# Continuous Improvement in CBM

## Presentation on theme: "Continuous Improvement in CBM"— Presentation transcript:

Continuous Improvement in CBM
What is it? How do we measure it? How do we make it happen?

Continuous improvement in predictive maintenance
First we define the Conditional Probability Density Function. Conditional Probability Density Function, CPDF t0 Working age Current time

First we define the Conditional (Probability) Density Function
It is like the Probability Density Function. Only it is calculated at the current moment in time, and it takes into account the current age and the current condition of the item. Conditional Density Function t0 Working age Current time

Next we define the Conditional MTBF
Is the well-known MTBF, except that it is measured from the current moment. That is, the state from which one needs to make an on-condition maintenance decision. Conditional Density Function t0 Working age Current time

Remaining Useful Life Estimate (RULE)
The Conditional MTBF is also known as the RULE (Remaining Useful Life Estimate). Conditional Density Function RULE t0 Working age Current time

Predictive performance is measured by the standard deviation.
One performance measure of a predictive maintenance program is the spread of the distribution around the mean (i.e. around the RULE). This spread  is quantified by the standard deviation. Conditional Density Function RULE t0 Working age Current time

Continuous improvement in predictive maintenance
Continuous measurable improvement in the RULE occurs over time as more experience is gathered. Not only is the RULE adjusted, but the spread narrows. Conditional Density Function Improved RULE t0 Working age Current time

Continuous improvement in predictive maintenance
And narrows further. The RULE improves. Confidence in prediction increases in a measurable way so that it may be reported as a CBM “KPI”. Conditional Density Function Improved RULE t0 Working age Current time

RULE and Standard Deviation reported in EXAKT
Every time the EXAKT decision agent runs it reports the RULE and the standard deviation.

How do we improve confidence in Predictive Maintenance?
By reporting both failures and suspensions when closing work orders. By continuously improving the RCM knowledge base when closing work orders. By linking work orders to RCM knowledge records. By generating samples from the CMMS. By applying reliability analysis techniques including EXAKT. These steps describe the LRCM process.