What is a Weibull database?

The smartest way to maintain a reliability database is in Weibull format and Weibull databases are available. Seldom do you see Weibull databases from vendors because they jealously protect their data for proprietary reasons-they life/die financially from the Weibull database information.

Why use a Weibull database?

The Weibull databases simplify the complications of failure data into two statistical values of great importance:
b tells you HOW things fail, and
h tells you WHEN things fail.
The results are key benchmark data that tell you how you're doing.

When to use a Weibull database?

Gather your failure data and create your own database. No one is going to give you their database because they put much sweat and tears into cleaning up the data so it is useful. The data needs to be locally generated because it tells you: 1) the life from the grade of equipment your purchase, 2) it describes the grade of operation of the equipment-do you operate it like 16 year-old teen agers or wise old men/women of 65?, 3) it describes the grade of maintenance you use to renew it's life, and 4) it tells you management's expectations for how to treat the system.

Where to use a a Weibull database?

The data starts out as a silly exercise by maintenance to accumulate data with much ridicule from the unknowledgeable about why are you spending this effort to build a Weibull database. Then suddenly when adversity arises, it becomes everyone's prized possession. Remember the worlds of Runyard Kipling about the English soldier: In peace time it's Tommy this and Tommy that, and Tommy get out of the way.....but you let the bullets fly in wartime and it Mr. This and Mr That and Mr. if you please! Everyone wants the baby but no one wants the dirty diapers that go with every baby! If you don't have a Weibull database, you're already too late because your competitor has one started and it using it for your disadvantage and he's not doing to tell you why you're left in the dirt!

These definitions are written by H. Paul Barringer

Return to Reliability Tools