Guides ·
Condition-based maintenance in a CMMS: thresholds that raise work orders
Condition-based maintenance (CBM) means doing maintenance when a measured condition says it's needed, rather than on a fixed calendar.
By Upvantix team
Condition-based maintenance (CBM) means doing maintenance when a measured condition says it's needed, rather than on a fixed calendar. A reading — vibration, temperature, pressure, oil condition, runtime hours or a counter — is compared with an alert or alarm limit, and crossing that limit triggers an inspection or a corrective work order. In a CMMS, that rule ties the reading to the asset, the job plan, the parts and the person who does the work.
CBM sits between preventive and predictive maintenance. Preventive maintenance runs on time or usage intervals whether or not the asset needs it. Predictive maintenance uses trends and models to estimate when a failure is likely. CBM is the practical middle step: fixed thresholds on the parameters that matter for each failure mode, set from a baseline and reviewed as data builds up. ISO 17359:2018, the general guideline for condition-monitoring programmes, describes the same path: pick critical equipment, identify failure modes, choose what to measure and how often, set baselines and alert/alarm criteria, then decide the maintenance action.
The cost of getting this wrong is real. In ABB's "Value of Reliability" survey (Sapio Research, July 2023), 88% of Indian industrial businesses said they experience unplanned outages at least once a month, compared with 69% globally, at close to INR 7 million (about ₹70 lakh) per hour for a typical Indian business.
Preventive, condition-based and predictive: how they differ
Most plants run all three: PM for statutory and simple tasks, CBM where a parameter reliably shows deterioration, predictive where the value justifies it.
| Criterion | Preventive | Condition-based | Predictive |
|---|---|---|---|
| Trigger | Calendar or usage interval | A measured value crosses a set limit | A trend or model forecasts failure |
| Data needed | Asset list and intervals | Readings per asset and limits per failure mode | History, continuous data and analysis |
| Typical first step | PM schedules in the CMMS | Meter or inspection readings with thresholds | Pilot on a few critical assets |
| Risk | Over- or under-maintaining | Badly set limits (false alarms or misses) | Effort before the data is ready |
How a CBM rule works in a CMMS
- A reading arrives: entered by a technician on an inspection round, scanned at the asset, or fed from a sensor or historian.
- The rule compares it with limits for that asset and parameter: an alert limit (plan work) and an alarm limit (act now).
- The action is a work order or an inspection, with the job plan, parts and priority set by the rule; repeat breaches don't create duplicate orders.
- The record closes the loop: what was found, what was done, the reading after the fix, so limits can be reviewed.
Choosing what to measure (the ISO 17359 way)
ISO 17359 starts with an equipment and criticality audit and failure modes, then selects the measurement technique, monitoring interval, measurement locations, baseline data and initial alert/alarm criteria, and only then data acquisition, comparison with criteria, diagnosis and the maintenance action. Its annexes list typical parameters by machine type, such as vibration, temperature, oil and wear debris, flow, pressure, power and speed. ISO 13374 (Part 1, general guidelines, and Part 2, data processing) covers how condition-monitoring data is processed, communicated and presented between systems.
Threshold-to-work-order rules (examples)
Example limits must come from the OEM manual, the relevant standard and your own baseline — the table shows the pattern, not values to copy.
Rules of thumb for the CMMS set-up: one rule per asset-parameter-failure mode; suppress duplicate work orders while one is open; require a post-repair reading to close; review limits after every false alarm or missed failure.
| Criterion | Parameter (how captured) | Alert limit → action | Alarm limit → action | Record on close |
|---|---|---|---|---|
| Motor / pump bearing | Vibration velocity (route reading or sensor) | Above baseline by your set margin → inspection work order within the week | Above OEM/standard alarm → priority corrective work order, consider stopping | Reading after repair; bearing replaced Y/N; cause |
| Gearbox | Oil sample: particle count or water content (lab result) | Above alert → schedule oil change / filter work order | Above alarm → inspect gears and seals | Lab report attached; oil batch |
| Compressor | Discharge temperature (gauge or sensor) | Trend above normal band → check cooler and filters | Above trip-safety margin → corrective work order | Temperature after fix |
| Conveyor / machine | Runtime hours or cycle counter (meter) | Reaches service interval → PM work order (usage-based) | — | Meter reading at service |
| Filter / strainer | Differential pressure (gauge) | Above clean-filter value by set margin → replace filter work order | Above maximum → immediate replacement | Pressure after change |
| Electrical panel | Hot-spot temperature (thermography round) | Above set rise over ambient → planned tightening/inspection | Above critical rise → urgent work order with isolation | Thermal image before/after |
Rolling out CBM without a big sensor project
Start with readings your technicians already take on rounds; add sensors only where the failure mode develops faster than your inspection interval.
- Pick 10–20 critical assets using a criticality ranking (impact on safety, output and quality).
- For each, list the main failure modes and the parameter that shows each one early.
- Choose capture: inspection round, portable instrument, lab sample or fixed sensor — match the interval to how fast the fault develops.
- Record a baseline on healthy equipment; get alert and alarm limits from OEM manuals and relevant standards.
- Set up the asset, meter/reading points and job plans in the CMMS; link each rule to a work-order template and parts.
- Decide who reviews alerts, and the response time for alert vs alarm.
- Run for one quarter; log false alarms and misses.
- Adjust limits and intervals; retire calendar PM tasks that CBM now covers, keeping statutory inspections.
- Track results with existing KPIs: unplanned downtime, MTBF, PM compliance, emergency work share.
- Only then consider continuous sensors or predictive analytics for assets where CBM falls short.
When to move from CBM to predictive
When you have reliable readings and work-order history on critical assets, and fixed limits keep giving false alarms or late warnings, trend-based or model-based prediction becomes worth testing. See our predictive maintenance guide.
Where Upvantix fits
Upvantix CMMS is designed to organise the asset register, preventive maintenance schedules, inspections and checklists, work requests and work orders, and completion records with photos, notes and measurements, with QR access to the asset at the point of work, and is designed for deployment in your AWS environment. If you're planning condition-based rules, talk to us about how readings, thresholds and work orders should connect in your plant.
This guide is general information as of 12 October 2026. Threshold examples show a pattern, not values for your equipment; set limits from OEM guidance, relevant standards and your own baseline, with your reliability and safety teams.
Common questions about condition-based maintenance
What is condition-based maintenance?
Maintenance triggered when a measured condition — such as vibration, temperature, pressure, oil condition or runtime — crosses a set limit, instead of on a fixed calendar.
How is CBM different from preventive maintenance?
Preventive maintenance runs on fixed time or usage intervals. CBM acts on the measured state of the asset, so work happens when readings show it's needed.
How is CBM different from predictive maintenance?
CBM compares readings with fixed alert and alarm limits. Predictive maintenance uses trends or models to forecast when a failure is likely. CBM is often the step before predictive.
Do I need sensors for condition-based maintenance?
No. Many plants start with readings taken on inspection rounds or with portable instruments. Fixed sensors make sense where faults develop faster than your inspection interval.
How do I set CBM thresholds?
Start from OEM guidance and relevant standards, record a baseline on healthy equipment, set an alert and an alarm limit per failure mode, and review limits after false alarms or missed failures. ISO 17359 sets out the general procedure, including baselines and alert and alarm criteria.
What does a CMMS do in CBM?
It holds the asset, reading points and limits, raises the work order or inspection when a limit is crossed, and keeps the record of what was found and fixed so limits can be improved.
Is CBM part of field service management?
No. CBM here is plant maintenance in a CMMS. Field service management handles service work at customer sites. Upvantix runs CMMS and Field Service Management as separate products.