My Concern

 Capability 

So what is my point in all of this? I am certainly not taking a pop at RCM, although it does concern me that it will be used for cost saving, not to improve availability (it may not in any case) unless safety or lives are at risk.

In his book ⁽²⁹⁾, Dubi makes the observation that we use the throw of a die in a 'statistical experiment' to demonstrate the meaning of probability. He goes on to explain that each throw is slightly different; the angle of the throw, the position of the die, the velocity etc. He asks whether there are "true" statistical experiments and if there can be a "unique deterministic" future.

My concern is that we don't understand availability and consequently the data associated with it.

What is availability?

What do we measure?

What data is required?

How do we interpret the data?

Key to answering these questions is the deterministic view of the quantitative, empirical, data that is currently taken by many professional engineers, even though there is evidence to dispute this view.

Dubi, Knezevic, Basu and Barlow essentially agree that classical statistics are logically untenable when it comes to predicting the future. Therefore application of statistics can only give a deterministic view of the historical availability up to a point in time and cannot give a probabilistic perspective with any level of confidence or regularity of availability at any future point in time. Therefore until we change the culture that largely prevails, we will remain in the maelstrom.

Finally the foregoing begs me to ask this last question.

If we understand science, shouldn't we have the conscience to use it?
"Fare il vino e semplice ma non facile"
(Making wine is simple but it's not easy)
- Italian Proverb

Or the British version.

"You can't make an omelette without breaking a few eggs"

 Back to Intro