The Problem with Calendar-Based Maintenance
Calendar-based maintenance operates on an assumption: if equipment has run for X days or X hours, it needs service. The interval is set based on manufacturer recommendations, historical failure data, or — in many operations — inherited practice from whoever set up the maintenance schedule years ago.
The assumption is wrong in both directions simultaneously.
Over-maintained equipment: A machine running in cool, clean, low-load conditions accumulates wear slowly. Its lubricants stay clean longer than average. Its bearings run well past the generic recommended replacement interval. A calendar-based schedule services it at the interval regardless — paying labor costs and exposing the machine to the risk of improper reassembly for maintenance it didn’t need.
Under-maintained equipment: A machine running heavy loads in harsh conditions, with frequent starts and stops, accumulates wear faster than the generic interval assumes. The calendar says it has 3 more months before service. The bearing says it needs attention now. The calendar wins — and the bearing fails 3 months early.
Research from the Plant Engineering industry consistently shows that approximately 82% of equipment failures are random — they don’t follow the age-based deterioration curves that fixed intervals are designed to catch. Calendar maintenance is well-matched to the 18% of failures that are age-related (wear-out failures where replacing a component at a defined interval prevents the failure). It is poorly matched to the 82% that are random and require condition monitoring to detect.
What Condition Monitoring Measures
Condition-based maintenance requires measuring operating conditions that indicate developing failure — not waiting for the failure to produce a visible symptom.
Vibration Analysis
Vibration is the most information-rich indicator of mechanical health. A bearing that is developing a defect produces vibration at specific frequencies determined by the bearing’s geometry. A gear mesh developing wear produces vibration at the gear mesh frequency. An imbalanced rotor produces vibration at the rotation frequency (1x RPM).
VX-Olympus connects to vibration sensors that measure:
- RMS vibration level: The overall vibration amplitude, in G or mm/s. Rising RMS indicates worsening condition.
- Frequency spectrum (FFT): The distribution of vibration energy across frequencies. Specific peaks in the spectrum identify specific failure modes (bearing outer race defect, inner race defect, rolling element defect, gear mesh).
- Time waveform: The raw vibration signal over time. Impacting and clipping patterns indicate specific failure types.
For most manufacturing applications, RMS vibration level with alert thresholds is sufficient for early warning. Full frequency analysis provides more specific failure mode identification for critical equipment.
VX-Olympus alert pattern:
- Watch alert when RMS vibration rises above 25% over baseline (establishes trend awareness)
- Warning alert when RMS vibration rises above 50% over baseline (schedule investigation within 2 weeks)
- Critical alert when RMS vibration reaches 2x baseline (schedule maintenance within 48 hours)
Temperature Monitoring
Temperature rises in rotating equipment are secondary indicators of mechanical problems. Bearing failure increases friction, which generates heat. Motor winding failure shows as abnormal temperature distribution. Gearbox lubricant breakdown shows as elevated gearbox case temperature.
Temperature monitoring is lower resolution than vibration analysis — it detects later-stage failure development than vibration, but requires simpler and less expensive sensors.
VX-Olympus alert pattern:
- Watch: bearing temperature rises more than 8°F above established baseline for that bearing position
- Warning: bearing temperature rises more than 20°F above baseline
- Critical: bearing temperature exceeds 200°F (absolute threshold for standard bearing materials)
Current and Power Monitoring
Electric motor current is a mechanical condition indicator. A motor driving a deteriorating pump shows increased current draw as the pump requires more torque. A bearing developing a failure increases the drag on the motor, raising current slightly. Motor winding failures show as current imbalance between phases.
Current monitoring is less specific than vibration but simpler to install — a current transformer around each motor supply conductor, without contact with the machine itself.
Operating Hours
Not all PM triggers should be condition-based. Some maintenance actions are lubrication, filter replacement, or inspection items where the correct interval is based on actual running time rather than a condition that can be measured directly.
VX-Olympus computes operating hours from current or status sensors: if motor current is above a threshold, the machine is running and hours are accumulating. The operating hours counter drives PM work order generation at configured thresholds.
The Transition From Calendar to Condition-Based
The transition is not binary — it is a progression of increasing condition-awareness applied to different maintenance actions.
Step 1: Baseline Current Practice
Before changing the maintenance schedule, characterize current equipment health. Deploy VX-Olympus vibration and temperature monitoring on the target equipment. Run for 60–90 days without changing the maintenance schedule. This builds the baseline that future condition thresholds are based on — what “normal” looks like for each specific machine in its actual operating environment.
This step prevents the common error of setting thresholds based on industry guidelines rather than the specific machine’s actual baseline. A machine that runs with inherently higher vibration than industry norms (due to its mounting, process load, or design) would trigger constant false alarms against a generic threshold. Against its own baseline, it triggers only when something actually changes.
Step 2: Add Operating Hour Tracking
Replace the calendar interval for time-based PMs (lubrication, filter changes) with operating hour triggers. Configure VX-Olympus operating hour counters on equipment where the relationship between hours and maintenance need is clearer than the relationship to calendar time.
This immediately corrects the over/under maintenance problem for time-based tasks: a machine that ran 60% of expected hours gets its oil changed proportionally. A machine that ran 140% gets its oil changed earlier.
Step 3: Add Condition Triggers for Failure Modes
For the failure modes that vibration and temperature monitoring can detect — bearing wear, imbalance, misalignment, overheating — add condition-based work order triggers to the VX-Olympus rule chains.
The rule chains generate work orders when conditions reach the warning level. The work order includes the current sensor readings, the baseline comparison, and the trend over the past 30 days. The maintenance technician has context before they examine the machine.
Step 4: Calibrate Based on History
After 6 months of operation with integrated work orders, review the maintenance history:
- Which condition-triggered work orders found real problems vs. false positives?
- Which calendar-based PMs found problems vs. found nothing?
- Which thresholds were too sensitive (too many work orders generated) or too conservative (problems developed further than expected before triggering)?
Adjust thresholds based on the actual history. A threshold calibration based on 6 months of real maintenance data is more accurate than a threshold set from manufacturer guidelines.
Real-World Results Pattern
Manufacturing operations that have transitioned from calendar to condition-based maintenance with VX-Olympus consistently report a similar progression:
First 90 days: The baseline period. Some surprises in the initial monitoring data — machines that have been running with elevated vibration that was never measured before. Condition-triggered work orders begin generating alongside calendar-based PMs.
Months 4–6: Calibration period. Some false positive work orders. Technicians build confidence in interpreting vibration data and temperature trends. Some calendar-based PMs get extended based on condition data showing healthy equipment.
Months 7–12: The productive period. The maintenance schedule starts to reflect actual machine conditions. Calendar-based maintenance continues for the maintenance items that don’t have measurable condition indicators; condition-based triggers handle the failure modes that can be monitored.
What Condition-Based Maintenance Doesn’t Replace
Condition monitoring is not a substitute for all maintenance activities:
Inspection-based maintenance: Some failure modes have no sensor signature until late-stage failure. Visual inspection — looking for cracks, corrosion, seal wear, fastener loosening — remains a necessary maintenance practice for equipment components that sensors don’t measure directly.
Lubrication: Lubricant quality degrades based on operating hours, temperature exposure, and contamination — not necessarily on vibration signatures. Lubrication intervals should be operating-hour based, not purely condition triggered.
Regulatory-required PM: Some industries have regulatory requirements for maintenance at specific intervals regardless of condition. These intervals cannot be extended based on condition data alone.
A mature condition-based maintenance program uses sensor data for the failure modes that sensors can detect, operating hours for the maintenance actions tied to usage cycles, and calendar intervals only for the maintenance actions that regulatory or warranty requirements mandate.
Conclusion
Calendar-based maintenance is a compromise — a practical response to the unavailability of real-time condition data. When you don’t know what a machine’s actual condition is, servicing it at manufacturer-recommended intervals is reasonable.
When you do know the actual condition — through vibration analysis, temperature monitoring, and operating hour tracking — the compromise is no longer necessary. Maintenance can be scheduled when the equipment actually needs it.
The transition requires a period of baseline measurement, threshold calibration, and organization adjustment. Operations that have made the transition consistently report the same outcome: less maintenance labor spent on equipment that didn’t need attention, fewer surprise failures from equipment that needed attention earlier than the calendar indicated.
VX-Olympus provides the monitoring layer, the analytics that convert raw sensor data into maintenance triggers, and the integrated work order system that closes the loop between detection and response.
Talk to our team about a condition-based maintenance deployment for your manufacturing operation.