Myself and just about every other Quality individual can identify with the 
frustrations of dealing with "Reactive" quality systems.  It seems that you 
have fallen into the trap of too much data, and not enough information.  I 
have worked hard over the last couple of years in trying to determine which 
data to collect within our manufacturing processes.  Every company is 
different and so will its data collection schemes be different.  There is, 
however a universal starting point.  "Pareto".  Most companies don't even 
know which problems are the most prevalent as far as resources spent.  They 
then spent time and energy on trivial problems that won't give them the most 
"Bang for Their Buck."  Start simple and collect only "Attributes" data. 
 Perform  simple Pareto analyses to determine your leading cause of 
rejection.  Work only this issue and continue to take data.  The next month 
you will see this rejection has moved down the scale and a new problem has 
taken its place.  This type of analysis will help you keep "score" on your 
continuous process improvement efforts as well as providing insight into the 
anticipation or recognition of developing problems.  Other supporting short 
term data collection schemes can then be generated, but only in support of 
identifying and separating the "Vital Few" from The "Trivial Many."

Good luck, its worth the effort.

Darrell Bonzo
Quality Engineer
EDO Corporation

 ----------
From: Todd Rogers
To: dabonz; jasatt; mimall; Jim Moffit (by Way Of Jack.cr
Subject: quality data collection


I am the quality supervisor of a single and double sided printed circuit
board manufacturer.  Over the past several years our company has developed
a "reactive" quality data collection system.  Every time there was a
problem we would start to collect a new set of data.  This has brought
about a huge amount of information and hours of time crunching numbers.
The bottom line is that there is no organization to all the information.
This makes it difficult to analyze and see trends that would be associated
with a continuous improvement effort.  This problem may sound very
elementary and easy to fix, but when you are caught up in the system it is
difficult to see obvious solutions.  My question is, What types of data are
other companies in the manufacturing of PCB's collecting concerning:
 1) Rejects and defects, ( reworkable vs. scrap)
 2) Customer Return information  (lots, parts, panels, ppm, etc.)
 3) Discrepancies ( to a department, process, reject type)
 4) Other information that may be necessary to develop improvement
planning.

Currently we have more data than is needed.  What is the minimum amount of
data for the maximum amount of effectiveness?  What I am looking to do is
annihilate the current system and only implement those measurements that
are needed to get an overview of how well the company is doing,
improvements that need to be made and if corrective action has been
effective.  I am also interested in the measurement unit.  A list of those
measurements, charts and associated formulas would help greatly and any
other information that companies are tracking would help get my perspective
back on-line.  I would appreciate any comments to this situation and
apologize for the lack of complexity of this particular situation.
Thank You!!!

Todd Rogers
[log in to unmask]



***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************