PM and PdM yield: how much real work your inspections actually find

The maintenance team spends a great many hours each month on preventive and predictive rounds. The question nobody has put formally is the obvious one: how much does all that inspecting actually find? Is it catching real problems before they become failures, or is it mostly walking routes and ticking boxes?

PM and PdM yield puts a ratio on it. Track how much corrective work those inspections generate against the hours spent inspecting, and you start to see whether the programme is earning its keep.

What it actually measures

PM and PdM yield is the ratio of corrective repair hours identified through preventive and predictive inspections to the total hours spent performing those inspections. A yield of 0.5 means that for every hour of inspection, the team found half an hour of corrective work that genuinely needed doing before something failed. It is a measure of catch rate: how productive your looking is, in the plain currency of repair hours surfaced per hour spent looking.

How to work it out

PM and PdM yield = Corrective hours found by PM and PdM / PM and PdM hours

In a month, the plant spends 1,200 hours inspecting, and those inspections turn up 720 hours of corrective work needed before failures occur.

PM and PdM yield = 720 / 1,200 = 0.6 hours per hour

A little over half an hour of useful repair work found for every hour of looking. Whether that is the right level depends entirely on the plant's reliability and the volume of inspection behind it, which is why the number only means something in context. As a rough orientation, many mature programmes find their healthy mid-range sits somewhere around 0.8, but that is an anchor to reason from, not a target to chase.

The grid that gives it meaning

Yield is almost meaningless read on its own; its power comes from setting it against the reliability of the equipment, because the same number can be excellent or alarming depending on the company it keeps. Picture a simple grid, yield on one axis and reliability on the other, and six readings emerge.

When reliability is high and yield is moderate, you are in the healthy zone: the programme is finding enough to justify itself and the equipment is staying well. When reliability is high but yield is near zero, you may be over-inspecting, looking far more often than the equipment warrants, and there is room to extend intervals and free the hours for better use. When reliability is high but yield is unusually large, the inspections are turning up a surprising amount on supposedly healthy machines, which often points to a redesign opportunity rather than a maintenance one.

The low-reliability row is where the warnings live. Low reliability with near-zero yield means the inspections are missing the failure modes that matter, a candidate for a thorough review of the tasks themselves. Low reliability with moderate yield suggests the programme is catching things but not the right things, and may need both review and redesign. And low reliability with a very high yield is the infant-mortality signature: fault after fault is being found, often because intrusive work keeps reintroducing defects, and the answer lies in better practices and design rather than yet more looking.

Why context is everything

Two further subtleties keep yield honest. First, never read it on a single asset or a single inspection; it speaks only as an average across a large programme over time, where the law of averages can do its work. Second, a low yield is not automatically a failure. A predictive technique chasing a failure mode with a short warning period, where little time separates the first detectable sign from the failure itself, will legitimately inspect many times before it catches anything, and its yield will look low for entirely sound reasons. Equally, a yield that has fallen can simply mean you are now inspecting too infrequently to catch developing faults inside their window, since stretching an interval past its useful limit lets problems mature unseen between visits.

Where it can mislead you

• Read yield alongside reliability, always. Low yield plus poor reliability is the combination to worry about; the same low yield on a reliable plant may be perfectly fine, or even a sign you can ease off.
• High yield plus high reliability can be a sign of over-inspection. Intervals may be safely stretched, but never cut work that exists for safety or regulatory reasons, whatever the yield says.
• It is meaningless on a single asset or a single inspection. Yield only speaks as an average across a large programme over time.
• It measures what was found, not whether the work was truly warranted. It is easy to flatter yield by writing corrective orders for trivial findings, so pair it with effectiveness to tell genuine catches from busywork.

PM and PdM yield answers whether your inspection hours are buying you anything: whether all that looking is turning up real, actionable problems or simply filling the schedule. Read with a clear eye on reliability, and never on a single machine, it tells you when to trust your programme, when to expand it, and when to trim it back, which is most of what you want a maintenance metric to do.