Pudu SH1 in Practice: Why a Walk-Behind Scrubber Belongs in a Robot Fleet

Physically the SH1 is an upright machine measuring 49 × 53 × 120 cm, running on an 18 Ah lithium battery with no mains cable, operated via a 4.2-inch LCD touch screen. You push it the way you push an upright vacuum, except that it scrubs, vacuums and squeegees the floor in one pass. There is no map, no LiDAR, no route, no docking station. Pudu markets it as a smart upright scrubber-dryer, never as an autonomous device, and every field we checked in the official documentation is consistent on that point: it works when someone is walking behind it, and it stands still when nobody is.

That is the honest category question, and it deserves a straight answer rather than a marketing one. An autonomous scrubber earns its keep on the contiguous, open, reasonably flat share of a floor — the aisles, the concourse, the sales area, the corridor. What it does not do is the rest of the building: the strip along the skirting boards, the inside corners, the floor under fixed tables and benches, the metre behind the counter, the small back room whose door is usually closed, and the coffee that goes down at eleven in the morning while the robot is parked. None of that is a defect. It is the shape of the job. Which class of machine covers which share of a building is the whole subject of our guide on which cleaning robot suits which operation.

So the SH1 is not the robot's competitor. It is the tool for the person who handles what the robot structurally cannot, and the reason we put a manual machine into a robot fleet at all: a cleaning concept that only plans the robot's share is a concept with a hole in it, and the hole is always in the visible places — edges, corners, entrances.

The cleaning end is officially documented and consistent across Pudu's product page, its datasheet and its launch press release: 44 cm working width at the brush, 49 cm squeegee width, 27 kg of brush down pressure at 350 RPM, and 20,000 Pa of suction. The brushes are dual disc brushes, not a roller — the official datasheet labels them "Disc Brush", and the consumables list in the manual carries two of them. Productivity is stated as 1,100–1,600 m²/h, and Pudu's own store is unusually candid about how that figure is built: an average straight-line walking speed of 0.7–1.0 m/s multiplied by the working width. That is a theoretical band with no overlap between passes and no refill stops in it. Treat it as an upper bound, not a plan.

Weight is the interesting one, and it is settled — but not the way the website says. Pudu's official product pages, English and German, still state 17 kg today. The official product datasheet of 13 June 2024 states 27 kg / 59.52 pounds, which is an exact kilogram-to-pound conversion rather than a typo, and two distributors publish 27 kg as well. The physics agrees: a machine cannot statically exert 27 kg of brush down pressure while weighing 17 kg. We settled it the only honest way available to us, on our own stock: it is 27 kg. That also means the machine is a two-hand lift into a van, not a one-hand carry up a staircase — worth knowing before you plan multi-site use.

Noise is not settled, and we are not going to pretend otherwise. Pudu's current official product page states 72 dB(A) in Standard mode and 69 dB(A) in ECO. Pudu's own official datasheet from June 2024 states 76 dB(A) and 71 dB(A) for the same two modes. Both are Pudu. We have not measured our own unit, so we publish both figures and plan on the louder end: this is a daytime-capable machine in occupied space, but it is not a quiet one, and it does not belong next to people who are trying to sleep or concentrate.

Two smaller discrepancies belong in the same paragraph, because they run the same way. The product page and the datasheet give 4-litre fresh and waste water tanks; Pudu's own web store lists 4.3 litres. The product page and datasheet give 70 minutes runtime in Standard and 100 minutes in ECO; the store says up to 90 minutes in ECO and adds a Heavy-Duty figure of roughly 50 minutes that the other two sources do not mention at all. Nothing here is dramatic, but it tells you how much precision to expect from any single Pudu document — which is why every number on this page is traced to more than one.

The mechanical argument against a mop is simple and does not need a manufacturer to make it. A mop redistributes the same water across the whole floor and leaves it there to dry; the dirt it picks up goes back into the bucket and comes out again on the next pass. The SH1 separates the three steps: the disc brushes loosen the soil with fresh solution, the vacuum lifts the dirty water immediately, and the squeegee pulls the rest off behind it. What stays on the floor is a thin film that dries within minutes rather than a wet surface you have to cordon off. Nine cleaning modes on the touchscreen tune that trade-off — Normal, Heavy Duty, ECO, Soak, Low Residue, Water Suction and Brush Max, plus Accessory and ECO Accessory for working with attachments. Pudu markets the set as "7+2"; the count is the same nine.

Pudu quantifies the advantage as up to 80 % less water and cleaning agent and a 70 % reduction in cleaning time compared with conventional mopping. Those two figures appear consistently in the official product page, the datasheet and the Interclean launch release, so they are solidly Pudu's position — and that is exactly what they are: a manufacturer comparison against an unspecified manual baseline, not an independently measured result. We repeat them as manufacturer claims and would not build a business case on the second number without measuring your own floor first.

One feature is worth more attention than the percentages, because it decides whether a machine survives daily use in a kitchen: the waste water tank carries an air-solid-liquid separation system, which Pudu describes as first to market. It separates gas, liquid and solid particles inside the tank so that food debris and grit do not end up in the drain path. The practical translation is that the machine is far less likely to arrive at Monday morning with a blocked wastewater pipe — the single most common way a small scrubber gets written off as unreliable. The modular, quick-release design of brushes, squeegees, tanks and battery works in the same direction: the parts that get dirty are the parts you can pull off without tools.

The floors it is officially cleared for are all hard: tile, terrazzo, granite, marble, epoxy and hardwood. It is a wet machine — that is the whole point of it — so textile flooring, unsealed surfaces and anything that should not see water are out of scope, and belong to a vacuum or a dry sweeper instead.

The pairing we run most often is the SH1 alongside the CC1, and the division of labour writes itself from the two spec sheets. The CC1 carries 15 litres of fresh and 15 litres of waste water and is officially rated at 700–1,000 m²/h in full-coverage cleaning: it takes the open area, unattended, usually outside opening hours. The SH1 carries four litres and a person: it takes the perimeter, the corners, the back rooms and whatever happened during the day. Neither machine is trying to do the other's job, and the differences within the CC1 line itself are covered separately in our CC1 versus CC1 Pro comparison.

In hotels that split is unusually clean. The robot runs the lobby, the corridors and the breakfast area at night; the SH1 handles the kitchen and back-of-house tiling, the metre behind reception, the lift cars and the stretch along the skirting that a round machine never quite reaches. The noise figure sets the schedule: between 69 and 76 dB(A) depending on mode and which Pudu source you believe, this machine works during service hours in public areas and in back-of-house at any time — but not on a guest-room floor at night. That is not a limitation of the SH1 specifically; it applies to any wet scrubbing machine with a suction motor.

In fitness studios the case is different again, because the problem there is not open area but hard, tiled, constantly wet zones. Changing rooms, the tiled corridors between them, the entrance where outdoor dirt comes in: these are small, cluttered, full of benches and lockers, and a robot spends most of its energy manoeuvring rather than cleaning. A person with the SH1 finishes them faster and more thoroughly, at any hour the studio is staffed, and can steer around members instead of waiting for them to move.

Every honest assessment of this machine comes down to tank size. Four litres of fresh water and four litres of recovery is very little, and Pudu says so itself: the official web store states that refilling and wastewater discharge are typically required about every 20 minutes, depending on site conditions. Put that against the official productivity band of 1,100–1,600 m²/h and one tank fill covers roughly 370–530 m² — and that is arithmetic on the manufacturer's own theoretical figures, so the real number on a floor with obstacles and overlap is lower, not higher.

What that means for planning is concrete. A 3,000 m² hard-floor area is six to eight refill-and-drain cycles, each of which needs a water point and a drain within reasonable walking distance, plus an operator on the machine for the entire time. The battery adds its own rhythm: 70 minutes in Standard mode, 100 in ECO per the product page, then roughly 2.5 hours on the charger. The SH1 is therefore excellent for a few hundred square metres of difficult floor and structurally wrong for a few thousand square metres of easy floor. The autonomous machines are the other way round, which is precisely why we run both, and how to pick among the autonomous ones is the subject of our scrubber-dryer guide.

So the recommendation is a two-part one, and it does not include buying an SH1 to avoid buying a robot. If your building is small, cluttered, hard-floored and someone is already cleaning it by hand, the SH1 upgrades that work immediately and on its own. If your building has real open area, the SH1 is the finishing tool that makes the robot look good — because the parts a visitor actually notices, the edges and corners and the entrance, are the parts the robot was never going to do. Planning the second job is not an afterthought to the first; it is half the concept.