Hierarchy and criticality
Plant, line, system and machine, with each asset’s criticality defined. Without that, everything is equally urgent, and so nothing is.
Guide
In a plant, maintenance is measured in lost production
That’s the translation that changes the conversation. As long as maintenance is discussed in technician-hours, it’ll be a cost center; once it’s discussed in line-down hours, it becomes a business decision.
What sets industrial maintenance apart
Three things, and none of them is the technical complexity of the equipment. The first is that downtime has a direct, calculable cost, which lets you justify investments that can’t be defended in other settings. The second is that maintenance competes with production for the same resource — machine time — and that negotiation happens daily. And the third is that much of the knowledge lives in the heads of veteran technicians, in an industry where generational handover has been a problem for years, not a forecast.
What a well-run operation rests on
Five things present in every plant where maintenance goes well.
A plan based on history
Not on the manufacturer’s catalog, which is written for the worst case and to avoid liability. The plan gets tuned to what actually fails in your facility.
Annual shutdown forecasting
What lets you walk into the scheduling meeting with a proposal instead of a request, and bundle everything due on that line into the same window.
Critical spares identified
Not stocking everything: stocking what stops the line. The difference between the two comes from consumption history.
Indicators that actually get looked at
MTBF, MTTR, plan compliance and cost per asset. Four numbers that fit on one page and that almost no plant has reliably.
The problem nobody puts on the agenda
In an average plant, the person who knows why that pump fails every summer is usually one specific individual, and they know it because they’ve been there fifteen years. That knowledge is written down nowhere and disappears the day that person is off, retires, or changes jobs. It’s a first-order operational risk and it appears on no dashboard. The only thing that mitigates it is having the history accumulate outside people’s heads: every breakdown with its cause, every inspection with its readings, every fix with what was actually done. It doesn’t replace experience, but it keeps the next person from having to learn it all from scratch through downtime.
The negotiation with production
It’s the structural conflict of any plant: maintenance needs to stop, production needs to run. It’s won with foresight, not arguments. Walking into the scheduling meeting with the annual plan already calculated — which inspections fall in which month, on which equipment, and how many hours they need — turns a request into a proposal, and lets you bundle everything affecting that line into a single window instead of stopping it three times. Just as important is recording what couldn’t be done: a plan that’s met seventy percent of the time, and you know it, is useful information; a plan that only shows what was executed isn’t information, it’s a flattering photo.
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.
What the indicators actually mean
MTBF, mean time between failures, tells you how long a piece of equipment lasts before failing again; MTTR tells you how long it takes to get it back into service. The first speaks to the equipment and the plan, the second to the organization: a high MTTR is almost never a skill problem, it’s usually time spent waiting for a spare part, a permit, or a person. Neither means anything if work orders are closed without real times or without being charged to the correct asset, which is why so many plants have indicators and no decision ever made from them.
The critical spare part, which isn’t the expensive one
It’s one of the confusions that ties up the most money in a plant. The spare part that needs to be in stock isn’t the most expensive one or the one used most: it’s the one that stops the line and takes a long time to arrive. A twenty-euro bearing with a twelve-week lead time for a machine with no redundancy is infinitely more critical than a three-thousand-euro part the supplier delivers in twenty-four hours. Organizing the warehouse by that criterion — downtime consequence of the replenishment lead time, not price — usually achieves two things at once: reducing total tied-up capital and eliminating the waits that inflate mean repair time today. And the data needed to do it, real consumption per piece of equipment, comes from the work order history.
Frequently asked questions
How much corrective maintenance is normal in a plant?
There’s no universal figure, and be wary of anyone who gives you one: it depends on how old the machinery is, the work regime, and the sector. What is comparable is your own plant against itself over successive quarters. If the share of urgent corrective work goes down and plan compliance goes up, you’re on the right track, even if the absolute numbers don’t look like anyone else’s.
Where do you start in a plant with no inventory?
With the most critical line or system, with the bare minimum per piece of equipment: identification, location and model. With that, real work orders are already flowing within weeks. Trying to inventory the whole plant before starting is the fastest route to an abandoned inventory.
Which indicators should I look at?
Four are enough to start: MTBF and MTTR per critical piece of equipment, preventive plan compliance, and accumulated cost per asset. The hard part isn’t calculating them, it’s making them reliable, and that depends on how work orders get closed.
How do you negotiate shutdowns with production?
With the annual forecast calculated in advance: which inspections fall in which month, on which equipment, and how many hours they need. It lets you bring a proposal and bundle everything affecting that line into one window.
Is predictive maintenance worth it?
On critical equipment with prior history, yes. Before that, no: without well-recorded breakdowns there’s no pattern to detect, and the instrumentation ends up feeding a dashboard nobody trusts.
How do you avoid losing veteran technicians’ knowledge?
By having the history accumulate outside people’s heads: every breakdown with its cause, every inspection with its readings, and every fix with what was actually done. It doesn’t replace experience, but it keeps the next person from having to learn it through downtime.
Start with one line
In the demo we build its hierarchy and a couple of real preventive plans, so you see results in a few weeks, not a year.