Every laser stops. The question that separates a profitable shop from a struggling one is not whether it stops, but whether the stop was on the calendar.
A planned stop costs you the maintenance hour. An unplanned stop costs you the maintenance hour plus everything that was queued behind it: the operator standing idle, the batch that has to be restarted, the parts half-cut and now scrap, the delivery promise that will be broken, and — in the worst case — the expedited freight or the discount offered to keep the customer.
That multiplier is not a small correction. It is usually the difference between maintenance being an expense and maintenance being the cheapest insurance in the building. And yet most shops treat the two identically: they fix things when they break.
This article is deliberately not a cleaning procedure. Two earlier articles on this site cover the routine — how to clean optics, what to check daily. This one covers the part that decides whether those procedures happen at all: when, by whom, at what cost when skipped, and what to keep on the shelf so that a stoppage is measured in hours rather than weeks.
Almost every failure in a laser shop announces itself first. The trouble is that nobody is recording the announcement.
The pattern across the table is that the machine degrades gradually and fails suddenly. Output drops a little each week; the operator compensates by nudging the power up or adding a pass; and the compensation hides the trend until the machine can no longer be pushed far enough.
That is why a maintenance schedule without a record is close to useless: it fixes things on a calendar while the actual wear is running on a completely different clock.
The default approach — clean on Monday, service in the spring — is inherited from machines whose wear is proportional to time. A laser's wear is not. It is proportional to beam hours, material volume and fume load, and those vary enormously between shops running the same model.
Three counters are worth keeping, because between them they predict almost every consumable:
Beam hours. The honest measure of tube life and of everything that wears with running time. A machine on two shifts accumulates this far faster than a one-shift machine, and a shop that services "every six months" is servicing a two-shift machine at half the interval it needs.
Material volume and fume load. This, not hours, drives filter life, optics contamination and extraction performance. Cutting foam, coated board or PVC-adjacent materials loads the system far faster than cutting acrylic sheet, and a schedule based on hours alone will under-service exactly the shops that need it most.
Cleaning and replacement events. Every lens cleaning, nozzle change and filter swap is a data point. The interval between them, and the reason for each, tells you what is actually wearing in your shop rather than in the manual's example.
The practical rule: take the manufacturer's interval as a starting ceiling, then let your own records pull it shorter. If a lens needs cleaning twice as often as the schedule says, the schedule is wrong — not the lens.
You cannot decide how much maintenance is worth doing until you know what a stopped machine costs. The good news is that the arithmetic is the same as the cost-per-part structure, read backwards.
Two things usually surprise shops when they do this for the first time.
The idle labour is often larger than the lost output. A machine stop in a shop with four people working around it stops four people's work, not one — and in small shops the finishing bench is nearly always starved by a cutting stoppage.
The rework line is frequently the biggest single item. A tube that fails mid-batch does not just lose the hour; it loses the material already cut into that batch, and the labour already spent on it.
Once that number exists, the stocking decision becomes almost mechanical: if an hour of downtime costs more than the part, and the part has a delivery lead time longer than you can tolerate, keep one on the shelf. Everything else follows from that rule, which is why the next section is short.
Spare-part stocking is normally decided by anxiety rather than arithmetic. The rule above replaces anxiety with a comparison.
The two errors are symmetric and both common. Shops stock nothing, then lose a week waiting for a nozzle that costs less than a lunch. Shops stock everything, tie up cash in a tube that ages on the shelf and is obsolete by the time it is needed.
The middle position — deep on cheap consumables, one of each critical item, a plan for the expensive ones — is almost always right, and it is worth writing down rather than leaving to whoever orders supplies.
One addition specific to lasers: ageing matters. Some spares degrade in storage — seals harden, optics can be damaged by careless handling on a shelf, coolant goes stale. A spare that has been sitting in a damp cupboard for three years is not a spare; it is a delay.
The most common maintenance failure in small shops is not a missed task — it is a task everyone believes someone else performs.
Two disciplines make this work. The first is that the operator must be allowed to stop the machine for maintenance — a shop where stopping is treated as laziness will run the machine into failure every time. The second is that the person who orders spares is named, because "we'll order it when we need it" is how a two-day repair becomes a two-week shutdown.
If a shop adopts one thing from this article, it should be the log.
The log is not a compliance document. It is a single sheet, per machine, recording: date, hours, what was done, and — most importantly — any deviation the operator noticed. "Needed two passes instead of one on the acrylic job." "Edge looked rougher on the right side." "Water temperature higher than usual."
Those notes are the entire early-warning system. Without them, a gradual decline is invisible until it becomes a failure, and the failure looks sudden when it was in fact predictable for weeks. With them, a pattern appears — and patterns are what turn maintenance from reaction into scheduling.
The practical format that survives contact with a busy shop is deliberately small: a few lines a day, written at the machine, reviewed weekly by whoever leads the shift. Anything more elaborate gets abandoned in a fortnight.
A maintenance log fails if it is written at the end of the week from memory. It works if it is written at handover, when the outgoing operator tells the incoming one what changed.
Three questions at handover cover nearly everything:
What did the machine do that it did not do yesterday? This catches degradation early, when it is still a setting rather than a fault.
What did you compensate for? Operators quietly increase power or slow the speed to keep output up. That compensation is the signal, and it is lost unless it is spoken.
What is running low? Consumables, material, gas, filters — better discovered at handover than mid-batch.
Shops that do this well rarely have dramatic failures. Not because their machines are better, but because nothing gets to fail unannounced.
When a component reaches the end of its life, the question is always the same and the answer is always in the records.
A tube that has lost capability gradually, with the log showing a long slow decline, has simply worn out — replace it, and adjust the interval that predicted it. A tube that failed early, with no preceding decline, is a different problem: cooling, power supply or water quality, and replacing it without finding the cause buys a second failure.
A lens that is damaged repeatedly is not a lens problem. It is a nozzle, alignment, spatter or assist-air problem upstream. The most expensive maintenance habit in the industry is replacing the same part on a schedule without asking why it keeps failing.
A chiller or cooling fault that appears only in summer is a capacity problem, not a failure — and it will recur every year until the duty cycle or the ambient conditions are addressed.
The log turns these from judgement calls into diagnoses, which is the whole point of keeping it.
Maintenance is not a cost to be minimised; it is a schedule to be designed. The tools are unglamorous — three counters, a downtime figure, a stocking rule, a named owner and a small log — but between them they convert a machine's wear from an invisible drift into a predictable line.
The shop that knows what an hour of stoppage costs will always out-invest the shop that does not, because it can see the return. And the shop that writes down what the machine did differently today will always be surprised less often than the one that trusts its memory.
None of this requires new equipment. It requires deciding, once, that a stop on the calendar is worth ten stops in the dark.
Hot News2026-11-16
2026-11-12
2026-11-10
2026-11-09
2026-11-06
2026-11-05