A cordless window cleaning robot rated 90 minutes gives 55 to 72 minutes of actual cleaning. The other 12 to 18 minutes goes to climb, re-wetting and returning to the start point, and that gap decides whether one charge covers your roof.
Battery figures on cordless window cleaning robots are measured on flat test glass in still air. Roof work is not that. Here is what the 12 to 18 minute gap is made of, and how to size a unit for a real skylight run.
The practical question is simple. Can one charge clean one roof section without a hand on the machine? Most 2026 answers are no, and buyers find out after the return window.
Where does a cordless window cleaning robot lose runtime?
Four places. Climb and positioning, at 5–8 minutes on a 6 m run. Re-wetting passes, because the water tank holds 60–120 ml and a dry brush burns motor load and leaves streaks. Corner and edge retries, which cost 2–4% of cycle time each. And the return leg, which is not optional if you want the machine back.
Noise: none of this is a defect. It is the difference between lab and roof, and any vendor who quotes a single number without a derate factor is telling you something useful about their testing.
| Rated runtime | Real cleaning time | Roof area covered (180 m²/h) | Verdict |
|---|---|---|---|
| 60 min | 35–45 min | 60–80 m² | Small panes only |
| 90 min | 55–72 min | 108–150 m² | One roof section |
| 120 min | 75–95 min | 150–200 m² | Most skylights |
| Swappable packs | Unlimited | n/a | Right answer above 300 m² |
Does a bigger battery always help?
No. Beyond about 8,000 mAh, extra cells add 180–250 g. On a pitched skylight that additional mass costs suction headroom and heats the drive motors sooner.
A lighter unit with a swap-in second pack usually beats one heavy pack. Two 4,000 mAh packs and a 20-second swap give you a fresh 55 minutes mid-roof without a cable. That is the configuration worth paying for on anything above 300 m².
How should you size runtime for a glass roof?
Multiply cleaning speed by usable time, then apply a roof factor. A unit at 180 m²/h with 60 usable minutes covers 180 m² in clean open conditions. On a framed, pitched roof with panel steps, apply 0.6–0.7.
So 180 m²/h becomes 108–126 m² per charge. That single correction is the difference between a one-day job and a two-day job, and it is missing from almost every product page.
Is cordless actually better than tethered for skylights?
For roofs under 200 m², yes. No tether means no drag on the pitch and no anchor point on a fragile parapet. Setup drops from 10 minutes to 3.
Above 400 m² the maths flips. A tethered unit runs all day, and the cable management pain is smaller than repeated battery swaps and charging breaks. Cordless is a convenience feature, not a scale feature.
Our ECOVACS WINBOT mini 2 review covers the small-roof case where cordless genuinely wins, including how the mini form factor handles tight frame returns.
What are the battery claims to distrust?
Any runtime quoted without load, ambient temperature or glass angle. Lithium packs lose 8–12% of usable capacity per year in daily service, and a 2026 unit running in an unshaded plant room can reach 45°C storage, which accelerates that.
Also distrust fast-charge claims without a charge curve. A 90-minute charge to 80% then 60 more minutes to full is normal. The last 20% is where the heat is.
How do you charge on a roof without power?
Use a 500 Wh portable power station for a day’s work at two packs, or run a 12 V feed from the site supply. Packs should charge at 1C or lower to limit heat, which means 90–120 minutes rather than the 60 minutes on the box.
Charging in direct summer sun pushes pack temperature to 45–50°C and shortens life. Charge in shade, and never on the roof surface itself.
Do batteries survive winter storage?
Store at 40–60% charge, not full, and above 5°C. A pack left full in a cold plant room for four months comes back with reduced capacity, and warranty rarely covers that.
What runtime do real 2026 owners report?
On framed skylights in the 180–250 m² range, owners report one charge covering 110–150 m² with a single re-wet. Two packs cover a full 400 m² roof with one 15-minute charging break, which is usually enough to change a filter and refill the tank.
Numbers drift downward by 10–15% in the second year. Plan the route around that now rather than discovering it mid-season.
Is a smaller machine with more packs better than a bigger one?
For skylights, usually yes. A 3.2 kg unit with two packs handles frame lips at lower risk than a 5 kg unit with one large pack, because the lighter mass needs less suction headroom to stay planted on a pitch.
What are the risks to plan for?
Two failures repeat on real sites, and both are cheap to prevent once someone names them in the contract.
- Battery packs stored in a hot plant room lose capacity 8–12% a year, quietly.
- Re-wetting passes eat runtime and are not counted in any published figure.
Key Takeaways
- Derate rated runtime by 20–30% before using it in a plan.
- Roof factor of 0.6–0.7 on framed pitched glass is realistic.
- Two swap-in packs beat one large pack on pitched roofs.
- Above 400 m², a tethered unit scales better than cordless.
Before you commit to a model, read our ECOVACS WINBOT mini 2 review and all skylight robot reviews. Both are written for buyers running real rooftops rather than spec-sheet readers.