Where are the rules for when a repeat minor nonconformance becomes a major?

[Marc]

Let's be clear. The example @Rameshwar25 gave above is a simple and common one - One or more measurement instruments are found to be out of calibration this year and are written up. A year later it is observed that some measurement equipment is found to be out of calibration. Measurement "labs" in a company are supposed to have a system which will "not allow" for this type of condition. A measurement lab CAN have equipment which is not "in calibration" as long as it is segregated or in some way clearly marked as such. That is elemental, and even in the earliest ISO 9001 such a system was required. Back in the 1980's aerospace, automotive, and just about any industry, really, Because this is such a simple, basic requirement, if in year 2 the problem still exists there is a systemic calibration system problem by the nature of it being a recurrent problem.

The secondary aspect is that for such a finding (the first one) had to trigger a Request for Corrective Action. Since this is a recurrent issue for which a corrective action request should have been initiated, the problem is such that it "should have been resolved/fixed" in less than a year. But lets look closer (back to my "every scenario is unique" mode).

Real world example: Some years ago I worked with a company which was quite large, had thousands of measurement devices of mixed nature. From micrometers to electronic test equipment, check fixtures, etc. A corrective action was started. It was not closed out when the auditor returned a year later. Reviewing the correction, we find that the company determined that the effort and the cost to quickly fix the problem so fast was excessive. The corrective action investigation did show that the uncalibrated equipment was effectively kept from escape into use. The corrective action included a plan which was a 2 year project. That plan was reviewed and found to be on-time and in progress. In this case the auditor wrote a minor, again, a duplicate of the one from a year before. The auditor can not write a finding against the Corrective Action system - There is no requirement that I am aware of that a corrective action be completed in a specific time frame. In some cases it isn't realistic to fully complete a corrective action in, say, a week, or a month, or even several months, and in some cases longer.

The worst an auditor could have possibly done, in my opinion, is write them up for a systemic system failure against the calibration department and classify it as a "Major". And what if they do? Worst case is the company will have its certificate yanked. Now, if this is a big company and is an important Tier I supplier, there will be some communication between them if their certificate is pulled. They will "work something out".

At times auditors have to make judgement calls where what is written in a requirement or standard doesn't perfectly align with a specific scenario. That is, of course, the problem when a company or group of any kind tries to write a requirement or standard. It is impossible to do in a way which will address every scenario. I'm not saying the process of writing a requirement or standard is a bad one. It's all we have. Some groups issue "Clarifications" or "Interpretations" to address some things more closely, but there will always be more specific scenarios that will occur where the standard or requirement does not address.

@Paul Simpson is in for some fun in TC 176 and @Marcelo Antunes is involved with one or more standards groups.

From Paul: "Thanks again, Sidney. I'll take your comments with me to the next TC 176 meeting in Portugal this month. The interpretations process is interesting. There is a balance with keeping the process simple and allowing all the TC 176 participating organizations (the National Standards Bodies) having a say on the outcome. There are a few other issues but I'll save those until there is more to say after the Azores."

The Azores - Make me jealous...

 

[Mr.Ruiz]

A much better rule would force lead auditors to justify in the report why similar nonconformities in consecutive audits do not represent a failure of the organization’s corrective action process. By forcing the auditor to justify their position, it would give the CB and the registrant a way out of an automatic major NC when and if the evidence and rationale justifies it. Everything is supposed to be processed via a risk-based line of thought. Why not in this case?

Totally agree

 

[Rameshwar25]

The corrective action included a plan which was a 2 year project. That plan was reviewed and found to be on-time and in progress.

IATF rulebook terms it as "100% resolved but open NC". It is not necessary to compete the entire corrective action(s) in one year only. IATF allows a corrective action plan which may extend beyond one year. If all action points planned for year-1 are completed at the time of next audit, it will be considered effectively implemented.
Text from IATF rule book 5.11.3
"In exceptional case(s) where the implementation of corrective actions cannot be completed within a maximum of ninety (90) calendar days from the closing meeting of the site audit, the certification body shall consider the nonconformity open but 100% resolved when containment of the condition to prevent risk to the customer has been taken, including a review of the systemic impact on the client’s process"

 

[Marc]

<snip> Consider the situation where several years ago you filled out a FEMA form "wrong" by not having a date. You made the correction, and learned your "lesson" but who knows how many other old forms might be missing a date. Welcome to two majors. It's nuts.

I had to laugh reading what you wrote - I first ran into the situation back in the early 1990's, like 1992 or so. Before the audit they made a frantic effort, successful but a waste of time in my opinion, to set a "benchmark date". They actually segregated documents and records as "Before" their implementing ISO 9001 from newer ones. A gap analysis had shown a serious problem with incomplete forms/records and many procedures were not controlled. Auditors were as weird then as they are today (not said as disparagement), the company knew it was unrealistic to review every old record and wanted to be ready for what they knew the auditor could drill them on (so to speak).

It worked. The issue came up and the auditor agreed that any record he might review that was prior to the company's "benchmark" date would not be written up. I don't remember if the company actually did a written, formal corrective action at the time they identified the problem, but it really didn't matter as the problem had been identified and resolved well prior to the registration audit.

In exceptional case(s)

Precisely. What this essentially says is the CB will review and render a decision. That, in turn, says the registrar can make a judgement call because, as I have been saying, there are many times where the specific scenario can't be aligned with the wording of the standard/requirement.

As I have recounted quite a few times in discussions here where an auditor stated what he was going to write the company up for, the company sat the auditor down while they called the registrar contact, and in short order the auditor was told by his boss to kill the nonconformance. Each made a judgement call. The auditor's boss "won", in a sense. I have actually seen this happen in a company that was something like 12 employees total.

 

[Sidney Vianna]

The Azores - Make me jealous...

Some people refer to ISO as the International Sightseeing Organization...

 

[AndyN]

What does it indicate? where are we going? The entire calibration calibration is likely to fail in future. and it becomes basis for a major NC.

I think that's a stretch. If there's greater than say, 400 items, 2, 3 or 4 items is NOT a breakdown (until we know the impact on product quality). It's simply NOT a numerical issue.

 

[Rameshwar25]

Dear Andy,
1.When I said two instruments are out of calibration, i gave example to explain a similar minor nonconformity. I could use ten instead of two. It was just an example.
2. Every minor nonconformity is raised with a justification that there is no impact on customer requirements. A single instrument found uncalibrated being used at final inspection may be a case for major nonconformity. Yes, it is not a numerical issue. It is impact on requirements.