OptimalTREND™Wiki

OptimalTREND™ Wiki / Predictive maintenance

The frame

Predictive maintenance

Predictive maintenance fixes an asset just before it would have failed, on the evidence of its condition, not on a fixed schedule and not after it breaks. It is the maturity step most asset-intensive operations are reaching for.

The maintenance maturity ladder

ReactiveRun to failure, then repair. Cheapest to plan, most expensive to suffer: unplanned downtime, collateral damage, safety exposure.
PreventiveService on a fixed calendar or run-hours. Better, but it services good assets too early and can still miss random failures.
PredictiveService on condition, when the data says failure is developing. Right work, right time, minimum disturbance to a healthy asset.
PrescriptivePredictive plus a recommended action and its effect. Not just when it will fail, but what to do and what it buys.

The strategy mix, not a straight upgrade

Predictive is not simply the top of a ladder every task should climb. Mature reliability runs a deliberate mix, chosen per failure mode by RCM and the DIPF frame:

Run to failureA conscious choice where the failure is safe, cheap and non-critical. Legitimate when the consequence is trivial, reckless when it is not.
Preventive, time-basedFixed-interval service or replacement. Right where a component has a clear wear-out age, wrong for the random failures that dominate modern equipment.
Predictive, on-conditionAct on the evidence of a developing failure. The default for critical rotating and electrical assets with a detectable P–F interval. This is OptimalTREND™.
Proactive, precisionRemove the defect before it starts: precision installation, alignment, balancing, lubrication and contamination control. The cheapest reliability of all, upstream of every monitor.

The DIPF frame places these in their domains: proactive and precision work sit in the design and proactive domains, condition monitoring in the predictive domain. See the DIPF curve for how the choice is made per failure mode.

Why on-condition wins

Most failures give warning. Something changes – vibration, temperature, current, wear debris – well before the asset stops doing its job. Predictive maintenance catches that developing failure and converts an unplanned breakdown into a planned intervention, which is cheaper, safer and far less disruptive. It also fits how equipment actually fails: most failure modes are random rather than age-related, so a fixed calendar cannot catch them, while a developing condition can be detected and acted on within its P–F interval. The value shows up as higher availability and lower maintenance cost.

What it returns

Across Optimal client engagements, structured reliability and predictive programmes return value in these ranges, shown for context, not a guarantee:

Asset availability+3–7%
Maintenance cost−8–20%
Productivity+10–25%