posted 24 July 2001 10:49 PM               

*********************

From: Nancy Jennejohn
Date: Mon, 23 Jul 2001 22:55:13 -0500
Subject: Re: Elem 14 related to employee suggestions /../Scalies/Bauer

From: "Russell Bauer"

Charley wrote:
> I had a registrar's auditor tell me that since the new standard requires
> procedures for CA and also PA, he expects to see 2 separate procedures.

I guess we are the strange ones, but I prefer 2 different procedures. This helps everyone keep the two processes separate. I have been trying for a long time to emphasize the practice of preventive action that is outside and not related to corrective action. it is a true art I believe, and preventive action must remain a procedure all it's own. Just my opinion...but strengthened with every external audit we have.

R. Bauer

*************************

From: Nancy Jennejohn
Date: Mon, 23 Jul 2001 22:59:17 -0500
Subject: Re: Elem 14 related to employee suggestions /../Scalies/Holtz

From: "John Holtz"

Charley Scalies wrote:

> I had a registrar's auditor tell me that since the new standard requires
> procedures for CA and also PA, he expects to see 2 separate procedures. I
> have a high regard for this guy - he usually has his head screwed on
> straight - so I figure he must have been off his feed that day.
> Has anyone else gotten the same silly "advice"?
>

Two procedures...one procedure...I can't imagine that being the issue with an auditor...but...if as an auditor I did not see a procedural differentiation between corrective and preventive action, I'd ask for the organization's understanding of the two concepts and follow the links or lack thereof to a process. The differences are subtle. 9000:2000 differentiates them by "prevent occurrence" (PA) and "prevent reoccurrence" (CA), but goes no further.. 9004:2000 implies that PA is data driven for potential events while CA is driven by root-cause of actual events. Hmmmm. Actually, 9001:2000 may give us the most information: from 8.5.2 "...take action to eliminate the cause of nonconformity"; From 8.5.3, "...determine action to eliminate the cause of potential nonconformity." Here's the tree-falling-in-the-woods conundrum as it relates to CA/PA: If in your corrective action you do what is necessary to assure that nonconformance will not happen again, is that a preventive measure or a corrective measure? According to what I've heard in 9001:2000 interpretation, that is CA, even though it "sounds an awful lot like" PA.

Sooo...IF that is CA, then what, pray tell, is PA? Although I fully endorse the data analysis interpretation of PA (for acting on trends before the trending actually leads to nonconformity), I go back to the concepts of mistake-proofing and FMEA at the design stage for clarification: if you design in the measures that eliminate the potential for nonconformity in the first place - and that is NOT easy to DO -- are you not practicing PA? To me, that's clear in theory and difficult-but-not-impossible in practice. My acid test is whether I can apply it to both a manufacturing and a non-manufacturing setting, and I can.

Back to wearing the auditor's hat, I'd allow some slack in interpretation, at least until there's a larger body of experience in interpreting 9000:2000. I'm known as being a "tough" auditor but also fair, and I see the value in differentiating the two processes in the manner stated above. I have also seen a lot of blank faces among my fellow quality professionals when I "talk like this," and so I'm using you all as guinea pigs in a way.

Have I added anything positive to the discussion of CA/PA or do you view this as further muddying of the already-silt-laden waters?

posted 25 July 2001 04:13 AM               

*************************

From: Nancy Jennejohn
Date: Tue, 24 Jul 2001 15:03:31 -0500
Subject: Re: Elem 14 related to employee suggestions /../Scalies/Holtz/Scalies

From: "Charley Scalies"

John Holtz wrote:
> Not silly advice, Charley, and this is a subject that will come up over and
> over again. (Are you just feeding the grist?) A FMEA is preventive and not
> corrective, as are mistake-proofing and poke-a-yoke. A mended or reworked
> part is corrective and not preventive. Data analysis before, during, and
> after the fact can be corrective (run charts) or preventive (process
> capability, trending, customer sat) depending on the nature of the data. I
> find that they plot out very nicely in a flowchart: corrective actions
> definitely have an arm that maps into a preventive cycle but the reverse is
> not true (I might be able to get my hands on that flowchart...gotta check).
>
> Whadayathink?

I classified this auditor's advice as "silly" because he was stressing form over substance. It's the kind of advice that makes "newbies" think ISO is about documentation when it isn't. But if you want to have separate procedures, I have no problem with it. It's (always) your call.

Since you asked what I think, here goes; First, I couldn't care less about the difference between corrective and preventive action just as I don't really care whether a problem discovered during an internal audit is called an observation or a nonconformance (requiring CA).

In the first place I don't like things with initials because everything with initials fails. Why? Glad you asked. Because things with initials become buzzwords and buzzwords always get trivialized. Just as ISO has.

A problem is a problem and it needs to be addressed based on its merits. If I see an angry dog heading my way, I don't need to be bit in the nether regions to realize I have a problem. I take the same action. A. I get the heck out of there. B. I don't go that way again. As for me, once I document a problem, I tend to be like that dog. I don't let go until I decide to let go.

Secondly, I believe we would be better off looking at the Quality Continuum (my term) as a single continuous process. It starts with controlling nonconforming product (keep bad stuff from getting out the door) moves to corrective action (find the causes of serious problems and make them go away), then preventive action (look for trouble before it starts and keep it from happening) and ends with continual improvement (when you have system that works, make it better.) Management needs to understand that they shouldn't stop at preventive action. Somebody will be making a better mousetrap and it might as well be them.

Who cares how you control a process as long as it works? Procedures, documents, flow charts, training, bosses with bullhorns and big whips (boy am I in trouble with the TQM'ers now) and on and on.

As for FMEA's, they do not define the preventive action process, though no doubt they could be used there I suppose. I don't really know how the QS9000 people use them but in my old military program days - yes FMEA's are that old - they were used during the design phase not during production.

Look, one of the major purposes of measurable objectives, especially process and product objectives, is as gauges of system performance. Not only can they tell you that you crossed the line (a nonconformance) but are early warning indicators that, if you don't make a course correction, you will be crossing the line soon. Either way, the actions I take to discern the cause and to eliminate it's recurrence are not at all affected by whether or not its an NC/CA or a PA.

(Looks like the one I really wound up was moi.)

posted 26 July 2001 09:11 AM               

quote:

---

Originally posted by ikar:
Don't delude yourself. Auditors like all documents, that you show them. It is another question - what do they do with only 6 mandatory procedures and 4 pages' Quality Manual?

---

What the auditor likes is that the procedure is well evolved, clear, concise, and actually followed! We actually have the rare gem among auditors that doesn't bog down an audit with extra focus on 'document control' - but rather audits the 'guts' of systems/processes and reviews documentation as necessary.

You might take some time to learn more about those of us that are registered members of the forum, before you make comments off the hip. I believe those who are familiar with my postings understood EXACTLY what I meant.

As for your other comment/question? concerning '6 procedures', I suggest you use the search feature of the site...there have been extensive discussions of this topic. (and I don't even want to know who would do business with a core document being a '4 page Quality Manual')

posted 26 July 2001 05:38 PM               

I agree. As one proceeds through an audit, document control, or lack of, is revealed. I have seen registrars come in and not specifically audit document control. Their reasoning was as stated above.

I can see a 4 page quality manual. It's all a matter of how one structures their systems and related documentation. Company size is often a factor. Some small companies have everything in one manual - so it's maybe 60 to 80 pages but that's everything.

Because of the differences in companies, including, but not limited to, size, it is silly to try to do anything other than to state one's expectations.

-> What the auditor likes is that the procedure is well
-> evolved, clear, concise, and actually followed!

posted 01 August 2001 02:38 AM               

*******************

From: Nancy Jennejohn
Date: Tue, 24 Jul 2001 15:03:31 -0500
Subject: Re: Elem 14 related to employee suggestions /../Scalies/Holtz/Scalies

From: "Charley Scalies"

John Holtz wrote:
> Not silly advice, Charley, and this is a subject that will come up over and
> over again. (Are you just feeding the grist?) A FMEA is preventive and not
> corrective, as are mistake-proofing and poke-a-yoke. A mended or reworked
> part is corrective and not preventive. Data analysis before, during, and
> after the fact can be corrective (run charts) or preventive (process
> capability, trending, customer sat) depending on the nature of the data. I
> find that they plot out very nicely in a flowchart: corrective actions
> definitely have an arm that maps into a preventive cycle but the reverse is
> not true (I might be able to get my hands on that flowchart...gotta check).
>
> Whadayathink?

I classified this auditor's advice as "silly" because he was stressing form over substance. It's the kind of advice that makes "newbies" think ISO is about documentation when it isn't. But if you want to have separate procedures, I have no problem with it. It's (always) your call.

Since you asked what I think, here goes; First, I couldn't care less about the difference between corrective and preventive action just as I don't really care whether a problem discovered during an internal audit is called an observation or a nonconformance (requiring CA).

In the first place I don't like things with initials because everything with initials fails. Why? Glad you asked. Because things with initials become buzzwords and buzzwords always get trivialized. Just as ISO has.

A problem is a problem and it needs to be addressed based on its merits. If I see an angry dog heading my way, I don't need to be bit in the nether regions to realize I have a problem. I take the same action. A. I get the heck out of there. B. I don't go that way again. As for me, once I document a problem, I tend to be like that dog. I don't let go until I decide to let go.

Secondly, I believe we would be better off looking at the Quality Continuum (my term) as a single continuous process. It starts with controlling nonconforming product (keep bad stuff from getting out the door) moves to corrective action (find the causes of serious problems and make them go away), then preventive action (look for trouble before it starts and keep it from happening) and ends with continual improvement (when you have system that works, make it better.) Management needs to understand that they shouldn't stop at preventive action. Somebody will be making a better mousetrap and it might as well be them.

Who cares how you control a process as long as it works? Procedures, documents, flow charts, training, bosses with bullhorns and big whips (boy am I in trouble with the TQM'ers now) and on and on.

As for FMEA's, they do not define the preventive action process, though no doubt they could be used there I suppose. I don't really know how the QS9000 people use them but in my old military program days - yes FMEA's are that old - they were used during the design phase not during production.

Look, one of the major purposes of measurable objectives, especially process and product objectives, is as gauges of system performance. Not only can they tell you that you crossed the line (a nonconformance) but are early warning indicators that, if you don't make a course correction, you will be crossing the line soon. Either way, the actions I take to discern the cause and to eliminate it's recurrence are not at all affected by whether or not its an NC/CA or a PA.

(Looks like the one I really wound up was moi.)