Re: Specifications

[Myron Tribus <mtribus@earthlink.net>]

John Constantine, it seems to me, mixes up the concepts of 
specifications and reliance upon inspections.

Specifications are necessary so that buyer and seller can come to an 
agreement regarding the goods and services purchased.  Inspection to 
assure that specifications have been met simply do not work.  The 
fault lies not with the specifications (though they may be faulty at 
first) but rather with reliance upon inspection.

Reliance upon inspection simply does not work.  Safety can only be 
achieved by designing processes which prevent accidents.  Even then, 
the people working in the process need to be continually involved in 
removing sources of error.

Examples from John

>Duke University had myriad specifications when transplanting a heart and two
>lungs recently. Inspections at every possible level, by the most medically
>capable persons, and yet the blood type was somehow mismatched and the
patient died after a second transplant ultimately failed.:

I would like to see a deployment flow chart which detailed the steps 
taken when a heart transplant is undertaken.  Does anyone know if 
Duke uses them?  This was a failure of process,not specs.
>
>The recent fire at a Rhode Island club indicated that all requirements had
>been met, all inspections were up to date, and yet over 90 persons died.

The latest information appears to show that the owners bought the 
cheapest wall covering.  When an evening is planned, does the 
management have a flow chart of what is to happen?

>Site inspections indicated that a type of missile exceeded allowable limits
>by several kilometers. This of course has implications for many nations.

This example hardly qualifies as an example of either the value or 
harm associated with inspections or specifications. We are dealing 
with a man who is deliberately attempting to threaten and conquer his 
neighbors.

>A recently designed housing ordinance relies on inspections in order to
>either approve or disapprove the construction of all new housing per
>specifications.

So???

>The Columbia space shuttle crashed recently, and a expert panel has been
>appointed to look into the reasons why it did.

No comment here as no one seems to know just what happened and why.

>Federal specifications have been applied to airport security, educational
>requirements, immigration policy, social security eligibility, SEC oversight
>of corporations, SUV (and other) safety considerations, and the list goes
>on.

The difficulties, it seems to me, lie not with the use of 
specifications but with reliance upon inspections and failure to 
analyze processes.

>As Deming and Wheeler point out, a specifications approach is restrictive by
>its nature, allowing for only one of two outcomes, either in/out, yes/no,
>good/bad, right or wrong. This is fraught with problems, as there will be
>occurrences which are outside the bounds of specifications.

To the extent that managements rely upon outcomes as the basis for 
making decisions, they will generate great difficulties for their 
enterprises.  Outcomes are lagging indicators and, as I have written 
before, managing by outcomes is the same as driving your car down the 
highway, guiding it by the yellowline in the rear view  mirror.

>
>There will be decisions made based upon the specifications (statutes,
>regulations, ordinances, policies, designs, programs). If there is not
>another means of interpretation than simply specification, there will be
>occurrences which will bring about more variation in the very process or
>system under consideration.

That's what continual improvement is all about!

Myron Tribus,   Web Master www.icelp.org
350 Britto Terrace,   Fremont,  CA  94539
Ph:510 651 3641  FX: 510 656 9875
mtribus@earthlink.net
"Experience is a great teacher.  Unfortunately there are few students."