Rating asset criticality: why it changes your maintenance schedule

Two identical air handling units, same make, same age, installed by the same contractor, can genuinely warrant different levels of maintenance attention — if one serves a plant room and the other serves an occupied space with vulnerable users, the consequence of either failing is not the same, even though the asset itself is. That's the basic case for rating criticality separately from just cataloguing what equipment exists.

What criticality actually measures

Criticality isn't a measure of how expensive or complex an asset is — it's a measure of consequence. What happens if this specific asset fails right now: does it stop a business-critical process, create a safety risk, mildly inconvenience someone, or go entirely unnoticed until the next scheduled check? Those are four very different answers, and a maintenance schedule that treats them the same is quietly under-serving the ones that matter most.

Building a simple, honest scale

A criticality scale doesn't need to be elaborate to be useful — in fact an overengineered scoring model that nobody applies consistently is worse than a simple one that actually gets used. A workable starting point is three or four bands: safety-critical, operationally critical, standard, and low-impact — applied consistently across the asset register rather than left to individual judgement each time someone happens to look at a specific asset.

• –Safety-critical: failure creates a direct risk to people — fire doors, emergency lighting, lifting equipment
• –Operationally critical: failure stops or seriously disrupts core building function — primary heating plant, water supply, main electrical distribution
• –Standard: failure is an inconvenience that needs fixing but doesn't create immediate risk or major disruption
• –Low-impact: failure is barely noticed and can reasonably wait for the next convenient window

How criticality should actually change the schedule

Once an asset carries a criticality rating, it should visibly change two things: how often it's checked, and how it's prioritised when something does go wrong. A safety-critical asset flagged as overdue shouldn't sit in the same queue as a low-impact one — it needs to surface immediately, not get buried in a general list sorted by date. This is really the same principle behind balancing planned and reactive work: the goal isn't treating everything equally, it's spending attention where the consequence of getting it wrong is highest.

The trap of rating everything as critical

There's a natural instinct, especially early on, to rate generously — better safe than sorry. But a register where 80% of assets are marked critical has, in effect, no criticality rating at all, because it no longer distinguishes anything. A useful scale has to be honest enough that most assets land in the middle, with only a genuine minority at either extreme — that's what makes the rating worth anything when a schedule is under pressure and choices have to be made about where limited hours go.