Condition-based maintenance
Maintenance triggered by the measured condition of an asset rather than by a fixed schedule.
How it works
Instead of changing oil every 500 hours, you check the oil and change it when it is degraded. Instead of replacing a bearing every year, you measure its vibration and replace it when the readings show wear.
The idea rests on the fact that many failures give a warning. Reliability engineers describe this with the P-F interval: the time between the point where a coming failure can be detected (P) and the point where the asset actually fails (F). Inspections must happen more often than that interval, or the warning is missed.
Common condition checks include vibration, temperature, oil analysis, ultrasound, pressure, current draw, and simple visual checks for leaks or wear.
Why it matters
Fixed schedules replace parts that still have life left and can miss parts that wear faster than expected. Condition-based maintenance spends effort where the asset needs it. It also catches problems before they turn into downtime.
In practice
Many teams start with route-based inspections: a technician walks a set of machines, takes readings, and records them. A reading outside its limit creates a work order. Fixed sensors come later, for critical assets where a missed warning is expensive.
The weak point is follow-through. A reading that is recorded but never compared with the last one is not condition-based maintenance.
Related terms
For a comparison of strategies, see Maintenance strategies compared.