What stops if it fails?
Nothing, a zone, a line, or the whole service. It's the heaviest-weighted factor and the easiest to answer.
Guide
Not every asset deserves the same maintenance
It seems obvious, and yet most plans treat a critical pump the same as an extractor fan in a warehouse. The result is overspending where it doesn't matter and underspending where it hurts.
What is asset criticality
It's a measure of how much it matters if that piece of equipment fails, built from two factors: the consequence of the failure and its probability. The consequence includes service downtime, risk to people, environmental impact, regulatory non-compliance and repair cost. Probability comes from history or, if there isn't any, from the team's experience. An asset with high consequence and high probability is the one that takes priority; one with low consequence can be allowed to fail even if it fails often, and that's also a maintenance decision.
What to look at to classify
Five questions are enough to sort an entire inventory.
Is there a risk to people or regulatory compliance?
If there is, criticality is high by definition and the plan isn't up for negotiation.
Is there redundancy?
A duplicated unit allows for a very different approach than a single one does. It's the factor most often forgotten.
How long does it take to replace?
A part with a twelve-week lead time turns a minor failure into a long shutdown.
How often does it actually fail?
If there's history, it's a data point. If not, it's the team's estimate, which is usually reasonable.
Doing it without turning it into a project
Formal criticality analysis methods are good and almost nobody finishes them, because they demand an effort that competes with the day-to-day work. What actually works is a pragmatic version: three levels — critical, important, the rest — a two-hour session with the people who know the installation, and the classification noted on each asset's record. It won't be perfect and it doesn't need to be: going from "treat everything the same" to "three groups" already completely changes how effort gets allocated. Refining it comes later, once history lets you discuss it with data instead of opinions.
What changes by level
For critical assets: a demanding preventive plan, spare parts in stock, a detailed inspection checklist and, if the cost of downtime justifies it, condition monitoring. For important assets: a reasonable plan and an identified but not necessarily stocked spare part. And for the rest, run-to-failure maintenance is a perfectly legitimate option: let it fail and repair or replace it. Saying that out loud tends to feel uncomfortable, but the alternative — maintaining everything the same way — in practice means the critical assets get less attention than they need, because effort is finite and it's being allocated badly.
Shall we look at it with your way of working?
Leave your details and we will get in touch to see whether we fit. No commitment, no lock-in period.
Criticality is not the same as urgency
They get mixed up every day and they're two independent axes. Criticality is a property of the asset and doesn't change: it describes what happens if that piece of equipment fails, how much the downtime costs, how many people it affects, and whether it has safety or regulatory implications. Urgency is a property of the specific incident and changes every time: it describes how long this particular thing that just happened can wait. A highly critical asset can have an incident that isn't pressing — a burnt-out indicator light on a machine that's still producing — and a low-criticality asset can have a genuinely urgent issue. When both things get mixed into a single priority field, what happens is that everyone marks things as high priority, because everyone weighs their own concern, and priority stops sorting anything. Separating them takes an afternoon and fixes half the problem.
The matrix people use, and the one they abandon
Almost everyone arrives at the same idea: score each asset by probability of failure and by consequence, multiply, and sort. It works, but only if you resist the urge to fine-tune it. Five-by-five matrices with weighted criteria produce beautiful numbers nobody can justify and nobody updates when the plant changes, which happens within six months. What actually survives is a three-level classification anyone on the team can assign in a minute and explain in a sentence. And a warning that saves a lot of work: the score isn't the goal. The goal is for the twenty percent of equipment that concentrates the risk to have a plan, and for the rest to stop competing for the same resources. That doesn't require decimal-point precision; it requires the list to exist and for someone to look at it once a year.
Frequently asked questions
How often should the criticality classification be reviewed?
Once a year is enough, and it should be a short review, not a redesign. What makes it go stale isn't the passage of time but changes to the installation: new equipment, lines that change use, processes that stop being a bottleneck. If the annual review turns into a project, that's a sign the classification was too fine-grained to maintain.
How do I prioritize if everything seems critical to me?
By forcing comparison instead of scoring in the abstract. If eighty percent of the equipment comes out critical, the classification isn't saying anything. A practical solution is to set a quota: only the ten or fifteen that genuinely stop something or carry a safety implication can be critical. Forcing a choice is uncomfortable, and that's exactly what makes the exercise useful.
How is an asset's criticality calculated?
With two factors: the consequence of its failure — downtime, risk, regulation, cost — and its probability. To start, three levels and a session with whoever knows the installation are enough; refining it comes later, with history.
Do I need a formal method?
Formal methods exist and they're good, but almost nobody finishes them because they compete with day-to-day work. Going from "treat everything the same" to three groups already completely changes how effort gets allocated, and that can be done in one afternoon.
Is it acceptable not to maintain a piece of equipment?
Yes, and it has a name: run-to-failure maintenance. For low-consequence, redundant, cheap-to-replace assets, it's the right call. Maintaining everything the same way in practice means neglecting the critical ones.
Where is that classification stored?
In the record of the asset and its location, so that an incident's priority comes from the system rather than depending on whoever receives it knowing what's in that room.
Classify a zone
In the demo we take one of your lines or locations and sort it into three levels, to see the effect on the plan.