RE: Process Capability - examples?

["Prevette, Steven S" <Steven_S_Prevette@rl.gov>]

Bob asked: Does anyone have an example of how the determination of
process capability has contributed to successful business improvement.
I am looking for real examples, rather than theory.

Determination of the process capability in itself doesn't contribute
much.  The primary realization is - do you have a stable process that
needs to be improved, or do you have a changed condition (trend) that
you must correct or redesign for.  Here is a real story:

My last job in the US Navy was at a torpedo maintenance facility.  We
were not capable of meeting the SUBLANT (US Submarine Forces Atlantic)
of processing 10 torpedoes a week.  This was definitely stable and
predictable.  We had tried reacting to high and low weeks, but nothing
changed.  

Now, this was when the Navy was "doing Deming", so our commanding
officer wisely did not push the goal on the sailors, but did push that
we needed to improve our production capability.  I ended up gathering
data on times to do tasks (there was an overall flowchart of what needed
to be done to a torpedo) and the flow rates through our automated test
equipment.  I found two things that once corrected we easily made the 10
goal without breaking a sweat.

First - there were a large number of torpedo engines waiting to be
cleaned, and a large number of torpedoes stalled in the process waiting
for engines.  Come to find out, at the 110 person facility, only two
sailors were qualified to clean engines, and they were the two most
junior sailors.  Cleaning engines was a dirty job no one wanted to do.
If either of the two were away on other tasks, engines didn't get
cleaned.  Several more senior sailors were then made to enjoy this task.

Second - we had four automated test lines to test the torpedoes at the
end of the process.  The technicians were very proud of their equipment,
and if any line went down, they were there all hours of the night and
weekend to repair the line.  They kept logs of the lines status.  Review
of the logs showed that, even taking into consideration downtime for
preventive maintenance, one of the four lines was ALWAYS idle.  Even if
we hit 10 torpedoes a week, it looked like one line would be idle.  I
convinced the commanding officer that one line was an installed spare.
If one went down, don't worry, don't do anything heroic to fix it.  Yes,
fix it, but no heroics.  But if a second line went down we were in
trouble.  Yes, some of the folks dedicated to the test lines learned how
to clean engines.

Bob's second question:  Any ideas on how to introduce management to the
concepts of the theory of variation, control charts etc would also be
welcome.

I would highly encourage you to use the Red Bead Experiment.  I got
farther in one session of the Red Beads than in six years of argument.

Steve Prevette
Occupational Safety and Health
Fluor Hanford, A Fluor Government Group Project
ASQ Certified Quality Engineer
steven_s_prevette@rl.gov
509-373-9371