Reactive Maintenance — Run to Failure

What it is

Reactive maintenance — also called run-to-failure or corrective maintenance — is the deliberate or default approach of operating equipment until it fails and then repairing it. No scheduled maintenance is performed in advance; resources are deployed only after a failure event.

When reactive is appropriate

• Low failure consequence: Equipment whose failure creates minimal operational disruption, no safety risk, no regulatory exposure, and no significant financial loss — the cost of a reactive repair is simply the repair cost, without the downtime cost amplifier that high-consequence equipment adds.
• Easily and cheaply replaced: Equipment whose replacement cost is low enough that carrying it as a spare (or ordering overnight) is less expensive than maintaining a PM program that costs more than the equipment itself.
• No identifiable failure precursor: Equipment that fails without detectable precursor symptoms — where PM inspection cannot catch a developing failure because the failure mode is instantaneous (electronic component failure) rather than progressive.
• Non-redundant low-frequency use: Equipment used infrequently enough that failure probability during use is very low and the cost of PM exceeds the probability-weighted failure cost.

When reactive is not appropriate

Equipment with safety consequences, regulatory compliance implications, revenue impact, or secondary damage potential should not be managed reactively — the cost of a single failure event may exceed years of PM investment.

Preventive Maintenance — Scheduled Care

What it is

Preventive maintenance is scheduled maintenance performed at defined intervals — calendar time, usage accumulation, or periodic condition checks — regardless of current equipment condition. It is the dominant strategy for most commercial equipment because it is systematic, predictable, and effective for equipment with wear-based failure modes.

When preventive is most appropriate

• Equipment with wear-based failure modes: Most mechanical, electrical, and HVAC equipment deteriorates predictably with use — bearing wear, belt fatigue, filter loading, coil fouling. PM addresses these deterioration modes before they reach failure threshold.
• Defined manufacturer maintenance specifications: Equipment with manufacturer-specified maintenance intervals and procedures has the PM program designed by the people who built it — following these specifications is the baseline appropriate strategy for most commercial equipment.
• Medium-to-high consequence with cost-effective PM: Equipment whose failure consequence is significant enough to justify PM investment but not significant enough to justify continuous monitoring — the middle tier of the criticality distribution where PM delivers the highest marginal return.
• Compliance requirements: Many regulatory maintenance requirements are expressed as PM requirements — fire suppression inspections, elevator certifications, generator testing. These are mandatory PM regardless of condition-based justification.

Predictive Maintenance — Condition-Based Service

What it is

Predictive maintenance uses sensor data, vibration analysis, oil analysis, thermal imaging, or other condition indicators to determine when equipment actually needs maintenance — based on its measured condition rather than elapsed time or usage. Service is performed when condition data indicates that a developing failure is approaching the threshold where PM will be effective.

When predictive is most appropriate

• High failure consequence: Equipment whose failure creates significant safety, revenue, or secondary damage consequences — where even a single prevented failure justifies substantial monitoring investment.
• Sensor-accessible failure precursors: Equipment with failure modes that produce detectable sensor signals before reaching critical failure — temperature rise, vibration increase, current draw change, pressure deviation. Not all failure modes have detectable precursors.
• Variable utilization intensity: Equipment whose wear rate varies significantly — highly efficient in some periods, heavily stressed in others — where calendar-based PM either over-maintains (during light-use periods) or under-maintains (during high-stress periods). Condition-based triggers automatically adjust to actual equipment state.
• IoT infrastructure already in place: Adding predictive maintenance to equipment that already has sensor coverage from IoT deployments for operational monitoring requires minimal additional investment — the sensor infrastructure exists, and alert threshold configuration produces predictive maintenance triggers.

The Strategy Allocation Matrix

Frequently Asked Questions