Renting or Buying a Scrubber-Dryer: Who Carries Which Risk

A scrubber-dryer performs three work steps in a single pass. It applies water and detergent to the floor, breaks the dirt loose mechanically with a rotating roller or disc brush, then lifts the dirty solution with a rubber squeegee and draws it into the waste water tank. What remains behind the machine is a walkable floor, not mop water drying in place. That is the whole point of the class: the dirt is removed rather than redistributed. In its manually guided form — a machine like the SH1, pushed by one person — the principle is over half a century old and unchanged.

Everything else in the class scales with two figures. The first is working width, which decides how much floor one pass covers: 440 mm on the guided SH1, 450 mm on the compact L4, 510 mm on the industrial L50. The second is tank volume, which decides how long the machine runs before someone has to intervene: 4 litres of fresh water on the SH1, 15 on the CC1, 38 on the L4, 55 on the L50. A small tank on a large area does not mean slow cleaning, it means frequent refill stops — which is a staffing question, not a machine question.

And the boundaries matter as much as the capability. A scrubber-dryer is a hard-floor machine. For the CC1 our product data names the approved surfaces explicitly — terrazzo, marble, tiles, epoxy resin, sandstone and artificial stone — with short-pile carpet handled in vacuum mode, dry. Carpet, deep textile and heavily structured surfaces are a different machine class, and no acquisition model fixes a wrong class. Which class fits which floor is the question our scrubber-dryer guide answers before this article becomes relevant at all.

The guided machine is the reference case. The SH1 is a cordless walk-behind scrubber operated by one person: 440 mm brush width, 490 mm squeegee width, 27 kg brush pressure at 350 RPM and 20,000 Pa suction, with 4-litre fresh and waste water tanks and a manufacturer-stated productivity of 1,100–1,600 m²/h. Its battery runs 70 minutes in standard mode and 100 minutes in ECO and charges again in about 2.5 hours. Every one of those figures describes the machine — none of them describes the shift. The guided machine works exactly as long as someone is pushing it, and that person's time is the largest single input in the whole calculation.

Take the operator out and nothing about the cleaning principle changes; a navigation layer is added. The CC1 orients itself with PUDU SLAM, combining visual and laser SLAM, and senses its surroundings with LiDAR, RGBD cameras and ultrasonic sensors; the L4 maps by LiDAR and keeps the map current automatically, and reaches to within 3 cm of the wall — the strip that otherwise stays manual. The L50 does the same thing at industrial scale: 2,200 m²/h, 55-litre tanks, up to 6 hours of runtime, 220 kg of machine, with forklift detection for mixed traffic in warehouses and logistics halls. The bigger change is not speed. It is that the water tank, not the shift roster, becomes the limiting factor.

Regulation has already followed the same logic. Automatic floor treatment machines for commercial use have had their own product standard since IEC 63327:2021, which applies in addition to IEC 60335-2-72, the standard covering floor treatment machines generally. The autonomous machine is not treated as a new species; it is the known machine with additional requirements attached to the fact that nobody is holding it. What that means in practice for safety, sensors and emergency stops is covered in our article on safety and standards for cleaning robots.

For the acquisition decision, autonomy adds one factor that a purely mechanical machine does not have: the machine carries software, maps and an account. Its usefulness then depends not only on its mechanical condition but on whether the manufacturer still supplies the series — and that is exactly where buying and buying used start to feel different for an autonomous machine than for a guided one.

Purchase is capital expenditure. The machine is on your balance sheet, you depreciate it, you keep whatever residual value it still has at the end — and you carry every risk attached to it: wear, repair, spare-parts supply, downtime and technical ageing. The upside is control. You decide when it runs, how it is configured, when it is serviced and when it is replaced, and once the machine is written down its running cost is genuinely low. The precondition is an organisation that can actually operate a machine: someone accountable for service intervals, a stock of wear parts, and a plan for the week it does not run.

Wear is not an exception on a wet machine, it is the operating mode. Brushes and squeegee blades sit in water and dirt every shift, and our own product data states a typical service life of three to six months for them, depending on the floor. Add filters, hoses, the suction path and the battery. None of these is dramatic on its own; together they are a recurring procurement and labour item that has to be somebody's job. If it is nobody's job, the machine does not fail on a specific day — it degrades. A worn squeegee leaves streaks long before anyone calls it a defect.

The item that decides the purchase, though, is the failure day. Not "does it break", but: how long until the part arrives, who fits it, what does the floor cost you in the meantime, and does anyone step in with a mop? An owner without a second machine has no answer other than manual work. That is not an argument against buying — it is the question you have to price before you buy, and it is the same question that separates all four routes. The complete list of what belongs in that calculation is in our article on what a cleaning robot costs.

Buying used is the only one of the four routes in which you take on wear that somebody else produced. The hour counter tells you how long the machine ran. It does not tell you the conditions: which floor, which detergent at which dosage, whether the waste water circuit was flushed after every shift or once a quarter, whether it was stored full over a long weekend. Two machines with identical hours can be in completely different states, and the difference sits precisely in the parts you cannot see from the outside.

The most expensive invisible item is the battery. Every battery ages with charge cycles and operating hours, and a machine standing still gives no indication of its state of health. Manufacturer runtimes — 70 minutes in standard mode on a guided SH1, up to six hours on an L50 — describe a new battery. A used machine that no longer gets through the shift is not a cheaper machine, it is half a machine: the area gets split across two runs, or a person finishes the rest. Ask for the operating hours, the charge cycles if the machine records them, and the battery replacement date, and treat a missing answer as an answer.

The second invisible item is specific to wet machines: everything that carries dirty water. Tanks, hoses, the suction path and the seals are in contact with waste water permanently, and without consistent flushing that leaves deposits, odour, blockages and reduced suction. A worn squeegee or brush is obvious and cheap to replace; a fouled waste water circuit is neither. Look inside both tanks, smell them, and check the suction hose over its full length rather than at the coupling.

For autonomous machines a third layer comes on top of the mechanics: software status, whether the series still receives updates and spare parts, whether the maps of the previous site were removed, and whether the machine was properly released from the previous owner's fleet account. A machine that cannot be enrolled in your account is a machine you cannot run — a class of problem that simply does not exist on a guided machine. And on the legal side: between businesses, warranty terms are largely a matter of contract, so how much cover you have and for how long can vary from one used sale to the next. Settle that in writing before you pay, not afterwards.

None of this makes used purchase a bad decision. It makes it a decision that fits specific situations: as a second machine covering exactly the failure day described above, as a reserve for peak loads, or where you have in-house workshop capacity and can judge and repair the machine yourself. What used purchase does not do is remove operating risk. It concentrates it — at the earliest possible point, and with the least information.

Leasing answers the question of how the machine is paid for, not the question of who keeps it running. Ownership stays with the lessor, the instalment is plannable, and the balance-sheet treatment differs from a purchase. Operation, however — maintenance, wear parts, repair, downtime — stays with you unless the contract explicitly says otherwise. A finance lease and a lease with a service component are two very different products that are routinely sold under the same word, so read which one is in front of you before comparing instalments: two contracts with the same label can allocate risk completely differently.

The second characteristic is the term. You commit to one specific machine for a fixed period. On a purely mechanical machine that is unproblematic — a brush and a squeegee will do the same job in five years. On an autonomous machine, whose usefulness rests on sensors, navigation and software, a long term is a bet that this series will still be the right answer in three or five years, in a product category that is still moving quickly. That risk is not eliminated by leasing; it is fixed in place by it.

And check the end of the contract before the start of it: return, purchase option, follow-on contract, and above all how the return condition is defined and who assesses it. On a wet machine "normal wear" is an elastic term — brushes, squeegee blades, tanks and the suction path all carry visible traces of use, and whether those count as normal is worth agreeing in writing rather than discovering at handover.

Renting is operating expenditure, and its decisive variable is not the rate but the scope. A bare rental contract shifts the payment pattern and leaves maintenance, wear parts and repair with you. A maintenance-inclusive one does not. This is where offers become incomparable without anyone lying, and it is why comparing two monthly figures side by side tells you almost nothing until you have read what each of them contains.

We work as a full-service rental model: maintenance, wear parts and repair sit with us, and if a machine fails we supply a replacement. What you procure under that model is not a robot but the availability of a cleaned floor — which also means the running-cost items in this article are not theory to us, they are our own cost lines. The structural trade-off is the mirror image of a purchase: less control over the asset, no residual value, and in exchange no exposure to the machine's bad days. How that works in practice is described on our service page.

On the rate itself we publish nothing here, and that is deliberate. What a machine costs with us depends on the area, the floor type, the cleaning frequency and the conditions on site — parking spot, socket, water connection, drain — and is therefore quoted individually. That is not a way of avoiding the question; it is the only honest answer to it. Anyone who names a figure without knowing those inputs is giving you an estimate dressed as a quote. The most reliable number available anywhere comes from your own floor: a trial deployment on site produces real square metres per hour, real cycle times including docking and manoeuvring, and an honest measurement of the manual work that remains.

Three questions settle this decision, and the first is utilisation. Purchase and leasing tie up capital regardless of whether the machine runs; renting and full service tie the cost to an ongoing service. A machine that runs every night carries a completely different structure from one that runs once a week — and utilisation is not the same as area. A large floor cleaned weekly is poor utilisation; a smaller floor cleaned every night is good utilisation. Work it out per machine and per week, not per square metre.

The second is failure tolerance, and it is worth being specific. On a sales floor that has to be walkable before opening — the situation in supermarkets and retail — or in a production hall between shifts, a two-week standstill is an operating problem, not an annoyance. Where the floor tolerates a delay, ownership is far more comfortable. Where it does not, the replacement-machine question is the entire decision, and it is a contract clause rather than a sales promise.

The third is technical ageing, and it applies only to autonomous machines. Sensors, navigation and software are still moving; a mechanical scrubber-dryer is not. The longer you commit to one autonomous machine, the more you are betting on the state of that technology. Put briefly: whoever wants to own the machine and can operate it, buys. Whoever needs availability and does not want to carry machine risk, rents. Leasing answers a financing question, not an operating one. And used purchase suits those who have a workshop or need a reserve. Which machine class fits your floor in the first place is the question before all of these, and we walk through it in which cleaning robot suits your operation.