Corrective Actions - Return On Investment

[GARY MOORE - 2009]

corrective actions

Does all material returns from vendors have to result in a corrective action? Thanks

 

[noboxwine]

SAVE THE TREES AND YOUR TIME

No. Let's walk through it. See the reams of paperwork that would generate and the amount of time to process all the C/A. imho it's Cost vs Benefit. It wouldn't be cost effective to issue a C/A for all returns but unquestionably, there will be a time that it's necessary. I have always done a simple Pareto Analysis to evaluate returns over time. It will show trends and repeat occurances and will illustrate where an opportunity may lie. Outside of that, if the dollar amount and impact of one event is significant enough, I would issue one as well. Let the data point to the improvement areas to ensure a return on investment. Hope this helps and good luck.

 

[Claes Gefvenberg]

Re: SAVE THE TREES AND YOUR TIME

noboxwine said:
----X--------
Outside of that, if the dollar amount and impact of one event is significant enough, I would issue one as well.
-------X-------

This thread could become interesting.

I agree with boxman, and can provide this example of when the price of the item was unimportant:

Let’s say that a customer returns a screw worth $1, or maybe less, and it came from a lot of 1000 pcs. So is it worth bothering with? Well, that rather depends, doesn’t it? Where, and for what is it used?

I can recall an occasion when I went through all the motions and returned a screw like that (Ok, I actually returned the entire lot). Both my superiors and the supplier first assumed that I had finally lost the tiny bit of sense they assumed I’d once had. Until I got the message across to them…

That bolt was to be used inside a rather complicated transmission.

If it had been used it would have caused a complete breakdown within a few hours of operation.
The transmission would probably have had to be scrapped or at least undergone extensive repairs.
The cost would have been about $7000, not counting the down time of the vehicle.

Lesson learned: You need to know how the item is used…

/Claes

 

[gpainter]

You need to decide when a car is needed for returns and be consistant in the application.

 

[M Greenaway]

Claes

If one screw can cause so much damage I suggest you re-design your product.

Screws are screws, in a bag of 1000 you may well get one or two with a stripped thread, or some such other fault.

If your company performed a good design FMEA and considered the function of this screw, and its potential for failure (and consequential effects) then you would have probably taken re-design actions.

The onus must be on the purchaser to know what he is buying, and use it correctly. You cannot make someone banging out screw by the bin load to suddenly upgrade the quality or specification of the screw just because you use it to hold the wings on your aeroplane.

Sorry for the exaggeration, but do you get my point ?

 

[D.Scott]

I'm for that!!

We are in the business of applying adhesives and sealant to our customer's bolts, screws and nuts. We work with lot sizes that average somewhere around 50,000 pieces.

The latest "buzz word" is ZERO DEFECTS. Unfortunately, they don't yet distinguish between a defect and a nonconformance. If we send out a part requiring 2 - 4 lead threads with only 1 1/2 lead threads it is counted as a defect and the entire lot is subject to rejection. I should point out that most of our customers have more sense, but there are some who would then insist on sorting every part and sending us the bill or rejecting the entire lot for re-work.

The importance of the nonconformance usually is determined by the customer. Even when we discuss it up-front and think we understand each other, it comes down to customer perception.

I guess when you work with screws you have to expect to get screwed.

Dave

 

[Claes Gefvenberg]

Sure I do Martin, and you are right of course.

On the other hand: Take just about any component in a gearbox... Just how much does it take to cause a full scale disaster? Not a lot, let me tell you...

As for screws, I don't agree. There are screws and screws.

Anyway, in this instance it was not as simple as a faulty thread or something similar and there was no question of upgrading the specification. The screw was going to be used for what it was designed for, and it was in fact overdesigned to fare thee well.

But: The heat treatment was at fault in part of the shipment.

/Claes

 

[Claes Gefvenberg]

Re: I'm for that!!

D.Scott said:

-------X-------------
I guess when you work with screws you have to expect to get screwed.

Dave

Errrrr.... Can't argue with that, Dave

/Claes

 

[energy]

Round and Round we go!

I agree with NoBox, Claes and Dave. What it comes down to is the "Determinator" decision to pursue CA. Every problem reported to me is considered by the reportee to be significantly important. A lot of it is crap to me. I view it like NoBox. Trend it. Also, we are not talking critical parts and disastrous equipment failures. But, it's still a problem for the "Determinator" to decide what needs CA and doesn't, which was the reason for the original post. I say, if you are the one in position to make the decision, be very reluctant to apply CA if you can. Assure the person who reported an incident that you will be monitoring these types of glitches in the system and the data will be used to report to Management during Mgt Review. JMHO

 

[Carl Exter]

We allow the gravity of the nonconformity to determine CA or not. We look at the effect the problem would have on the finished good. If it's significant, a CAR will be issued. If not, then we will still file a nonconformity report, which will be copied to the supplier. We also publish monthly and quarterly supplier ratings, like a report card, of which # of nonconformity reports/# of lots received is a significant part. So far it's worked pretty well for us.

Likewise, with product returned to us from our customers, each piece is analyzed and that data goes into a database from which we can discern problem trends, return rates for various SKUs etc. That information is fed-back into engineering design.

Usually we don't issue a corrective action from product returns per se. If a problem is significant enough to the customer to require corrective action, we'll hear about it long before the product gets back to our dock! Hope this helps.