dear all, does anyone know what the different between failure and deviation? i'm trying to make a SOP for handling deviation & Failure in my company.
thx:(

Why don't you just call it nonconformity which is the correct term for what you want, avoid confusion, and continue to march?

dear all, does anyone know what the different between failure and deviation? i'm trying to make a SOP for handling deviation & Failure in my company.
thx:(


i've got this in Wikipedia.org

Deviation (statistics), the difference between the value of an observation and the mean of the population in mathematics and statistics.

Failure (colloquially fail, phail, flop or flub) in general refers to the state or condition of not meeting a desirable or intended objective

i've got this in Wikipedia.org

Deviation (statistics), the difference between the value of an observation and the mean of the population in mathematics and statistics.

Failure (colloquially fail, phail, flop or flub) in general refers to the state or condition of not meeting a desirable or intended objective

The OP is clearly talking about an ISO management system issue here.

Thanks anyway

The OP is clearly talking about an ISO management system issue here.

Thanks anyway

I thought she just confuse about the meaning of difference between deviation and failure, so i try to search for both of them in wikipedia for her.
and then if she just understand both of them, then easier for her to make that SOP,

Thanks and Best Regards

Deviation is, as defined in Dictionary.com:

de⋅vi⋅a⋅tion
   /ˌdiviˈeɪʃən/ Show Spelled Pronunciation [dee-vee-ey-shuhn]
–noun
1. the act of deviating.
2. departure from a standard or norm.
3. Statistics. the difference between one of a set of values and some fixed value, usually the mean of the set.
4. Navigation. the error of a magnetic compass, as that of a ship, on a given heading as a result of local magnetism. Compare variation (def. 8).
5. Optics.
a. Also called deflection. the bending of rays of light away from a straight line.
b. angle of deviation.
6. departure or divergence from an established dogma or ideology, esp. a Communist one.

Randy is right to advise you not to create confusing sets of terms. However, it's up to you to define the difference if your organization finds it important to do so.

If you do, a deviation could be seen as an opportunity for improvement, if it is not severe enough to be rejected as a product or service, but is not up to your desired standards. In manufacturing such a designation could be given to a part that failed inspection but was shipped after material review board, and/or upon customer permission. Your procedures could describe it that way.

In the nuclear repair industry a deviation from procedure or spec wouldn't be allowed. But if you are not nuclear, you may want to give yourself that gray area so you have something to point to when you want to say, "See here, we can do better than this."

This subject is bound to raise some disagreement, so I will be interested in other people's responses.

You could define the following in your procedure like this

Deviation - a planned departure from the normal practice, a predetermined and planned action. For example , the lack of a certain raw material that could be substituted with another raw material.

Failure - an unexpected result - probably from in house QC testing, or a customer complaint.

You could define the following in your procedure like this

Deviation - a planned departure from the normal practice, a predetermined and planned action. For example , the lack of a certain raw material that could be substituted with another raw material.

Failure - an unexpected result - probably from in house QC testing, or a customer complaint.


Both of which are a nonconformity.........keep it simple, understandable and usable

Both of which are a nonconformity.........keep it simple, understandable and usable

"Deviation" might be being used in the sense of using NC material with permission (from a customer, e.g.), in which case a nonconformity is involved, but the problem is disposition.

The deviation is a formal recognition of an authorised change, you can call it a non conformity if you want - but its just a recognition of an authorised change (displays some form of admission of control).


A failure is a non conformity.

Can help to categorise like this in terms of analysis of data.


You can see how many deviations were raised and perform trend analysis to put some actions in place to prevent recurrence(s).

As Randy stated keep it simple. I would follow Q9000-20005 QMS Fundamentals and vocabulary that does not reference deviation & Failure only nonconformity 3.6.2.
:2cents:

The deviation is a formal recognition of an authorised change, you can call it a non conformity if you want - but its just a recognition of an authorised change (displays some form of admission of control).
While this use of "deviation" is common, it's far from being a universally-accepted definition. We don't know what the OP was referring to.

"Deviation" might be being used in the sense of using NC material with permission (from a customer, e.g.), in which case a nonconformity is involved, but the problem is disposition.

In that case it would be handled under the nonconforming product requirement

I'll defer to my standard answer

"Different strokes for different folks" :-)

I would have to agree with Denis on this. In our organization if we have a deviation, it is something that is customer approved and therefore we do not catagorize it as a nonconformance. But it is really up to the company or specific industry to define.

I would have to agree with Denis on this. In our organization if we have a deviation, it is something that is customer approved and therefore we do not catagorize it as a nonconformance. But it is really up to the company or specific industry to define.

How do you categorize it before the customer approves?

It would start as a nonconformance for sure. So we do handle it under our non conforming process as an "accept as is" or "customer approved".

I agree that in most cases a deviation would be a non-conformance. However, a deviation under a controlled and documented situation could be considered conforming. The deviation might include a product characteristic that does not meet a design bill yet is acceptable per the standards agreed to by the customer. Further testing could prove the deviation as good design and therefore would no longer be a deviation. Failure disregards critical characteristics while testing has shown that they cannot be deviated without unwanted consequence.

I agree that in most cases a deviation would be a non-conformance. However, a deviation under a controlled and documented situation could be considered conforming. The deviation might include a product characteristic that does not meet a design bill yet is acceptable per the standards agreed to by the customer. Further testing could prove the deviation as good design and therefore would no longer be a deviation. Failure disregards critical characteristics while testing has shown that they cannot be deviated without unwanted consequence.

It is still a nonconformity no matter what you want to call it. It does not meet design characteristic, even though the Customer has accepted it.

Let me take a whack at this dead horse. :horse:

A condition starts out as suspect, pending disposition.

The dispostion can be:

-Scrap (Failure)
-Us as is, no issue, no nonconformity (e.g. measurement error resolved)
-Deviation (approved for use after customer approval per statutory limitations) after:
--Rework
--As-is

Let me take a whack at this dead horse. :horse:

A condition starts out as suspect, pending disposition.

The dispostion can be:

-Scrap (Failure)
-Us as is, no issue, no nonconformity (e.g. measurement error resolved)
-Deviation (approved for use after customer approval per statutory limitations) after:
--Rework
--As-is

There are times when a product is obviously nonconforming in a state of nonconformity bad, so the "suspect" tag might not be appropriate in all (perhaps most) cases. We generally know when something doesn't meet the requirements, and it comes down to what to do about it--the disposition.

It's important to recognize if this is about a product/service, or procedure.

Using terms like deviation really can be too confusing. All we should be doing is noting that something has not met the target - then we address that.

It could be that the question is meant to capture an acceptable non-perfect area. We should establish procedures that allow appropriate controls and do not lock uses down in unreasonably strict practices, because honestly any deviation from procedure is a nonconformance.

Now for the product aspect. In my NDT days, we would never address issues like undercut or weld porosity as defects; we had to use the term discontinuity and evaluate for acceptance using specific criteria.

But discontinuities weren't deviations; they were merely imperfections; some were acceptable and others failed inspection.

If the intent of the original question is supposed to involve this kind of scenario, I would strongly advise against using the term "deviation" to describe an imperfection that could be acceptable.

There are times when a product is obviously nonconforming in a state of nonconformity bad, so the "suspect" tag might not be appropriate in all (perhaps most) cases. We generally know when something doesn't meet the requirements, and it comes down to what to do about it--the disposition.

Without getting into the "What is 'is'?" as we have with nonconformance/nonconformity, I prefer the singular tag of Suspect simply because whether it is known to bad or needs to be determined it is bad or good enough, it still needs disposition. Why have 2 inputs into disposition? One will get it there. If the system requires tagging the product for identification of product condition (and I believe it would) why buy a Suspect tag and a Scrap tag? Use a Suspect tag to go to disposition. Saves the operators from having to make a preliminary disposition that still goes to...well, disposition!

Without getting into the "What is 'is'?" as we have with nonconformance/nonconformity, I prefer the singular tag of Suspect simply because whether it is known to bad or needs to be determined it is bad or good enough, it still needs disposition. Why have 2 inputs into disposition? One will get it there. If the system requires tagging the product for identification of product condition (and I believe it would) why buy a Suspect tag and a Scrap tag? Use a Suspect tag to go to disposition. Saves the operators from having to make a preliminary disposition that still goes to...well, disposition!

You only have to have a "Nonconforming Material" tag, and the understanding that verification is required prior to disposition.

You only have to have a "Nonconforming Material" tag, and the understanding that verification is required prior to disposition.

Sure. That will work. You can even use 'Rejected' - it is a common tag to purchase.