In most small and mid-sized manufacturing companies, maintenance looks the same. A machine goes down. Someone runs for the mechanic. The mechanic tries to remember who last worked on that injection moulder. They find them. They fix it. An entry in a paper notebook — if anyone remembered. Next breakdown in a week. Repeat.
That isn't maintenance. That's firefighting in real time. The difference between the two is a CMMS — a Computerised Maintenance Management System. It sounds heavy, but in a 15-40 person company it boils down to four simple functions: a machine register, breakdown history, a preventive maintenance schedule, and a mobile app for the mechanic. Nothing more, to start.
Before you start comparing vendors, though, it's worth understanding what a CMMS really changes — and what it doesn't.
What a CMMS actually does
A CMMS doesn't fix machines. It fixes your knowledge of the machines. That's a subtle but crucial distinction.
Without a CMMS, the knowledge about your machine park lives in the heads of two or three people. When one of them retires or leaves for a competitor, half of the maintenance know-how walks out the door. A blank page is left behind. The next mechanic starts from scratch — including rediscovering that injection moulder number 7 has a tendency to overheat when the raw material is changed.
A CMMS turns that individual knowledge into a company asset. Every breakdown report, every inspection, every part replacement — all recorded under a specific machine, accessible to every mechanic, to the production manager, to the owner. After a year, patterns become visible. After two, you know which machine was simply a cheap purchase mistake that eats more manhours than it's worth.
In practice, it gives you three things:
- Problem history under each machine — who fixed it, what they did, which parts were used.
- Preventive maintenance schedule — automated reminders instead of relying on the mechanic's memory.
- Downtime statistics — at month-end you know exactly how many hours you actually lost and where.
A concrete example
One of our clients — a 35-person plastic packaging producer near Poznań, with 12 injection moulders plus 6 packaging lines — was losing 8-10 hours per week to unplanned downtime across the shop floor. Half a day of production per week gone. Nobody measured it precisely, because "breakdowns just happen".
We started with the basics. A scan of the machine park, a machine catalogue with inventory numbers, QR codes on the housings. A spare parts catalogue — not perfect, but sufficient to link a purchase order to a specific machine. A mobile app for mechanics instead of a paper notebook in the workshop. Integration with Comarch Optima — so parts orders flow through the normal purchasing process, not "Krzysztof will call his mate".
After six months, downtime dropped to about 3 hours per week. Six hours of recovered production time per week is roughly 250 hours per year — and at a machine-hour rate of around 180 PLN that's a return on investment any CEO understands. On top of that came something nobody had measured before: mechanics stopped ending their shift with a list of five things they needed to hand over verbally to the next shift. Because it's all in the system.
The cost? CMMS licences at this scale run about 20-40 PLN per user per month. The implementation with inventory, ERP integration, and mechanic training was a one-off cost comparable to six months of licences. Payback appeared in the fourth month.
Preventive maintenance — the core
The biggest value a CMMS brings isn't in recording breakdowns; it's in preventing them. A preventive inspection of an injection moulder every 500 operating hours costs 45 minutes of a mechanic's time. An emergency repair of the same machine after a cooling system failure — 6 hours, plus waiting for a part, plus a batch of raw material ruined.
The problem is, without a system, nobody remembers that moulder 4 has just hit 500 operating hours since the last inspection. "You were supposed to check last week, weren't you?" Yes, they were. But they forgot. It was Friday.
A CMMS automatically issues a work order when an operating-hour counter crosses the threshold. The mechanic gets a task on the tablet, does it, marks it as done, and a thermal image ends up in the machine's card. Three months later, when the same moulder does go down, you open the history and see that the last inspection was completed but a slight filter contamination was reported. Diagnosis in 5 minutes instead of an hour.
This only works if the schedule is realistic. Entering 40 inspections per week into a system that supports two mechanics guarantees that within two weeks nobody will click anything anymore. Start with the 5-8 most critical machines and realistic frequencies. Expand later.
What a CMMS won't fix
A few things worth saying plainly, to avoid disappointment.
A CMMS won't replace a competent mechanic. If the company has one maintenance technician who barely keeps up, adding tablet forms to their duties won't magically free up time for preventive work. Hire a second one first. Then deploy CMMS.
A CMMS won't tidy up your spare parts warehouse. It will create a register, but the shelves still need to be physically sorted, labelled, and stock counts reconciled. That's a separate project, often preceding the CMMS itself.
A CMMS won't work if the mechanics haven't bought in. "They forced this system on us from above" is a guarantee that within a month everyone clicks "done" without entering any details. Talking to mechanics before choosing the tool, and testing a few options with them, turns "they forced it" into "we chose it". Three hours of workshops save six months of frustration.
Where to start
If you don't measure downtime at all, the first two weeks are a simple list of machines and tallying on paper how many times each one went down. No system, no licences. The count in the second week is often instructive enough to speed up the decision.
Next — tool selection. On the Polish market, solutions like CMMS Machine, Fabryka 4.0, Simfy work well, as do Fiix or Limble in versions with Polish support. Key criteria: a mobile app for mechanics in Polish, QR codes on machines, a sensible API for ERP integration, reasonable pricing at 3-6 users.
An implementation makes sense as an iterative rollout. Month one — machine inventory, QR codes, entering the top-10 most frequent breakdowns from the mechanics' memory. Month two — mobile app, everyone reports breakdowns in the system instead of via WhatsApp. Month three — a preventive schedule added for the 5-8 most critical machines. Month four — first reports, decisions based on data. By year-end the system has full historical data, and the CAPEX discussion about replacing old machines rests on numbers, not anecdotes.
When to talk
If you have a dozen or so machines, don't measure downtime in hours, and every month suddenly some critical part is missing, then a CMMS will pay for itself. The only question is how to implement it well — because a badly deployed one becomes just another system the mechanics work around.
The first call is free and takes about 30 minutes. What comes out of it is either a straightforward tool recommendation and a list of things to sort out before deployment, or an honest signal that it's not time yet — because there are more important things to do first.
A CMMS doesn't fix machines — it fixes your knowledge of the machines. A subtle difference that turns firefighting into normal maintenance after about a year.