Subjective parameters and Quality costs - VDA 16

We have many subjective parameters to chack on our products, many surfaces, (VDA 16).

Of course, operators being humans the more products that go through the station the more faults are found in our CSP control.

The problem is, the more faults that come there, the higher risk that they come out to the customer.

Has anyone any experience of how we can reduce the number of faults being overseen by the operators?
I am thinking of maybe a study concerning time at station contra Q faults found after that station

Re: Subjective parameters and Q-costs

David Bamford said:

We have many subjective parameters to chack on our products, many surfaces, (VDA 16).

Of course, operators being humans the more products that go through the station the more faults are found in our CSP control.

The problem is, the more faults that come there, the higher risk that they come out to the customer.

Has anyone any experience of how we can reduce the number of faults being overseen by the operators?
I am thinking of maybe a study concerning time at station contra Q faults found after that station

Click to expand...

If you have identified the failures why not concentrate on finding the cause(s) and ultimately trying to avoid the failures and the causes.

Best regards,

Antoine

Re: Subjective parameters and Q-costs

David Bamford said:

We have many subjective parameters to chack on our products, many surfaces, (VDA 16).

Of course, operators being humans the more products that go through the station the more faults are found in our CSP control.

The problem is, the more faults that come there, the higher risk that they come out to the customer.

Has anyone any experience of how we can reduce the number of faults being overseen by the operators?
I am thinking of maybe a study concerning time at station contra Q faults found after that station

Click to expand...

If faults are being found after products have been inspected you have to ask why the inspectors can't find them. Perhaps an attributes MSA would be helpful--you have each inspector evaluate a controlled lot of parts (some with defects) and see how well the inspectors agree on what's acceptable and what's not. If there's significant disagreement, you'll know that you have a training issue, or that the requirements are too subjective.

If you haven't already done so, you might also want to use boundary samples--parts that are acceptable and parts that aren't--so that inspectors have a more objective frame of reference.

Re: Subjective parameters and Q-costs

Jim and Antoine made some good suggestions.

An attribute study should include inspectors that are at the end of their shift or just before a break. (to show influence of fatigue)

You should focus on eliminating the defects and finding the defects upstream (at the source).

In the mean time:

Are your inspectors checking for too many faults?
Are they getting fatigued? At what time are the defects found?
How is the lighting in the inspection area?
How are the other ergonomic aspects of the inspection process?
Are they finding more or fewer faults at certain times of the day?
How are they trained? Do you have well defined boundary samples?
Do they have aids to help with inspection?
Are they getting distracted (noise, traffic, people, etc)?


Create a fish bone / cause effect chart for the inspection process.
Eliminate cause and improve the inspection process

Re: Subjective parameters and Q-costs

We find that a number of them originate from the anodization process, this has been optimized as much as possible though it must be remembered that we actually control products to a standard higher than VDA 16, it is the inspectors that are being fatigued over a period of time

Re: Subjective parameters and Q-costs

We have valid MSA's and trained operators, our physical enviroment is VDA 16 compliant, and a low noise level.
The actual border samples are well defined and documented, it is simply a matter of fatigue/boredom when inspecting the same products over a longer period.
This is why I assume it will help with a break, (different product/different station).

I'm confused by the statement "to a higher standard than VDA-16". I am not that familar with VDA-16 but I thought it applied to class A decorative type surfaces, usually with 1000 lumens at 120 cm (ouch!). Line of sight items for the car driver and passengers.

A simple solution to minimize or determine the impact of human error is to do an attribute study, 50 pieces x 3 x 3 and focus on the kappa measure (see MSA 4th edition). This will ensure every inspector understands all the possible defects. As a TS auditor, I asked a steel mill I audit to do this and they used photos of the various defects as part of their samples (beats having 50 coils!) and it drew the interest of the other staff. Lo and behold, it turned out the quality manager scored the worst, so this should tell them he shouldn't override his inspectors.

I agree with the earlier poster to try to eliminate as many sources of the variations, but if you are still dependent on visual inspections to screen out defects, as many are , you should see if that inspection can be optimized by eliminating the human error. If you can't, invite a few vision system salespeople in to show what they can do for you.

Good luck,
-outoftown