Re: procedures for procedures?

[Bob Flood <bflood@SLAC.Stanford.EDU>]

At 02:09 PM 6/23/98 -0500, you wrote:
>Doug asked about the "What if" method for procedures. I find his
>question interesting for historical reasons. When I first started in
>this position at Princeton over 30 years ago, what we did was set up a
>system of carefully planning out an administrative task by using logic
>diagrams to track the flow of paperwork and to help in the development
>of computer programs that were associated with the task.

In a previous life, when TVA's nuclear power program was shut down to heal
itself, a consultant was brought in to "fix" the procedure system.

The cure proved to be far worse than the disease, however. The system
imposed on the nuclear power program required the use of a very specific
writing technique, the usual document reviews we all think of as part of
the procedure development/revision process, and also required that the
process described in the procedure be diagramed (as in software
flowcharting). There was a review checklist that was 10-12 pages thick to
be completed by everyone reviewing the document. There was also a "test
drive" of the procedure to carried out by a qualified person with a
specific checklist to complete. There was a final review and concurrence by
a person previously certified as a subject matter expert and qualified as
an "independent reviewer," meaning this reviewer could not be part of the
writer's chain of command. This process was applied to all procedures. Each
feature of the system had its merits, but together they brought procedure
development and revision process to a halt. (a government tradition:
anything worth doing is worth overdoing.)

After learning the hard way (the way I learned most of what I know), the
system was simplified to include the usual review and the test drive using
the checklist. The independent review was retained for plant safety-related
procedures.

The test drive process was probably the most useful part. It meant that
someone who would have to use the procedure actually tried to execute the
procedure as-written. A great way to find flaws.

---------
Bob Flood
Dosimetry Group Leader
Stanford Linear Accelerator Center
(650) 926-3793
bflood@slac.stanford.edu