Quality metrics for electronics assembly

I am a procurement engineer for printed circuit assemblies (PCAs) and as

such the quality and technical interface between several external

contract manufacturers and my company-internal customers. Product

portfolio is highly complex, high-mix/low-volume with monthly

requirements ranging from 5 to 500. We have in place a well structured

supplier evaluation and rating process, where quality plays a

significant role in the scoring.

The main quality metric, both for absolute and comparative supplier

quality, is customer reported defect rate (ppm per component). While

this metric is fairly accurate, the time-lag between PCA production and

defect reporting is easily electronic assembly several months. This not only renders it

almost useless for corrective action and process improvement at the

supplier, but also means that I have no direct visibility to what I’ve

got in the pipeline.

Historically, PCAs were inspected (on sample basis) upon receipt before

we sent them on to the customer, but this was discontinued for the

biggest part with the argument that as the CMs are doing both in-circuit

test and 100% final inspection, we shouldn’t have to duplicate this (I

fully go along with this line of argument). Problem is only that I also

lost a quality metric in the process.

Another quality measure I am/was looking at is the internal quality

level at the CMs, but found this to be difficult to compare. In

addition, I don’t really see much correlation between this number and

what my customers are actually finding. My biggest gripe about internal

defect rate is that it is internally focused, i.e. tries to provide data

for process improvement, while I’d like to know the quality level going

out the door and ending up at my customers (externally focused).

However, the data are available for any potential statistical

evaluation.

I’ve been working on this metric issue for quite some time now, but

don’t seem to be getting any further. What I’m looking for ideally is an

early warning sign that makes aware that I am going to be hit, but which

also would allow me to initiate appropriate (internal) action to either

counteract any ‘hiccups’ or provide my customers with a warning so that

they can react before they are getting a potential line-down situation.

I know that theoretically the supplier should do all of this internally

and shield me, the customer, from any (excessive – whatever this means)

defects, but in real life the priorities are not always set this way.

I’d appreciate any ideas, further thoughts, experiences or other input

that would help me get a better handle on this.