Photovoltaic Cleaning Robots
-- When They Actually Pay Off
Photovoltaic Cleaning Robots: When They Actually Pay Off
A photovoltaic cleaning robot is a small self-propelled machine that crawls across panel glass, scrubbing off dust and dirt with a controlled water spray. The technology is real and it works. But here is the part most pages skip: robots pay off on utility-scale arrays with thousands of square meters of glass, not on the 60–100 W panel of a street light. We supply street lights for a living, and we do not put cleaning robots on them — this page explains the technology honestly, then tells you what we do instead.
How the machines work
A cleaning robot is small but packs in several subsystems. An automatic spray system meters water onto the glass, loosening dust and dirt so a brush or wiper can remove them. Onboard sensors read the panel surface in real time — dust thickness and distribution — and the controller adjusts cleaning parameters accordingly, spending effort where the soiling actually is. The control system plans its own path across the array and steers around obstacles: junction boxes, mounting feet, frame edges. The practical benefit over manual cleaning is endurance: the machine keeps working in heat, height, and dust conditions where a crew with a squeegee would be slow, expensive, and inconsistent.
Why soiling matters at all
Dust on glass blocks irradiance before it ever reaches the cells. Commonly cited soiling losses on PV arrays range from a few percent in rainy climates to 20% or more in hot, dry, dusty regions — the number depends entirely on your site, its rainfall, and its dust sources. Sustained soiling also creates another problem people miss: uneven shading across cells drives the shaded cells into reverse bias, which shows up as hot spots that degrade the panel permanently. If your site is dusty, cleaning is not cosmetic — it is output protection.
Our position: robots on street lights do not pay
Run the arithmetic the way we do on a project. A street-light panel is a fraction of a square meter. A crew member with a soft brush and water cleans it in under a minute, once or twice a year in most markets. A robot that can traverse that same glass costs orders of magnitude more, needs its own power, storage, transport, and supervision, and introduces a machine that itself needs maintenance. The robot wins when the glass is far away, extensive, and awkward — rooftop and ground-mount power plants. On a lighting fixture, it is a solution priced for a different problem.
What we actually specify for street lights:
- Tilt the panel at least 15° so rain self-cleans. A flat panel in a dry climate is a dust trap; the same panel at 20° sheds most of it with the first real rain.
- Site survey for dust sources — unpaved roads, construction, agriculture — at design time. A pole moved 3 m upwind of a dust source outperforms any cleaning schedule.
- Annual cleaning aligned with the dusty season , scheduled before the season that matters for your load.
- Check for hot spots at the same visit — a 30-second thermal or IV check catches the permanent damage early. We wrote up the symptoms in hot spot solutions for solar panels.
If you are sizing a system and want to know what soiling is costing you in kWh, the solar energy yield calculator lets you set the derate explicitly — run it before and after your cleaning assumption.
What this page does not cover
This page evaluates cleaning robots as a technology and states our position on their use in street lighting. It does not cover utility-scale O&M contracts, dry-cleaning (brushless) robots, or panel electrical fault diagnosis beyond hot spots. For complete fixture-level maintenance guidance, see the all-in-one solar street light documentation, which includes the cleaning interval we print in every manual.