Supplier Auditing (2nd Party relationship)

Final Inspection 5 step plan, or How to eliminate Wrong P/N Rejections

I noted over a period of years, a persistent reported problem in supplier defect data. One of the top 5 problems every month
was related to incorrect part marking. The rejection may have been called ‘wrong part received’, ‘wrong part number sent’,
‘wrong PO number’, ‘paperwork does not match part’, or other. This rejection category was often the #1 supplier defect
month-to-month, it was always in the top 5, and was typically the #1 defect code when data was totaled for the year. It did
not matter what quality system was in place at the suppliers (ISO, QS, MIL, or other) if a supplier had the problem at all it
was typically a repeat problem.

Supplier Corrective actions had been ineffective at stopping the problem and I was at a loss as to what to do to significantly
reduce the number of occurrences. The typical SCAR (Supplier Corrective Action Request) reply was that the supplier had
instructions in place to check the part number and revision of the part, they attributed the problem to operator error, mixed
paperwork, mixed factory order packets, etc. and took ‘action’ to prevent the problem. However, the problem persisted, it
would typically dip for a time when a concerted effort was directed at the supplier, but it seemed to always come back.
Finally, I took the time to do some benchmarking and examined data to find suppliers that did not have problems with
shipping us the wrong part number. In some cases it may have been purely a result of ‘luck’ or dedicated personnel that was
the cause of not shipping us wrong part number since I could find no significant difference in their systems as compared to
others. I finally came across a supplier that had taken effective corrective action a couple of years before and had eliminated
the problem. Both their data and mine showed that they had not had a problem with mis-marked parts for the two years
since. It had become simply a non-issue for them. I had a chance to visit their facility and look closely at their method.
They had been using the method so long that they were curious that I would find it interesting.

Over time, I have adapted and expanded their method and applied it to other suppliers. When it is properly implemented at a
supplier that has problems with ‘wrong part number’, the problem disappears. I call this a 5-step ‘sanity check’ for want of a
more scientific name, also to avoid confusion with typical “Final Inspection” methods. I tend to get very prescriptive with
suppliers when I know that something works well, and this has worked well. It provides an ‘outside loop’ of original data
that aids in the elimination of wrong part number shipments. This type of defect fell from a consistent top 5 rating, to rarely
showing up on the top 10 listing for all suppliers using it. The ‘sanity check’ has to be a separate, distinct function with clear
instructions of what information is to be compared for accuracy. It may be combined with Final Inspection, Shipping, Pack
& Ship, etc., or other function depending on how the supplier is set up. It is best if the check occurs immediately before
shipment and must (especially) be applied to in-stock product. The method need not cause any disruption of flow if it is
implemented correctly.

NOTE: One thing that I take as a warning sign in a supplier is if they are doing a detailed mechanical or electrical inspection
at final on a regular basis. This strongly implies that they do not have adequate process controls in place and they are trying
to ‘catch’ any bad product before shipment. This is valid only on a short-term, ‘special’ basis, or as part of a periodic
internal audit schedule. If the inspection is done on a regular basis then something is very wrong with the process controls.
If this is noted as part of a supplier evaluation for new business then I often will recommend that we do not let a contract
until the supplier can provide justification for the added expense of the ‘extra’ inspection. This justification involves a close
examination of their statistical data for the process and an examination of the process steps for the product.

The 5 step ‘sanity check’ is the data that should be at Final Inspection (or the last step before shipment of customer product,
could be shipping). The 5 data sources and an explanation of each is below:

1. The original customer contract along with any change notices. Often companies object to having prices of
product available on the floor, simply cover the pricing area with a piece of paper and copy.
This provides a last check to be sure that what is being shipped matches what the customer ordered; it also
verifies that any change notices have been included. The customer PO and any revisions should be provided
from customer service directly at the time that the production order is moved to final. This provides the
‘outside loop’. Putting this information in the original factory order packet is not as effective since the
paperwork can be mixed in the process.

Many suppliers/manufacturers have sophisticated order entry and tracking systems. For the most part these are
excellent systems. The flaw in these systems is that one slip of the fniger and the part number can be
transposed (AMP had many similar part numbers) or an incorrect revision entered. Also, a person can mix two
Norman Black (black4982@aol.com) 5 step sanity check pg 1 of 2
 factory order packets on similar parts and the mix-up might not be discovered until the customer reviews the
part. This outside loop gives the supplier a strong opportunity to prevent their customer from getting the wrong
part.

2. The factory order packet. This needs to be reviewed (recommend that it is done prior to final at the last step
before final) to verify that all the required steps have been completed. A control sheet showing all the steps
should be included and reviewed. If there is not a unique control sheet for that customer part number then there
has to be a generic format for doing this. The factory order packet needs to be reviewed at final with an eye to
the customer specific information, i.e. part number, revision level, PO number, internal tracking number, and
other. This review is typically on a check sheet and is signed off. One supplier folded the check sheet into their
C of C. (I do not agree with the generic C of C that is typically asked for by many customers. It is a useless
piece of paper. The only weight that it would carry is in a court of law if you had to sue a supplier over a part
number or performance issue, it would only be another piece of paper on the pile. The only document that is
significant is the PO, since it is the contract. The C of C only presents an opportunity for failure, i.e. if it isn’t
signed, wrong part number, wrong PO number, on and on. The only other document that may be of value is test
data or statistical data for the part in question.)

3. The customer drawing. This is not meant to be used to ‘measure’ parts at final. It is for reference to verify
part number, revision and to physically look at the part to make sure that any part number marking is correct
and that the part ‘looks’ like the part on the drawing. The drawing also needs to be compared to all the other
data for part number and revision. The supplier may have made the best part possible, but if it is to the wrong
revision, they may have made scrap.

4. The part itself. A quick visual check to verify that the part looks like the drawing and has the correct marking
is generally all that is needed. A visual inspection of additional parts may be done but a detailed inspection is
not appropriate. No matter how good the detailed inspection is, some ‘bad’ product will pass through if the
process is not adequate. If the factory depends on final inspection to catch the mistakes for on-going
production, then I will most likely not recommend this supplier for continued business.

5. Review the packaging and labeling. (This required the only major change at a few of the suppliers since they
typically made and applied labeling at shipping after final. However, this also eliminated a major source of
problems since the labels are reviewed prior to application). Are there any special requirements called out on
the PO or drawing? Has the supplier met all packaging and labeling needs of the customer? Is the product
being sent to the correct customer at the correct address?

Yes, I know that this is very basic, but American industry is sometimes slow to catch on to some basic concepts. They will
spend huge amounts to improve the process (good) or thousands for improved inspection methods (not as good) but miss the
obvious on something that costs them little to nothing. If they ship the ‘right’ product with an incorrect part number, it has
the same negative effect on the customer as a ‘hard’ rejection for poorly made parts. I want to emphasize that although this
is a very simple method, it works well to reduce a specific type of rejection. ‘Wrong Part Number’ went from #1 on a
consistent basis in the company where I used this, to not even making it on the top 10 chart, over a period of 2 years - when
applied any time a rejection occurred. After I had tested this method for about a year with a couple of suppliers and saw a
significant reduction, I began requiring that suppliers implement this method any time that they had this type of rejection.
The first time that this type of rejection occurred I would sent out a description of what was required to be changed along
with the SCAR (reverse corrective action?). This would include a detailed description (as above) of the types of data to be
present, along with an example procedure for final ‘inspection’. If the same defect occurred a second time at the supplier, I
made a trip there as soon afterward as possible to give a presentation to their management and quality people. I would show
them before and after examples for ‘properly implemented’ final ‘sanity checks’ (usually Pareto charts of supplier defects
and supplier quality ratings) and an estimate of the implementation cost. This was usually enough to get them to implement
the change, especially if the examples given were their competition. (We were actively involved in a supplier reduction
program.)

Norman Black (black4982@aol.com) 5 step sanity check pg 2 of 2