For most commercial skylights, a skylight cleaning robot works on 8-19 mm monolithic or laminated glass; low-E and ceramic-fritted coatings need softer brushes and a lower squeeze-out pressure, because the coating is what fails first, not the glass.
Buyers ask about suction and battery. Almost nobody asks about the glass itself. That is backwards. The glazing on a mall atrium or a factory roof bay decides how much downforce the robot can use before it starts to polish a coating off, and how much water the squeegee leaves behind.
Why does glass type change robot setup?
Suction pads grip by creating a pressure differential. On smooth, flat glass that differential is stable. Add a ceramic frit pattern, a tinted film, or a heavy low-E stack and the surface changes: some spots are rougher, some are slicker, and the pad seal breaks in different places.
Brush pressure matters even more. Soft-coat low-E glass typically carries a coating only a few hundred nanometres thick. A nylon brush loaded at 30 N will scratch it. The same brush at 12 N will not. That is a setup decision, not a hardware purchase.
What glazing types do skylight robots handle in 2026?
| Glazing | Typical thickness | Grip | Watch out for |
|---|---|---|---|
| Monolithic tempered | 8-12 mm | Excellent | Edge chips at cut corners |
| Laminated (PVB/SGP) | 13-19 mm | Very good | Delamination near old seals |
| Hard-coat low-E | 10-16 mm | Good | Squeegee streaks on coatings |
| Soft-coat low-E | 10-16 mm | Good, lower pressure | Coating scratch, haze over time |
| Ceramic-fritted | 10-19 mm | Reduced seal | Frit debris clogging pads |
Lingdu Intelligence builds three machines with different pressure profiles: Lingkong K3 for larger open roof bays, Lingyun Y3 for tighter framed skylights, and Lingfeng S1 for lighter servicing work. Asking which one suits your glazing beats asking which one is “best”.
Does a coated skylight need a different squeegee?
Yes, and this is where most complaints come from. A rigid 70 Shore A squeegee pulls water cleanly on bare glass and leaves visible drag lines on soft-coat low-E. Dropping to a 50-60 Shore A blade, and lowering angle by a few degrees, usually removes the streak without changing the pump.
Hard water makes it worse. At 200+ mg/L calcium, spots appear within minutes on a hot skylight even after a perfect pass. A reverse-osmosis rinse of 0-15 ppm TDS is the fix. You can read a deeper treatment in our water system guide.
Who should not buy a robot for coated glass?
If your glazing is heritage single-pane glass with a hand-applied frit, or a soft-coat unit already showing hazing, a robot is a risk, not a saving. Those roofs need hand work and a conservator. Robots excel on modern tempered and laminated units where the coating, if any, is factory-applied and hard.
Key Takeaways
- Check glazing thickness and coating before suction specs; 8-19 mm covers most commercial skylights.
- Soft-coat low-E needs lower brush load, roughly 12 N vs 30 N on bare glass.
- Squeegee hardness matters more than pump pressure for streak-free coated glass.
- Keep rinse water at 0-15 ppm TDS to stop hard-water spots on hot roofs.
- A robot is a poor fit for heritage frit or already-hazed soft-coat glass.
How thick is too thick for suction pads?
Pad force has to overcome machine weight plus the pull of the squeegee. At 20 kg machine weight and a 1.5x safety factor, you want at least 300 N of total grip. That is easy on double-glazed units seated in rigid frames. It gets harder on a 25 mm triple-glazed unit with heavy mullions, where the pane itself is stiffer but the frame tolerates less point load.
The practical ceiling for most machines is around 19-25 mm of glazing. Above that, weight climbs and the frame becomes the limiting part. Ask the supplier for the point-load rating they assume, not just the suction number.
What about tinted and reflective films?
Aftermarket films are the fastest way to ruin a cleaning program. A solar-control film applied to save cooling load often softens under a hot rinse and lifts at the edges within a year. Check whether the film is warranted for mechanical cleaning. Many are not, and a robot brush is mechanical cleaning by any definition.
If the answer is no, your options are manual soft washing or a film removal decision. Do not let a crew discover this after a pass has scuffed four panels.
Does frame material change the cleaning plan?
Aluminium frames conduct heat and stay warm, so sealant next to them cures faster and cracks sooner. Timber frames hold moisture and swell, changing the gap the squeegee meets. Steel is stiffer and more forgiving. None of this is about the robot, but all of it changes where water ends up.
What is the single most common buying mistake?
Choosing the machine before walking the roof. Buyers read specs, pick a model, then discover the pitch, mullion spacing or glazing type rules it out. Walk the building with a tape measure and a level first. Ten minutes on the roof beats a week of datasheet comparison, and it usually shrinks the shortlist from five machines to two.
The second mistake is trusting a video. Demos run on clean glass under good light. Your glass is dirty, angled, and possibly filmed. Ask for a trial bay, with your glass, at your pitch, in daylight. Anything less is a guess dressed as a decision.