Yes, a skylight cleaning robot can clean solar glass, but only with a soft roller and controlled pressure. Soiling typically costs a solar glass roof 3 to 5 percent of output in temperate regions and up to 20 percent in dusty areas. A robot pass can recover most of that, provided the brush does not abrade the anti-reflective coating. Hard brushes and dry scrubbing are the main risks.
Hybrid roofs, where photovoltaic panels and vision glass share the same plane, are increasingly common on 2026 commercial buildings. They create a cleaning problem with two sets of rules.
Why do solar and vision glass need different care?
Solar panels carry an anti-reflective (AR) coating, often porous and thin. Vision glass usually has a hard or soft low-emissivity coating for thermal performance. Both dislike abrasion, but the AR coating on modules is more forgiving of soft cloth and less forgiving of pressure.
You cannot use the same aggressive cycle on both without checking. A robot tuned for a dusty factory skylight may run a stiff roller and higher downforce, which is exactly wrong for a coated module.
How much output does soiling actually cost?
| Environment | Annual output loss from soiling | Cleaning upside |
|---|---|---|
| Temperate urban | 3-5% | Recover 2-4% |
| Dry and dusty | 12-20% | Recover 6-12% |
| Coastal with salt | 5-9% | Recover 3-6% |
| Near agriculture | 8-14% | Recover 4-9% |
Recovery is not 100 percent of the loss, because some soiling that was cleaned early would have washed off with rain anyway. Clean before the dirt bonds, not after.
What rules protect the coating?
- Soft roller only: microfiber, low downforce, no abrasive bristles.
- Clean water: filtered to avoid minerals that bond onto AR coating.
- Morning or evening: never shock hot glass with cold water.
- Test panel first: clean one module section and compare output before a full run.
Thermal shock is a real risk. Spraying cold water on glass at 60°C can crack a laminated panel. On a sunny solar roof, that is a five-figure mistake. Work when the glass is below roughly 35°C.
Is a robot better than a crew here?
On a large, uniform solar array, yes. Crews on a photovoltaic roof risk stepping on modules, and a single boot point can crack a cell and kill a string. A robot stays on the glass surface without concentrating load the way a human foot does.
On small or irregular arrays, no. The machine cost and access setup rarely justify it under 1,000 m2 of solar glass. Also be cautious with bifacial modules and elevated structures where the robot’s retention system cannot reach an edge safely. Our hybrid roof notes cover the layout decisions, and the contact page covers compatibility questions.
Key Takeaways
- Soiling costs solar glass 3-5% of output in temperate zones and up to 20% in dusty regions.
- AR coatings need soft microfiber rollers and low downforce, not stiff bristles.
- Cold water on hot glass risks thermal-shock cracking; clean below about 35°C.
- A robot beats crews on large arrays because it spreads load instead of stepping on cells.
- Skip robot cleaning for solar areas under 1,000 m2, where payback rarely lands.