PM and PdM work orders overdue: a queue of deferred risk, sorted by what matters

End of the month, and the maintenance manager pulls the open work order list. Dozens of preventive and predictive tasks are sitting there past their due date. Some are a week late. Some are three months late. A handful are on assets that would stop the whole plant if they failed.

Without a structured way to count and sort that pile, it is easy to wave it off as "just a few slipped jobs". This metric turns the pile into a table, ranked by how late each task is and how critical each asset is.

What it actually measures

PM and PdM work orders overdue counts every active, still-open preventive and predictive work order that has passed its due date. The orders are grouped into age bands, commonly up to 30 days overdue, 30 to 90, and more than 90, and then mapped against asset criticality. For runtime-based tasks the same logic runs in hours: up to 25% past the interval, 25 to 50% past, and more than 50%. The result is a snapshot of deferred care, stratified by urgency, so that an intimidating list becomes a short list of things that genuinely demand attention.

How to work it out

Days overdue = Current date - Due date

Count every open PM and PdM order, band it by how overdue it is, and lay it against criticality rank. A month-end snapshot might total around ninety overdue orders, which sounds alarming until you read the structure: most are low-consequence jobs only a little late, while the part that matters is far narrower, perhaps a dozen tasks on the most critical assets and already more than ninety days overdue. Those few get a recovery plan before anything else on the list is touched.

Why band it at all

The instinct with an overdue pile is to work it oldest-first, or simply to panic at the headline count. Both miss the point. A hundred overdue lubrication routes on standby equipment are a housekeeping problem; a single overdue protective-system test on a critical asset is a safety and reliability problem. Banding the list by age and criticality together lets you triage like an emergency room rather than a queue, clearing the dangerous end first. And the trend matters more than any single reading: a rising count in the high-criticality, over-ninety-day corner is one of the more reliable early signals that failures are coming, because that is deferred risk maturing in plain sight.

How it differs from compliance

It is easy to muddle this with the compliance measures, but they look at different moments. Compliance scores a period after the fact: of the tasks that came due, how many were done on time? Overdue work orders are a live snapshot of what is open and late right now, today, this minute. Compliance tells you how the month went; the overdue count tells you what is currently at risk and waiting. A programme can post respectable compliance over a quarter and still be carrying a dangerous overdue pile at any given moment, which is why the two belong side by side on the same dashboard rather than standing in for each other.

What good looks like

The target is fewer than five per cent of all work orders overdue, and the ideal is zero. Zero is rarely sustained in a real plant, so the practical discipline is twofold: keep the overall figure low, and make sure that whatever is overdue is concentrated in the low-consequence bands, not the critical ones. Any task more than thirty days overdue, or past a quarter of its interval by runtime, should carry a documented plan for resolution rather than simply rolling silently into next month.

Where it can mislead you

• Use a Pareto or criticality analysis to prioritise. Clear the high-criticality, long-overdue tasks first, not simply the oldest items on the list.
• The count moves continuously, almost by the minute, so treat it as a snapshot and trend it. A single high reading matters far less than a rising pattern.
• Any task more than 30 days overdue, or past 25% of its interval by hours, should carry a documented plan for resolution.
• A large pile in the 90-day-plus band usually means those tasks are being deferred indefinitely. Decide honestly whether they can still be done as planned, or whether the task itself needs re-engineering or retiring.

Overdue PM and PdM work orders are best understood not as a backlog but as a queue of deferred risk. Counting them is easy; the value is in sorting them by consequence, so your recovery effort goes to the dangerous end of the queue first, where a slipped task is most likely to become the next breakdown.