Condition-based maintenance (CBM) is a simple idea with complicated plumbing. The idea: do the work when the asset condition indicates it, not when the calendar indicates it. The plumbing: sensor data, thresholds, decision logic, and a maintenance system that turns a condition event into a real work order with the right technician, the right parts, and the right closeout evidence. A CMMS is where that plumbing lives. Without it the signals pile up on a dashboard and the work never happens.
PricewaterhouseCoopers’ “Predictive Maintenance 4.0” study, one of the most-cited multi-country maturity surveys with 250 companies across Belgium, Germany, Netherlands, Norway, and South Africa, found that two-thirds of respondents still operate at PdM maturity levels 1 and 2, and only 11 percent reach level 4. The gap between levels 2 and 4 is almost entirely about work execution discipline, not sensor count.
What Condition-Based Maintenance Needs From the CMMS
A serviceable CBM program asks the CMMS to do six things.
One, hold an accurate asset register, with criticality ranking, failure modes, and the sensors tagged to each asset. Without this the condition signals have nowhere to land.
Two, ingest telemetry and convert threshold breaches into work. The trigger rules live inside the CMMS or an adjacent layer that feeds the CMMS with pre-filtered events.
Three, prioritize the resulting work orders by consequence, not just by alert severity. A small deviation on a critical pump usually beats a larger deviation on a redundant asset.
Four, dispatch the work with the right technician, parts, and permits attached. Condition triggers that generate work orders without parts or permits stall in the queue.
Five, record the intervention with failure codes, findings, and photos, because the next model iteration depends on clean closeout data.
Six, close the loop back to the condition monitoring layer so that the trigger library improves with each cycle.
The Signals a CMMS Can Act On
Most plants have more signal than they realize. A partial list of what a CBM program routinely uses:
- Vibration amplitude and spectral bands on rotating equipment
- Temperature and thermal deltas on bearings, motors, and electrical panels
- Pressure and flow across filters, heat exchangers, and pumps
- Motor current signature analysis on driven assets
- Ultrasonic leak detection on compressed air and steam systems
- Oil analysis results logged against lubrication PMs
- Runtime and cycle counters from PLCs and drives
Each of these feeds a threshold or trend rule. The AI-powered maintenance layer adds multi-variate models where single thresholds miss. Both approaches share the same execution endpoint: a work order in the CMMS with a responsible technician.
Typical Outcomes After a CBM Program Matures
Operations that run a disciplined condition-based program inside their CMMS for 12 to 18 months typically report:
- 20 to 40 percent reduction in unplanned downtime on covered assets
- 15 to 30 percent increase in MTBF on rotating equipment
- 10 to 20 percent reduction in emergency work order volume
- 25 to 45 percent reduction in secondary damage from deferred repairs
- PM labor freed up for upgrade and reliability improvement projects, often in the 15 to 25 percent band
Siemens AG’s “True Cost of Downtime 2024” report provides the cost context: large plants lose an average of 27 hours per month to unplanned downtime, down from 39 hours in 2019, with auto OEM line stops running up to $2.3 million per hour. The hours recovered by a working CBM program translate directly into avoided downtime cost at those rates.
Where CBM Programs Stall
The same three failure modes recur.
Over-alerting. Trigger libraries set by the sensor vendor, not tuned to the asset context, produce thousands of notifications. The CMMS has to filter by consequence before the work lands in the technician queue.
Missing closeout discipline. If work orders close without failure codes or findings, the CBM program loses its feedback loop. The first six weeks of program maturity usually involve rationalizing the failure-code picker.
Parts misalignment. Condition-based work orders that arrive without the right part reserved stall. Integration between parts and inventory and the CBM trigger layer keeps the flow moving.
The Rotating-Equipment Case
Rotating equipment is the textbook CBM domain. Pumps, motors, fans, gearboxes, and compressors produce readable vibration and temperature signatures that mature into dependable condition triggers. For an energy or process operation with hundreds of rotating assets, the program typically starts with the top 50 criticality-ranked units. Once the CBM trigger library on that cohort holds for two maintenance cycles, the program can expand to the next 100 assets with confidence.
For electrical and thermal systems, infrared inspections on a route schedule feed the CMMS as periodic condition reports, producing inspection findings that become work orders the same way continuous sensors do.
Governance the Program Depends On
Three governance decisions determine whether the CBM program survives the first year of staff turnover.
The trigger library is owned by a named reliability engineer, reviewed quarterly, and versioned.
The failure-code picker is owned by the same engineer, not left as a free text field, and enforced at work-order closeout.
The consequence-weighted priority logic is written down and reviewed when the criticality ranking changes.
Frequently Asked Questions
Is CBM the same as predictive maintenance?
CBM uses current condition signals. Predictive maintenance uses models to forecast future condition. Both live in the same CMMS workflow and are often used together.
How many assets should we start with?
The top 10 to 20 percent by criticality, measured by downtime consequence. This usually covers 60 to 80 percent of downtime risk.
What sensors are mandatory?
None, in the strict sense. Many CBM programs start with runtime meters and manual vibration readings on a route, then add continuous sensors where the failure mode justifies them.
Does CBM replace preventive maintenance?
Not entirely. Lubrication, inspection, and cleanliness PMs remain calendar-driven. CBM replaces or extends the overhaul-interval type of PM where wear rates vary by duty cycle.
What does closeout discipline look like?
Failure code, findings, photo of the part removed, time on task, and parts actually used. That package is the training data for the next iteration of the trigger library.
Condition-based maintenance is a coordination problem as much as a sensing problem, and the CMMS is where the coordination happens. Book a Task360 demo to see the CBM workflow applied to your own asset base.
