A 100% PM completion rate looks good on a scorecard, but it does not prove a preventive maintenance program works. A maintenance department can close every scheduled PM and still watch the same failures recur if the underlying tasks are poorly written, scheduled too often, completed late, or unconnected to real failure modes.
Steve Janow, senior reliability consultant at MVP One, teaches maintenance teams in MVP 306: PM Development & Optimization, to look past compliance percentages and examine four signals that indicate whether a PM program protects equipment.
The first signal is the early discovery of corrective work. A useful inspection catches deterioration while a repair can still be planned, not after a failure stops the line. Technicians should find early evidence such as a bearing beginning to run rough, a VFD cooling fan packed with dust, an oil leak at a gearbox seal, or vibration readings that exceed baseline.
Janow cites a benchmark of roughly one corrective action for every six PMs performed. He notes the ratio will not fit every facility or PM type, but the underlying principle holds: Inspections should generate actionable findings. When hundreds of condition checks produce no corrective work, the team should question whether the instructions are specific and followed correctly.
The second signal tests whether a task is written precisely enough to repeat. Instructions such as "inspect machine" or "grease all bearings" leave too much room for interpretation. A strong PM names the component, location, condition, acceptable range, required tool, material, and safety precaution, and it tells the technician how to respond to an abnormal reading.
Rather than asking a technician to check vibration, the procedure should specify the measurement points, the equipment's operating state, the sensor or route used, the expected baseline, and the escalation threshold. Janow describes the goal as delivering "a detailed, step-by-step procedure," so results stay consistent no matter who performs the work.
The third signal measures whether frequency matches failure behavior. A weekly inspection producing no findings for a year signals a candidate for a longer interval, while a monthly check that keeps uncovering advanced deterioration may need to run more often or move to continuous condition monitoring.
Overmaintenance carries its own risk: unnecessary disassembly can disturb alignments, damage seals, introduce contamination, or trigger an early-life failure after reassembly. Optimizing a PM program does not mean adding tasks. It means matching the right work to the right interval.
The fourth signal asks whether the task can be executed as planned. Even a well-written PM fails when a technician arrives without the correct filter, belt, lubricant, gasket, or bearing. Reliable execution depends on inventory discipline: spare parts lists attached to assets, parts tied to work orders, and storeroom balances that reflect actual usage.
Min/max thresholds should reflect lead time, criticality, consumption, and failure consequences rather than guesswork. Cycle counting exposes discrepancies before an outage and can reveal workbench hoarding, duplicate SKUs, mislabeled bins, and parts issued without a transaction, each one a warning that a future PM or repair could stall.
Register for MVP 306: PM Development & Optimization.
