Re: Article on ISO & Quotas

["John Bruman" <jbruman@precisiondie.com>]

> "You cannot manage what you don't measure, You cannot achieve
what you
> don't measure (Lord Kelvin)."

I am sure you are very sincere in your opinions, and in your
interpretation of Lord Kelvin's statement, and the context from
which it was taken. I however cannot agree with you simply due to
the fact that what you are suggesting goes against nearly all of
the tenets I hold dear, and most of the important things things I
have learned through many years of management mistakes and
learning experiences.

> "Actually, measuring the number of inspections is not relevent
becasue
> one should measure the outcome result and not the means."

Measuring or counting outputs is inspection at it's weakest, and
has nothing to do with continual improvement or the true
understanding of Shewhart's Plan-Do-Study-Act cycle.
A much more enlightened approach is to control the inputs to a
process and thereby cause success rather than counting, measuring,
or analyzing failures.

What Mike was referring to was not inspection as you suggest, but
process auditing (which is also required by ISO 9000) to ensure
the process is conducted in accordance with the foundational
hypothesis upon which it was planned and designed. Without
assurance that the process is performed according to plan, we lose
all ability to form conclusions, make predictions, or learn from
our experiences. If his definition of success is to "Do" Receiving
Inspection according to the plan, what better way to ensure that
than to monitor and control the frequency of audits in the
Receiving Inspection Department.

Merely looking at the output as you suggest, and going back and
making adjustments based on the output, without reviewing and
verifying the Plan and Do step is mis-interpretion of the PDSA
cycle as a trial and error process.

Suppose I am to weld two parts together to form an assembly. I can
choose to control my process with a destructive "hammer and
chisel" weld test. This leaves me in a real predicament whenever I
fail a test because I never know exactly when my process went bad.

What do I do with all the parts run since my last test?
Are they Good or Bad?
Furthermore, Which ones are Bad?
Additionally; What do I do to correct my process?
What can I adjust? How much?

If through profound knowledge and/or designed experimentation, I
determine that a good weld is controlled by a certain combination
of voltage, clamp load, and dwell time however, I can chart
periodic readings of those variables, and monitor them for changes
which may produce a bad weld. Before long, I'll know when to
adjust what, without increasing the variability through twisting
knobs and tweaking dials (tampering).

Eventually, I should be able to abandon the destructive tests
entirely by controlling the input variables of my process. By so
doing, I eliminate output inspection, rejects, scrap and rework,
and all the costs they involved, while providing a more consistent
product to my customer.

Do you really have a better way?

"Beside, one
> should not forget that quality is "built-in" by an operator and
not
> added-on by an inspector".

Amen. You just supported my earlier statements.

> In many instances, I have seen companies required "a measurable
> improvement" without specifying a specific "rate". Just show
that you
> have improved by any percentage.

It all depends on how you define "improvement", and "success". As
a grandparent my satisfaction, joy, and loyalty to my
grandchildren grows every year, but I haven't measured any of them
even once. Believe me, they continue to improve in everything they
do. Their improvement is not from achieving goals or numerical
targets, but by a steady, slow process of continual maturation and
learning.

> Can any one play tennis, badmington, bridge, football, golf, or
any
> sport without measuring a result ????

Of course you can. If you define your success as having fun,
enjoyment, or fellowship. If you define your success by scoring
more points, or "beating me" you better bring along an auditor
(referee) to make sure I follow the rules. I've been known to
cheat.

> Food for thought

> Joseph Kelada
> HEC Business School
> University of Monteal

Yes Joseph, it certainly is.

John A Bruman
QA Manager