Apr 08 2011

We’re making too many rejects

Published by at 1:18 pm under Uncategorized

That’s the essence of the phone call I get from clients.  Nice.  Doesn’t tell me a thing other than I might be able to send a Whopper invoice. “Too Many” rejects is a value judgment. But they can be real profit wasters if you’re not careful.  If you hear this cry of anguish read on.

Rejects come in two flavors:

Rejects from the customer – the problem.
If you’ve got reasonably good specifications and inspectors whose IQ is higher than a squirrel, rejects from the customer come from two sources.

First is that someone missed something gross and you shipped a production lot that was drastically different from previous shipments and didn’t have any hope of passing function or cosmetic requirements.  This is a righteous reject.  In this instance a witch hunt is appropriate for both the quality and production people too dumb not to know what a good part was.

The second cause of a customer reject is:  (1) Somebody pulled a quality audit and of the hundreds of dimensions on the specification they found one you didn’t comply with and therefore rejected it. Nobody ever bothered to see if the parts worked or would pass cosmetics.  The minute they found a dimensional discrepancy they shipped the whole lot back. (2) In the normal course of commercial manufacturing you shipped a lot of parts that were functional and cosmetically acceptable but the dimensions were slightly different.  Because you didn’t stay within a CpK of a zillion (+/- .001 mm) it was thought you lost control of the process and therefore there must be some unacceptable parts in the production lot and therefore “we’re rejecting them because their different” (3) Sales have slowed down and the customer’s inventory levels are too high.  They either reject them for any reason they can think of or no reason at all.

Rejects from the customer – the solution
You need a policy on how to handle customer rejects.  Go to wjtassociates.com  click on ‘free stuff’ then download the “Policy Manual”.  In essence it says when a reject happens here’s what you’ll do.:”
(1)    In order to be a valid reject a part must be shipped overnight express as an example of the entire lot with a description of what’s wrong.
(2)    Right or Wrong, all work on that part will halt for two weeks while you investigate the problem.
–If it’s your fault, after the Slaughter of the Usual Suspects, you need an All Hands Effort (24/7 if need be) to fix the problem and try to get replacement parts to your customer ASAP
— If it’s not your fault a stated policy of two weeks without shipments will cause your customer to go into a major pucker before rejecting anything.
(3) Know what quality is.  See below

Rejects from the production floor – The problem
This is usually a head scratcher.  Sometimes the mold goes crazy and it won’t make a good part.  Other times you’re just irritated at a low yield from this mold.  Regardless of what it is the solution is the same.

First, understand quality.  Second make sure your customer has the same definition of quality as you do.  When it doubt go to the industry standards.  Yes, there is a standard for cosmetics, and tolerancing, AND part finishing.  The Society of the Plastics Industry sells them.  Get a set for yourself and one for your customer.

“Parts free of defects” is not a quality standard; it’s an invitation to rejecting every part.  An inspector staring through a 5X magnifying lamp for 30 seconds at part the size of your thumb IS NOT IN ANY cosmetic standard.  Usually it’s 24 -36 inches under normal lighting conditions for 1 second per 2 square inches.  During that time the inspector must make a color match, and find all of the 17 cosmetic defects, AND know which won’t comply.   The last time you bought a car; did you whip out of magnifying lamp and stare at every square inch (interior and exterior) for a minimum of 30 seconds before you bought the car?  I don’t think so.

There are three components to the Universal Definition of Quality; it’s what your spouse thinks of you: (1) You work, for the function your serve.  (2) You are acceptably good looking, for the function in #1.  (3) With conditions #1 and #2 satisfied, size doesn’t matter.

Rejects from the production floor – The Solution
Sixty percent of your in-house production rejects are usually from not knowing the definition of an acceptable part.

An acceptable part is one that the end use customer (the guy/gal buying the product your customer is selling) will pay what he/she believes is an acceptable price and will get the expected use from the product.  NOTE – While little Johnny’s bicycle helmet might be shiny when his mommy bought it from the store; she isn’t going to return it because it is scratched after it bounces off the sidewalk with or without Johnny’s head in it.

An unacceptable part is one the end use customer will DEMAND an immediate refund/replacement because it (1) doesn’t work (Function) or (2) the aesthetics look so poor it screams inferior workmanship of impending functional failure (Cosmetics).

When was the last time you bought a box of nails and measured each one?  If you did, you’ll find they are not particularly precise.  But you didn’t inspect them because all you did was get your hammer and what you wanted to nail together.  If the nails were reasonably straight and not horribly rusted AND the only reason they bent was because you’re not very good with a hammer, you didn’t care how long they were.  Moral – the end use customer doesn’t dimensionally inspect the products.

Yields – Molds don’t make scrap.  You do.
If your setups aren’t repeatable train your people.  Document your processes; do what works.  Figure out how to translate conditions between machines of different sizes OR run the mold in the same press every time.

Test your machines for consistency.  If a telescopic sight isn’t firmly screwed to the rifle, while you’ll always put the target in the crosshairs, but the site will only occasionally be aligned with the barrel. (Hint: the bullet ALWAYS goes where the barrel is pointed) Great equipment is useless if you don’t maintain it.

Maintain your molds – are vents/waterlines blocked?   Clean them.  Do the techs hook up the water the same way every time?  Train Them

Make a Pareto Chart of Scrap of the scrap from this mold.  This is a bar chart with the largest amount of scrap to lowest.  Analyze it using the 80/20 rule – 80% of your scrap comes from 20% of the possible causes — Is it cavity specific?  Fix the cavity.  Is it random?  Look at your process, something isn’t consistent.  Is it specific to some machine?  Fix the machine.

For a part with an 80% yield that represents $5,000 in annual sales with only a 5% profit margin, spending $1,000 to raise the margin to 10% is a waste of money.

Training your people can triple your profits.  Demanding your customer give you reasonable requirements and not tolerating silly rejects can double your profits.

“We’re making too many rejects” isn’t a factual statement; it’s emotional.   Figure out what’s acceptable and work on it.

* * * * *

You can read this and think “I’ll never settle of anything less than the quest for zero defects.”  Feel free to do that until you bankrupt your company by trying to ship the Crown Jewels but invoicing for Rock Salt.  Or, you can put this in the Do Not File file and continue to complain just to hear your lips flap.  OR you can do some real work and change things.  You’d be surprised how once you implement and embed change you’ll look back and be embarrassed at how poorly you ran your company. OR you can print this out and use it to line the litter box of the cat who you keep for in-plant pest control.

It’s only your career, you choose

Bill Tobin

No responses yet

Trackback URI | Comments RSS

Leave a Reply