I generally think of Document Control as what happens to a "good" document
(documents are the "plans" for performing an action versus a "record" which is a special kind of document which "records" the action taking place.) in the time from approval to actual production and after initial production
(or action) if circumstances indicate a change is required. Documents may be revised from time to time by a systematic method
(some folks have a "document change" or "engineering change" process.)
"Control" means determining who may
- see
- copy
- revise
- store
- use
a document.
If you have a good document control system, you can still have BAD documents, like some of the examples above, simply because a typographical error or an error in judgment can slip through ANY system, no matter how many checks and balances are in place.
The frequency and egregiousness of the errors increases when one or more of the folks who are supposed to perform the checks and balances are ill-prepared and ill-motivated to perform those tasks in a competent manner.
The incident described above with the ship poorly designed to make the "power pack change out" procedure impossible to perform as intended is really a failure to truly follow a Configuration Management system as part of the overall Document Management and Control System.
At its simplest level, Configuration Management is intended to assure an obsolete document is not used on a current project. As envisioned and practiced by the aerospace industry for decades, Configuration Management is ALSO supposed to entail ensuring all Associated Documents are reviewed and checked to determine they will jibe with the revised document and with each other.
Elsewhere,
here in the Cove, I outlined a relatively simple array of things which need to be checked and possibly ALSO modified as a result of one tiny change in a product design (changing a fastener from Phillips head to Torx.)
Analyzing a change and reviewing Associated Documents is not a simple clerk's job - it requires knowledge of the entire process and ability to spot how each change may affect form, fit, and function of other parts of a process, some of which may be subtle, while others can have catastrophic consequences.
My point in the first post I made in this thread is that it is not necessarily the catastrophic lapses which cause the headaches in industry, because they are relatively few and far between. Far worse are the nickel and dime lapses which add up to far greater overall damage.
Consider a weak analogy between the number of American casualties in Iraq over the last four years (about 5,000) [source:
http://www.icasualties.org/oif/USbyYear.aspx ]
and the number of deaths due to automobile accidents in 2007 alone (about 41,000) [source:
http://en.wikipedia.org/wiki/List_of_motor_vehicle_deaths_in_U.S._by_year]
My local paper lists the death toll in Iraq (thousands of miles away) every day, but it doesn't list the deaths (about 24,000) in California each year due to air pollution
A state report (.pdf) to be presented today [May 22, 2008] says that "as many as 24,000 deaths annually in California are linked to chronic exposure to fine particulate pollution," according to the LA Times.
[source:
http://laist.com/2008/05/22/24000_californi.php]
My point being the relatively unnewsworthy events wreak more damage than the newsworthy ones.