Condition-Based Maintenance: Fixing Equipment on Evidence

Key takeaways

• Condition-based maintenance (CBM) triggers work on the actual measured condition of an asset (vibration, temperature, pressure, and similar), acting when a reading crosses a threshold rather than on a fixed calendar.
• It sits between calendar-based preventive maintenance (which can over-maintain healthy assets) and predictive maintenance (which forecasts a failure date). CBM reacts to present condition, not a forecast.
• The payoff is doing maintenance when it is actually needed: fewer surprise failures than run-to-failure, and less wasted work than fixed-interval PM.
• CBM only works with a measurable degradation signal, a sensible threshold, and a committed response. Monitoring with no threshold or no action is just data collection.

What condition-based maintenance is

Condition-based maintenance means you watch a measurable indicator of an asset's health and act when it degrades, not when the calendar says so. A bearing's vibration creeping up, a motor running hot, a pressure drifting out of band: each is a condition signal that triggers maintenance before the asset fails, but no sooner than necessary.

The logic is that most components do not fail on a neat schedule; they show signs first. CBM catches those signs and converts them into timely, targeted work.

CBM versus preventive versus predictive

The three are often confused:

• Preventive (calendar-based): service every N hours or weeks regardless of condition. Simple, but it over-maintains healthy assets and can still miss early failures. See the preventive maintenance schedule.
• Condition-based: act when a measured condition crosses a threshold. Responds to the present state of the asset.
• Predictive: use trends and models to forecast when failure will occur and plan around it. More data-hungry and harder to do well.

CBM is the practical middle ground: it needs real measurements but not a full predictive model.

When CBM pays off

CBM earns its cost on assets where three things hold: the asset is important enough that failure hurts (see asset criticality analysis), the degradation is measurable with a sensor or inspection, and there is enough warning between the signal and the failure to act. On low-criticality assets with no measurable signal, run-to-failure is usually cheaper.

Common mistakes

• Monitoring with no threshold. Collecting vibration data nobody acts on is cost without benefit. Define the trigger level and the response in advance.
• Applying it everywhere. CBM on a trivial, redundant asset wastes sensors and attention. Target it by criticality.
• Confusing CBM with predictive. CBM reacts to current condition; it does not forecast a date. Claiming a failure prediction from a simple threshold oversells what the data supports.

How Fabrico fits

Fabrico connects condition signals and downtime events to the asset and its maintenance history, so a threshold breach can raise an alert and become a work order in the same system, with the event logged for analysis. It focuses on acting on measured condition and true-cause downtime rather than overselling forecasts. The closed path from detection to a tracked fix is covered in automatic downtime tracking. Fabrico is built and hosted in the EU with data residency in mind and is ISO 27001 certified. To act on condition rather than the calendar, book a demo.

Related reading

• Preventive maintenance schedule
• Asset criticality analysis
• Automatic downtime tracking
• OEE for manufacturing

For a practical next step, compare the leading options in our guide to the condition monitoring software.

Teams putting this into practice often review our roundup of the affordable CMMS software.

Frequently asked questions

What is the difference between condition-based and preventive maintenance?

Preventive maintenance is done on a fixed schedule regardless of condition. Condition-based maintenance is triggered by a measured indicator crossing a threshold, so work happens when the asset's actual state calls for it rather than on the calendar.

Is condition-based maintenance the same as predictive maintenance?

No. Condition-based maintenance reacts to the present measured condition. Predictive maintenance goes further, using trends and models to forecast when a failure will occur. CBM needs measurements but not a predictive model.

What should we monitor for CBM?

Whatever degradation signal precedes failure on that asset: vibration, temperature, pressure, current draw, or periodic inspection readings. The signal only helps if it gives enough warning to act before failure.

Which assets are right for CBM?

Critical assets with a measurable degradation signal and enough lead time between the signal and failure to respond. Low-criticality assets with no measurable signal are usually cheaper to run to failure.