

In my opinion, the greatest benefit from SPC is to predict future behavior of a process, not to look at where the process was. SPC should be first and foremost used as a prediction tool. This is where the savings and benefits really are with SPC.

My opinion.

Regards,

Dirk

SPC is great for quantitative data (when operators measure and say good or bad), but in case of alternative data with small ppm and 100% control, SPC waste time (when even 2 defects cause point out of limits). Am I right?

SPC is great for quantitative data (when operators measure and say good or bad), but in case of alternative data with small ppm and 100% control, SPC waste time (when even 2 defects cause point out of limits). Am I right?
Yes, in cases where a process is highly capable and has a very good history, a small number of defects close together will cause a signal. That in itself is not necessarily a waste of time. The waste of time comes in as to how the workers and managers respond to the signal.

I also need to point out that with a highly capable process, the emergence of a small number of defects could be very important, and heading off an emerging problem early has many benefits. If you catch the problem at its onset, it is much easier to correct than if you let it run long enough to have a significant impact on your production. And if it runs long enough, it becomes ingrained in the process and is much harder to fix.

Generally I have found SPC, and the way of thinking that goes with it, to be useful to most anything. The idea is we are trying to separate the random noise that occurs on a day to day basis (common cause variation) from a signal that something is changing (special cause variation). SPC (in my opinion) is the easiest statistical tool that allows that separation to be made reliably. You may wish to see some of the materials on the Hanford website about SPC at http://www.hanford.gov/rl/?page=1144&parent=169

Steve, You provided excellent reference information about SPC and other similar/companion areas. I have enjoyed your contributions on various ASQ forums as well. Have a great day. Feel free to contact me with a PM or e-mail!

My company builds high mix low volume products. We used to collect data but it was not used and was a wasts of time. We started to collect "incidents" as they happen daily but we are not sure this will help either. Does anyone have any thoughts on using a simple SPC for high mix low volume manufacturing?

Yes, in cases where a process is highly capable and has a very good history, a small number of defects close together will cause a signal. That in itself is not necessarily a waste of time. The waste of time comes in as to how the workers and managers respond to the signal.


Steve:

I am a bit confused on your answer. Do you mean to say that a small number of out-of-control product close together will cause a signal rather than "defects".

Steve:

I am a bit confused on your answer. Do you mean to say that a small number of out-of-control product close together will cause a signal rather than "defects".

The time between defects can be much more sensitive (and effective) in detecting a change in defect rate rather than simply counting defects per time interval. See http://www.hanford.gov/rl/uploadfiles/VPP_TrendLow-Rate.pdf for an example which I took from Dr. Wheeler's book "Understanding Variation - the Key to Managing Chaos" (a very good introductory book on SPC, I recommend it highly).

The time between defects can be much more sensitive (and effective) in detecting a change in defect rate rather than simply counting defects per time interval. See http://www.hanford.gov/rl/uploadfiles/VPP_TrendLow-Rate.pdf for an example which I took from Dr. Wheeler's book "Understanding Variation - the Key to Managing Chaos" (a very good introductory book on SPC, I recommend it highly).

Steve:

Sorry I misunderstood you. You are talking about attribute data rather than variable. In the automotive supplier base, attribute charts are rarely used.

Steve:

Sorry I misunderstood you. You are talking about attribute data rather than variable. In the automotive supplier base, attribute charts are rarely used.

Definitely - if the defect is defined by a measurement, you are much better off plotting the measurement, rather than just analyzing the defects.

A bit more input: rare defect rates where the average count is less than one will produce an out-of control signal when the one rare event occurs. This is obviously a false alarm. Additinally many rare event conditions tend to 'cluster'. this is when 2-4 defects happen within a relatively short time (single defects with a small number of good events in between.) This is typically followed by a relatively large number of good events. (you know how "celebrities die in 3s", people have "Winning streaks" and accidents seem to come in bursts? It's not always true of course but if you plot these things out in time sequence you'll see that clusters definitely happen. Now either there is an assignable cause for the clustering or it's just the nature of random events. In any event, a better way to chart rare events is to use the number of good runs between a defect and plot that number. This distribution can be modeled by the geometric distribution (mode near zero with a long tail to the large side...)

I use the following formulas:

calculate the average RUN (it's the inverse of the defect rate by the way)
The center line of the chart is .7 * average RUN.
The UCL = (alpha/2)*average RUN
The LCL = -Average RUN*LN(Alpha/2)

This works really well - I've used it successfully for many years.

Reference: Goh, T. N., “A Statistical Procedure for Defect Control in High Quality Manufacturing”, Sensors Controls and Quality Issues in Manufacturing, ASME 1991, pp395-401.

Yes It Is A Great Tool But Only If Used For The Long Run. It Can Infact Bring Some Degradations As Well If The Results You Expect From A Shorter Period.

That is a great question!

SPC is useful for process optimization (Design of Experiments). After going through the SPC steps, the process is characterized well enough to set parameters and have them stick. SPC simplifies DOE randomization (long story).

I would encourage any reader that has a solid SPC program to cash in their chips and use DOE.

I have found that process engineers use SPC. They call it calculating "deadband".

Hope this helps,

Tom