Renumbering Inspection Report Numbers, do you do this?

Example: 100 parts for delivery, 105 parts made (5 as extras). All are engraved with IDs + a sequence number

During inspection parts 7, 23, and 65 FAIL. Parts 101, 102, and 103 are selected, engraving buffed out, then re-engraved as 7, 23, and 65, with the Inspection Report rows for those parts updated with the measurements from 101, 102, and 103.

I've been told we do this because the Customer wants a contiguous string of sequence numbers with no gaps.

How common is this?

Not so much a complaint as a "sonar ping" on prevalence.

Thanks in advance.

I guess whatever the customer wants.
However, I think on your in-house paperwork you should leave the original number.
ie: 6,101,8... etc
The reason being , if something is wrong with 7 you don’t want to be checking 6 and 8 for similar problems.

You should be checking 100 and 102.

ttrager said:

...
I've been told we do this because the Customer wants a contiguous string of sequence numbers with no gaps.

Click to expand...

OUCH....assume this customer not big auto.
I think this a terrible idea but the customer is always right.
So more like a serial number for shipped and I hope some way to track that back to actual run and your internal SPC.
Do not make parts that do not pass and this not a problem.
If parts scarp on a sub op how do you handle the part count or serial number tracking? Is it parts only on final that get numbered?
If so is final inspection the last op and one just skips the bad? How to link to the actual run data?
Bob

I don't know how common it is, but I really don't like pre-engraving the extras then removing their marks and re-engraving. Several reasons, but main one is that it looks like what it is: a deliberate attempt to renumber an individually tracked and controlled part. Could bring a lot of very hot questions a couple of years down the road, and entirely unnecessarily.

I'd recommend leaving the extras entirely unmarked until and if needed. I'd also recommend a process that immediately destroys (not just drop in a wastebin) the failed parts, because the last thing you need is a pair of parts with the same number floating around.

I think it lead to a mess, in your report 7, 23, 64 failed but you shipped 7 23 and 64... are these the old bad ones or the new renumbered ones?? Or maybe 64 is a bad one you shipped by mistake and you junked the good 64...Phil

If the parts delivered to the customers have to be engraved with consecutive numbers, I strongly suggest having two completely different numbering systems: one for production (it could be a sticker, written with a sharpie, or engraved, whatever works), and one for the customer, engraved just before it leaves the doors of your shop. Of course, you need to keep a spreadsheet/database to track what production ID corresponds to what customer ID.
Given that engraving the customer ID is done after inspection, you can keep them in the same order with the order of production.

Example:
Production:
ABC001 (good) -> 001
ABC002 (good) -> 002
ABC003 (bad)
ABC004 (good) -> 003

Paolo

+1 on using Production ID, then assigning SN. Treat SN's as holy, never to be changed, then you get no duplicates or tracking problems.

I echo sentiments of the others- your internal SPC is basically holy, and the identity of every part produced, wether fail or pass, needs to be indelible, and for at least a half-dozen reasons.

I would consider having two markings... perhaps the first is the PIC (Production Identity Code (which follows it from start to finish of process) and then the QCSN (Quality Assurance SERIAL number) Which is applied immediately upon passing all QA tests.

If it were me, I'd make the PIC an encrypted alphanumeric, somehow etch it, or possibly make it in a bar or QR code if your shop process has capability to do so...

And for the QCSN, make sure that number, when assigned, identifies the PIC, so it can be correlated to all date codes, batches, processes, and finally, the test documents supporting that component's successful passage.

That way, a failed component is never awarded a QCSN, never goes out the door, but your Quality Control's defect history remains rock-solid and investigateable for every future need.

The integrity of production history necessitates having all learned lessons retained, as the failures of our past came to us as costly tuition that should NOT be squandered.

I would like to thank each and every one of you for your considered advice, no matter the subject isn't overly technical.

Didn't think about using a separate internal control number on the parts. I'll think that through even though that means two separate engraving cycles. Sometimes the precise syntax is identified by the Customer.

I don't like how I'm doing this at the moment, it gives me the hives: Having to buff off the ID from the Spare to be used to replace a delivery part, engrave with the delivery part's ID, then update the Inspection Report line for the delivery part with the Spare's measurements. I get away with that at the moment because our quality rate is high, with few fails, in addition to our job batches being relatively low. Still, that's a messy process IMO, too much re-handling, multiple spots a mistake can happen.

I'm going to think this through some more considering what each of you have said in this thread.

Thanks much for your time, and I hope you are all well in this weird times we find ourselves in with business closures and what have you.

Phil,

What is happening is the original 7, 23, 64, have been physically replaced with spare parts. The spare parts have been renumbered as 7, 23, and 64 and the inspection measurements for the spares are used to update the entries for 7, 23, 64.

I don't like this. Lots of opportunity for a mistake, and basically data collection to facilitate future evaluations / metrics is disabled IMO.

If you are going to do this why not number 7,23,64 as 7r 23r and 64r?...Phil

I'll think about that. Currently, I have an "intelligent" spreadsheet setup to auto-number rows and ID inspection point columns based on just entereing number of Parts and number of Inspection Points.

I'll think about how I want to adjust the method.

Thanks.