A cordless window cleaning robot in 2026 runs 35–70 minutes per charge, covering 45–90 m² depending on suction mode. High suction roughly halves runtime. For interior windows it wins on convenience; for full glass roofs above 8 m² the corded version usually finishes the job faster.
The pitch is obvious. No cable trailing across the glass, no snagging at the frame, no outlet hunt in an empty building. The catch is that suction is the most power-hungry thing the machine does, and battery capacity is capped by the weight the robot can hold against the glass.
How long does a cordless window robot actually run?
Quoted runtimes are measured at low suction on a dry indoor window at 20°C. Real numbers, from units I have watched run on roofs and curtain walls:
| Mode | Quoted runtime | Measured runtime | Area per charge |
|---|---|---|---|
| Eco suction (1500 Pa) | 70 min | 55–62 min | 70–90 m² |
| Standard (2200 Pa) | 55 min | 40–48 min | 50–65 m² |
| Max suction (3000 Pa) | 35 min | 22–28 min | 25–35 m² |
| Max at 5°C ambient | 35 min | 16–20 min | 18–25 m² |
Lithium cells lose capacity fast below 10°C. A cold warehouse skylight in winter is the worst case in the table, and it is also the case nobody tests before buying.
Why is there a weight limit on cordless units?
Suction has to hold the robot’s own mass, the water it carries, and its payload against gravity. Adding battery capacity means adding cells, which means adding weight, which means more suction to hold it, which drains the battery faster. The engineering converges on roughly 2.2–2.8 kg for most cordless designs, and about 60–90 minutes of run time.
That is why cordless units tend to have smaller water tanks — typically 60–100 ml — and why they finish a room but not a roof. A corded unit can be heavier, carry a bigger tank, and hold higher suction continuously. The cable is the price of that.
What are the real downsides nobody mentions?
- Mid-run drop risk. When a battery sags near cutoff, suction weakens before the machine stops. On a vertical pane it slides; on a 35° roof it slides faster.
- Charge cycle time. Two to three hours to full. Two units do not solve it unless you also schedule the swap.
- Cell ageing. After 300 cycles, capacity drops 15–20%. Buy a replaceable battery or plan to replace the unit.
- Tether still required. Cordless means no power cable. It does not mean no safety line, and every manual says so on page 7.
The sliding-at-cutoff behaviour is the one that catches people out. Manufacturers rarely state the low-voltage cutoff threshold, and it differs meaningfully between models. Ask for it, or watch a unit run to empty before you trust it on a roof.
Corded or cordless: which should you buy?
| Use case | Cordless | Corded |
|---|---|---|
| Interior home windows | Yes — clear win | Fine, just inconvenient |
| Single skylight under 3 m² | Yes | Yes |
| Glass roof 8 m²+, regular schedule | Marginal | Yes |
| Curtain wall segments | No | Yes |
| Job site with no power nearby | Yes | No |
The interior case is genuinely strong. Dragging a cable across a living room, over furniture, to a socket behind a sofa makes people stop using the robot by week three. Convenience is a real feature.
Where does the battery maths get ignored?
Facilities teams buying for atrium roofs frequently spec cordless because the roof has no power outlets, then discover that one charge covers only half the glass. The workaround is two units and a scheduled swap, which costs 80% more than one corded unit plus a 30 m extension run.
For a fuller picture of how suction mode changes throughput, the Lingyun Y3 review breaks down the runtime-versus-suction curve on a real roof run, and the skylight robot review archive has side-by-side battery and area figures for every unit tested.
One more thing worth saying plainly. Battery capacity claims are the most inflated number in this category. A 5000 mAh pack at 14.4 V is 72 Wh, and no amount of marketing makes 72 Wh do more than about an hour of high-suction work. If the spec sheet lists a runtime but not watt-hours, that is a choice the vendor made.
There is a workaround worth knowing about. Two mid-range cordless units cost less than one premium corded unit, and the swap takes about ninety seconds if the roof is walkable. That setup beats a single battery unit on any roof over 60 m², provided you actually schedule the swap rather than discovering the limit mid-run. The failure mode is always the same: one unit dies at the far edge, someone drags it back across wet glass, and the clean is ruined.
Does faster charging fix the runtime problem?
Partly. Fast-charge units reach 80% in about 75 minutes in 2026, which is far better than the two to three hours the previous generation needed. But charging at that rate also heats the cells, and heat is what kills capacity over the long run. A pack that fast-charges every day will hit 300 cycles sooner than one that trickle-charges overnight.
For facility use that trade-off is fine, since the machine charges between scheduled cleans anyway. For daily domestic use, charge slowly and accept the wall time. Most manuals do not say this and the omission costs buyers a battery every two years.
Key Takeaways
- Expect 35–70 minutes per charge, with high suction cutting that roughly in half.
- Cold weather drops runtime another 20–25% — the worst case for unheated warehouse skylights.
- Cordless units are capped near 2.5 kg, which limits tank size and continuous suction.
- Suction sags before battery cutoff; on a sloped roof that is a slide, not a stop.
- Cordless wins indoors; above 8 m² of roof glass, corded plus an extension run is usually cheaper.