Beyond PDCA by Praveen Gupta

[qualitytrec]

Steve I agree with you. Thanks.

Jim,
Actually, I have not worked in a place yet that values control data nearly so much as we here at the cove do. I believe most of it is a problem of not truly understanding the purpose of control and the charts. We have at my current employer a situation where weight is one of the key indicators that something is going wrong in the process. Our weight logs show a consistant lack of attention to even the out of spec product not just the out of control points. Ask a technician what the key indicator is and they will all say weight. Ask them why they did not deal with the situation and they will answer it did not seem like a big deal at the time they had other fires. Probably a bad example but I see charts made all the time in many companies and is just paper. Many times they just are not used. sometimes it is because they are not understood sometimes something else. I would bett there are alot of trees hanging on our walls that never needed to be killed. And the ones that did die for a cause are almost in vain also.

But to get back on topic. I believe PDC/SA if used just like the charts mentioned above would work for almost any business that truly uses and understands the methodology.

 

[Jim Wynne]

Ask them why they did not deal with the situation and they will answer it did not seem like a big deal at the time they had other fires.

I've seen lots of useless charts as well, but I was referring to actually doing statistical analysis, not just saying that you do. As for "other fires," it seems clear that weight is not a "key indicator" or is only a key indicator if its place on the fire-to-put-out list is high enough. Important attributes or characteristics can't be established in a vacuum, or by simple fiat. If operators's priorities lean in another direction, the outcome is preordained. Operators will naturally give priority to things that have the potential to cause them (personally) the most trouble. Try asking water flowing downhill to follow some path other than that of least resistance and you'll get a similar result.

 

[qualitytrec]

in response to other fires you are absolutly correct. The techs tend to try to get as many machines running at once as is possible. This means sometimes they "ignore" another problem in the interest of fixing the things that are less complex to boost production numbers. There are several systemic issue involved in this, resources and training are probably among the greatest. But I really did not mean to hi-jack this thread.

 

[Steve Prevette]

But I really did not mean to hi-jack this thread.

Well, I'm glad to see some real-world context for this discussion.

A "Best Practices" conference was held last week in Colorado by the US Department of Energy. An Assistant Secretary of Energy came out strongly in favor of my work being done here at Fluor Hanford, and he challenged the other DOE contractors to follow suit. He also stated that he has directed his direct reports that if they have data to present to him, it better be in a control chart.

It's going to be interesting to see how this plays out.

 

[ralphsulser]

Alright Steve, way to go

Glad to see one of our "Cove family" get deserved recognition for doing a good job

 

[Wes Bucey]

In your opinon, folks, is there any probative value in inviting Praveen to join this discussion? I know him by sight from being a member of my ASQ section, but I know two of his brothers, Kam and Pradeep, better. There are five Gupta brothers, all of whom are (or were) long time members of ASQ.

It may take me a while to connect with him, but I will make the effort if you believe this can be a serious discussion and not a slamfest. I would not like to be embarrassed by inviting a colleague to be the main course at a barbecue.

 

[Jim Wynne]

Letting him know that his article is being discussed is not equivalent to leading him to the slaughter. If he wants to discuss it, he will.

 

[Helmut Jilling]

Some companies might not even ship a product which exceeded statistical control limits - likely Toyota would not use the part. But even if you choose to ship the product, if there is a statistical signal that something has changed, then it should be investigated, not ignored.

I have to agree with Jim, and several others. Not shipping the part seems unnecessary, and Toyota would likely not even be aware whether it was out of control (your example said the process was much in control overall).

However, I would agree, investigation should be made to determine if adjustment would be beneficial. It should not simply be dismissed out of hand.

 

[qualitytrec]

I do not really see the differences (except in symantics) that Gupta is claiming. What really is the difference between Prepare and Plan, Perform and Do, Check/Study, and Perfect (defined by Gupta as "...assessing whether the process performed as planned..."). I see no change to the current cycle except the change of names to protect the innocent or uninformed.

 

[Jim Wynne]

In your opinon, folks, is there any probative value in inviting Praveen to join this discussion? I know him by sight from being a member of my ASQ section, but I know two of his brothers, Kam and Pradeep, better. There are five Gupta brothers, all of whom are (or were) long time members of ASQ.

It may take me a while to connect with him, but I will make the effort if you believe this can be a serious discussion and not a slamfest. I would not like to be embarrassed by inviting a colleague to be the main course at a barbecue.

I just posted a fisking of the Gupta article here, and notified the author of the posting, and the discussion here, through his business e-mail address.