QS9000's New Dock Audit Requirement

I have a question for anyone out there .

How are you going to meet the new dock audit requirement ? (4.10.4.2)

We are finding it very difficult to pull parts out of a box prior to shipment and then re-do all inspection and testing in order to meet the intent of this requirement before we ship.

Even if we specify a frequency of once a year we can't justify the lost time and expense. And I'm afraid that we will not be able to justify a general frequency such as once a year because it will not be based on PPM levels.

Our first thought was to have our existing Dock Audit function verify that all in-process inspections had taken place by checking for a series of stamps on the Kan-Ban traveller, but our registrar has told us that this dos not meet the intent of the standard so we are fresh out of ideas.

Any help would be greatly appreciated.


------------------

[Roger Eastin]

I thought the intent of this requirement was to look for things like obvious damage to packaging, checking for paperwork errors, etc.? I know the section states that the audit covers all specified requirements, but I thought it was basically a "packaged product" check.

[Kevin Mader]

Bill,

I'm guessing a bit here, but I think that your registrar may be saying that you need to add to what you originally proposed. Product audits generally require you to do an appraisal (inspection and tests) of the product itself. Aside from your records, does the product meet specification? You must develop a program to ensure that the product you manufacture meets the Customer's requirements. This subparagraph allows you to determine frequency. Plan this activity as appropriate (status and importance). How were the results of the last Product Audits? Is there evidence to support a reduction in frequency or should frequency be increased? Be sure you can defend your position, especially if you plan to be liberal.

It seems that your organization may have a slight commitment problem. You need to satisfy the Customer requirements each and every time, not only when it suits you. But that doesn't mean that you have to open a bazillion finished goods to determine conformance. There are many suitable sampling plans (statistically proven) that can reduce work and give you information on acceptability of goods (alpha/beta risk).

Roger's points are well taken here. Product Audits are performed to determine if the process is yielding the proper results (Does product meet spec?), inclusive of product records. By your posting, it appears that you have the traveler records (acceptance stamps), now you need the product records of your audits. Did you find the shipping label in the correct location? Where they the correct size and contain correct information? These are things your traveller probably does contain.

we check the packaging,labeling,shipping q'ty and the reports of OBA as dock audit before shipping. I think the packaging and labeling
inspection are the main requirements of dock audit of QS-9000.

[Dawn]

We are shooting for registration in the spring, and we have a procedure stating we will perform dock audits once a year.
When we perform the audit, we will take six random boxes off the shelf, open them up, check for contamination and other visual problems, check 6 pieces out of each box for inspection verification, and pack them back up. I have created a special form to verify everything has been done and will be checked off by QC Technician who performs the work. This seems to be ok to meet the intent to me, and its really not that big of a deal.(Compared to alot of other requiremetns anyway).

[Kevin Mader]

Dawn,

Why 6 pieces and not 7 (is there any statistical significance to the number being 6)? Are once a year dock audits suitable (frequent enough to reduce the risk of sending nonconforming product)? These are some tough questions, but be ready to answer them.

One thing to be careful of is setting the trap for yourself. I speak from experience here. While many of us struggle with commitment of resources to achieving ISO goals, ISO does not care about the situation. They care that things are being taken care of as prescribed by your procedures. If you can avoid it, try not to tie yourself into fixed numbers (6 items, 3 days, by the 15 day of each month, etc.).

Sampling plans are a huge bear trap. Plans that are set arbitrarily (5 items since I have 5 fingers on one hand) have no statistical significance. You can still use them, but you will need to show a registrar that the sampling plan you have selected/created is effective (both in detering a nonconformance, and, in cost). This is generally why organizations use established sampling plans (MIL, ANSI, etc.). Why reinvent the wheel so to speak. So what sampling plans are useful when auditing?

The book "The Quality Audit" by Charles Mills has some useful information in the arena of audit sampling. He describes three types of sampling, two are more commonly used than the third. In order they are Binomial, Poisson, and Hypergeometric. Sounds complicated, but their bark is much tougher than their bite. These methodologies can significantly reduce the amount of work while yeilding bonified results. Predetermined confidence levels and levels of acceptability/rejectability may give you some answers to the question of having enough time or resources.

The form is a good idea, neat and organized in the record collection department. Be sure your QC Tech also has auditing techniques training beyond general inspection technologies and that the associate training record reflects this. Otherwise, good luck next spring. Hope all goes well.

As an FYI, I've been assisting this week in an ISO Lead Assessor class and the subject of the new dock audit came up. The instructor of the class, a well-respected ISO/QS9000 certfied lead auditor, stated that he would expect to see someone randomly hold out a package prior to sealing to inspect its contents. A Motorolan in the class (parts of Motorola are QS9000 certified) stated that that was going to be their approach.

I would have to agree with this approach because it does seem to make the most sense. Having to reopen then re-close a package to check its contents doesn't make much sense.

------------------

The above reply should include packaging and labeling inspection, too.

------------------