The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : 8D Approach - Verification vs. Validation


Periannan
2nd October 2007, 07:08 AM
Dear Guys,

Need you guys help to clear my doubt! Recently I had an argument with team member regarding 8D approach in answering the corrective action.

- In brief, D3 column taking about define and implement containment action to isolate the effect of the problem from internal and external customers until corrective action is implemented. There is verification and validation column

Base on my understanding, verification is taking about what method has used to perform the re-screening (is it full proof and will not induce other failure?)

As for validation, basically it will show screening result (after rework or sorting), disposition of non conformance part, scarp and result of buyoff by third part such QA.

My team member says it won't sound like that. Do you have an example? I have attached a 8D replies of mine

Thank you,

Periannan :bonk:

Stijloor
2nd October 2007, 07:36 AM
Hello Periannan,

Look at paragraph 4.5 in the 8-D Work Instruction attached to this (http://elsmar.com/Forums/showpost.php?p=188458&postcount=3) post. This post was from Mr. John Nabors.

Stijloor.

Geoff Withnell
2nd October 2007, 11:48 AM
From the ASQ Quality Glossary:



Verification: The act of determining whether products and services conform to specific requirements.

Validation: The act of confirming a product or service meets the requirements for which it was intended.

I have always taken this to mean:

"Verification" is does the action meet the specified requirement, in this case is 100% containment implemented?

"Validation" is does the action produce the desired results, in this case is nonconforming product actually removed from the product flow?

Geoff Withnell

Valeri
3rd October 2007, 09:20 AM
Periannan,

Verification: did the action you take to "fix" the problem permanently correct it?

Validation: is this "fix" still working?

Periannan
4th October 2007, 09:54 PM
Dear Velari,

Thanks, I’m pretty clear with validation. If i applied in interim containment, the validation will see whether the sorting method done at supplier site is really good enough or not. The data or evidence can validate by monitoring my in coming acceptance (sampling) and my production escalation (reject by workstation).

As for the verification (referring to above scenario), can I say, how do you determine the rework flow for interim containment to see whether it is fully proof to arrest the defect or not.

Let see the an example below


Interim containment action

Action taken
a. Arrange sorting activity to segregate the defect goods
b. Label (box level) as “After sort” for segregated part
c. Scarp the reject part


Verification
a. The defect is visible and able to detect by visual inspection
b. Sorting flow and defect sample has catered to sorting team

Validation:

a. Total sorted Qty: 1200pcs
b. Reject: 100pcs
c. Scarp: 100pcs
d. Customer site: No defect found for the sorted lot

First: In above case, is it rite, the disposition status should written under validation, verification or action taken.

Second: is it the rite way to answer the verification column.


Periannan :bonk:

Valeri
5th October 2007, 09:49 AM
Periannan,

What this sounds like to me is you are sorting temporarily to "stop the bleeding" - a short term solution at best. Are you are going to sort forever because of the problem?

You need to find the root cause for the defect, initiate the solution to that problem and make sure (verify) it's working. If it's still working 3 months down the line, it's validated.

Does this make sense or did I miss the mark all together?

Valeri
5th October 2007, 10:02 AM
Periannan,

I just found our training. Perhaps this will help.

Jim Wynne
5th October 2007, 11:41 AM
Periannan,

Verification: did the action you take to "fix" the problem permanently correct it?

Validation: is this "fix" still working?

If the first thing is true (the problem has been permanently fixed) then the second isn't necessary, no? While the verification/validation dichotomy might be useful in some contexts, sometimes we spend time unnecessarily trying awkwardly to fit actions into standardized language. Something goes wrong, and we attempt a fix. In order to understand whether or not the fix is efficacious, we test it. It's really just the scientific method:

Form a hypothesis (The problem is caused by some phenomenon x)
Make a prediction based on the hypothesis (if I invoke a specific control, the problem won't happen)
Test the hypothesis, and see whether or not the prediction is realized.
Rinse, lather, repeat (as necessary)The words used to describe the steps (verify/validate) are not important to the process so long as everyone understands how the process must be invoked.

Valeri
5th October 2007, 04:12 PM
If the first thing is true (the problem has been permanently fixed) then the second isn't necessary, no?


You would think so, wouldn't you. Unfortunately, in my industry, that doesn't work. Let's say the correction does work, it was tested and parts are good and they stay that way for awhile. Somewhere along the way, something squirrely happens and the issue raises it's ugly head again; same issue/different day.

The words used to describe the steps (verify/validate) are not important to the process so long as everyone understands how the process must be invoked.

I agree, it's semantics.

Jim Wynne
5th October 2007, 04:55 PM
You would think so, wouldn't you. Unfortunately, in my industry, that doesn't work. Let's say the correction does work, it was tested and parts are good and they stay that way for awhile. Somewhere along the way, something squirrely happens and the issue raises it's ugly head again; same issue/different day.


Then the problem wasn't permanently fixed. :cool:

Periannan
7th October 2007, 10:55 PM
Periannan,

I just found our training. Perhaps this will help.

Dear Valeri,

Thanks for the training material. do you have a copy of replied 8D thats in line with the given material. this will help me to further understand on practical application. regarding Jim concern, yes as long the people understand the process or the issue had address successfully no other things to be required. it already meet the objective. i'm now working in how to generate a good report of 8D (reporting wise) therefore each defination and content of each D elements are important to me. I need people view to success it. The main reason of doing this is, my boss always has something to say to reject the report. i believe the rejection is base on his understanding and what he want to see. getting majority view can strenghten my understanding and Challenge him. I believe teaching by referring to theory is not enough, giving a good sample or example of 8D report in line with theory will give better understanding. usually my boss will give the theory. attached is the material that he use to explain to us

periannan :rolleyes:

kedarg6500
8th October 2007, 07:48 AM
Zip File Is Password Protected.please Let The Password Known

Periannan
8th October 2007, 08:20 AM
Zip File Is Password Protected.please Let The Password Known


as far as i know there is no password required. let me extract for you

Periannan :cool:

Periannan
10th October 2007, 03:05 AM
Periannan,

I just found our training. Perhaps this will help.

Dear Velari,

do you have a sample of replied 8D report

periannan :rolleyes:

Yew Jin
10th October 2007, 11:42 PM
I will not ever forget Plan-Do-Check-Act cycle or Plan-Do-Study-Act cycle when dealing with any problem.

8D technique is still follow the cycle above.:rolleyes:

zancky
9th June 2008, 01:51 PM
from my experience I have seen a lot of different 8D forms (even an 8D form with only 7 items!) one from each customer but never a validation requested for a containment action.
One of the biggest customer into world (german automotive company) ask only for effectivity (i.e. how much sure we are to contain problems) in this case.
If I have to fill a form like Yours I will suppose that validation is how to prove the containment action is enough. If I decide to sort something I will prepare a controlled batch with some known defects and I will check that inspection/sorting will find every defect I have inserted into the batch.
Verification may be just control that every box into production and warehouse has been sorted and list how many defects have been found.

Raffy
1st July 2008, 06:12 AM
We were audited last week and one of the observation that we have is that the term 'permanent corrective action and prevention of recurrence' is one and the same. I would like to confirm the statement above. Is there a new 8D Report that was been published? Please advise.:confused:
Can somebody share the new format for 8D? Thanks in advance.
Thank you very much in advance for the usual kind attention you will give this request.
Raffy:cool: