How does the app change day-to-day use?
The app controls pattern selection, manual driving and schedule. The useful part is manual driving, because that is how you handle a window with a handle, a latch or a plant blocking one corner. Pattern selection matters less — the default Z-path is fine for a rectangular pane, and the fancier patterns mostly add passes.
The part that frustrates people is connectivity. These robots talk over 2.4 GHz and they do it from outside a building, which is the worst place for a weak signal. If the pane is on the far side of a thick wall, expect dropouts and plan to use the physical remote or the on-unit buttons.
Schedules are useful for commercial sites and nearly useless at home. On a site, a scheduled cycle starting at 7 a.m. means the operator arrives to a machine already working. At home, nobody wants a robot crossing the bedroom window at dawn.
Key Takeaways
- W3 suits 1.5–4 m² vertical framed panes; W2S Plus is the slower mid-tier; mini 2 stops near 3 m² indoor glass.
- All three struggle on frameless butt-glazed curtain wall, where edge registration is weak.
- None of them is the right tool for a sloped glass roof with vents or laps.
- Buy two pads per machine — drying time, not runtime, limits daily throughput.
- App pairing failures are usually router-side, not robot-side.
For the full set of machines we have tested on both windows and roofs, see the skylight robot reviews hub.
The ECOVACS window cleaning robot line in 2026 splits three ways: the WINBOT W3 for larger vertical panes, the WINBOT W2S Plus as the mid-range workhorse, and the WINBOT mini 2 for small indoor panes up to about 3 m². Choose by pane size and frame type first, grip second — all three hold glass reliably within their stated range.
People shop this line by price and then complain about coverage. The three machines are not trim levels of one product; they are sized for different glass.
Which ECOVACS window cleaning robot suits a large pane?
The W3. It is built for panes in the 1.5–4 m² range on vertical facades, with a pad area that covers a 1.2 m pane in roughly two passes. It handles windows with frames and does the standard Z-path, then an edge pass. Where it struggles is frameless butt-glazed glass — no frame means the edge sensing has less to register against, and the machine can overrun. Our ECOVACS WINBOT W3 review documents exactly where that happens.
The W2S Plus sits below it at a slightly smaller pad and a lower grip margin. It cleans the same kind of glass, just slower, and it is the model most offices end up buying because the price gap is real.
Is the WINBOT mini 2 enough for a skylight?
Usually not. The mini 2 is designed for domestic window panes: thick double glazing, small bathroom windows, and panes up to about 3 m². Put it on a 4 m² sloped skylight and it will finish the middle and come up short on the edges, because the pad geometry means it needs more passes near frames and the edge logic on a shallow incline is not the same as on a vertical pane.
That said, for a home with two skylights above a stairwell, the mini 2 is the correct product and a commercial machine is overkill. Read the WINBOT mini 2 review for the exact size cut-off we measured.
How does grip compare to commercial skylight machines?
All three ECOVACS units are designed for household and light commercial use, and they hold well on clean, framed glass. Where they differ from a machine built for roof work is the seal margin. A skylight machine like the Lingdu Intelligence Lingyun Y3 is engineered to keep grip with a wet pad on a sloped surface and to cross raised vents — our Lingyun Y3 review shows the wheel and sensor changes that make that possible.
For an office with framed windows below 3 m, the ECOVACS line is a sensible buy. For a glazed roof with vents, laps and a 25° pitch, it is the wrong tool. That is not a knock on the product; it is a category boundary.
What are the recurring complaints?
Three show up repeatedly in owner reports. First, edge passes leave a thin wet line on some window frames because the spray nozzle sits slightly inside the pad edge. Second, the app pairing drops after router firmware updates, which is a network issue but users blame the robot. Third, pad cleaning is required every cycle for good results and most owners do it every third cycle.
The pad point is the one that changes outcomes most. A pad with dried residue from two cycles ago will streak glossy glass no matter which model you bought.
Should you buy the W3 over a commercial machine?
Yes, if your glass is vertical, framed, under 4 m² per pane, and indoors. No, if the glass is a roof, sloped above 15°, or above 4 m. The W3 will clean a tall window beautifully and will be a frustrating purchase on an atrium skylight. Buy by surface, not by brand loyalty.
One more practical note: for a building with more than a dozen panes, the difference between models matters less than having two pads per machine so one can dry while the other works. That single choice roughly doubles daily throughput.
How does the app change day-to-day use?
The app controls pattern selection, manual driving and schedule. The useful part is manual driving, because that is how you handle a window with a handle, a latch or a plant blocking one corner. Pattern selection matters less — the default Z-path is fine for a rectangular pane, and the fancier patterns mostly add passes.
The part that frustrates people is connectivity. These robots talk over 2.4 GHz and they do it from outside a building, which is the worst place for a weak signal. If the pane is on the far side of a thick wall, expect dropouts and plan to use the physical remote or the on-unit buttons.
Schedules are useful for commercial sites and nearly useless at home. On a site, a scheduled cycle starting at 7 a.m. means the operator arrives to a machine already working. At home, nobody wants a robot crossing the bedroom window at dawn.
Key Takeaways
- W3 suits 1.5–4 m² vertical framed panes; W2S Plus is the slower mid-tier; mini 2 stops near 3 m² indoor glass.
- All three struggle on frameless butt-glazed curtain wall, where edge registration is weak.
- None of them is the right tool for a sloped glass roof with vents or laps.
- Buy two pads per machine — drying time, not runtime, limits daily throughput.
- App pairing failures are usually router-side, not robot-side.
For the full set of machines we have tested on both windows and roofs, see the skylight robot reviews hub.