The two main skylight cleaning machines are an automated robot that climbs the glass and a water-fed pole kit that reaches it from a walkway below. Robots cost 300 to 4,000 USD and suit glass roofs above traffic, while pole kits cost 200 to 900 USD and suit reachable roof lights up to about 8 metres.
People conflate them constantly. A search for skylight cleaning machine returns both categories mixed together, which is why so many buyers end up with the wrong one and think the technology failed.
What does a robot do that a pole cannot?
Two things: it works where nothing can reach, and it cleans the outside of glass above a space you cannot stand under. A pole needs a standing position with line of sight. An atrium skylight over a four-storey void gives you neither.
Robots also scrub. A pole with a water-fed brush relies on flow and light agitation. Dried-on contamination, pollen cemented by a few rain cycles, or construction residue needs mechanical scrubbing, and that is where a robot with a rotating brush or a weighted pad wins.
The trade is weight on glass. A robot presses 2 to 3 kg against the pane through suction. On old or thin glazing that is a genuine risk, and our Lingkong K3 review notes the glass thickness a machine is rated for, which is the number people skip.
Where does a water-fed pole still win?
Single-storey and two-storey roof lights you can see at an angle, low-volume maintenance, and any glass you are not confident about loading. Pole work has no suction, no battery, no firmware. It also reaches the edges, which robots famously skip.
The edge band is the recurring robot complaint, and on a 2 metre roof light a 5 cm margin on all four sides is a visible frame of dirt that looks worse than the middle being clean. You end up redoing the edges by hand, which defeats the point.
How much does each option cost over five years?
A pole kit at 400 USD with a deionised water system running 150 USD a year in resin and consumables lands near 1,150 USD over five years. A robot at 900 USD with pads, seals and one battery replacement lands around 1,700 to 2,000 USD, plus your time supervising it.
On a commercial site the maths flips because labour is the dominant cost. A 500 USD per-visit rope access job done four times a year is 10,000 USD over five years. A 4,000 USD robot with 300 USD a year in consumables lands near 5,500 USD, and removes four descents a year from the risk register.
What about the water quality question?
Both methods care about it, differently. Poles use deionised water specifically to avoid spotting, which is why the resin cost exists. Robots typically use tap water with a small dose of cleaning agent and rely on a squeegee or pad pass to lift it before it dries.
On hard water, neither approach is clean. Dissolved solids above roughly 300 mg per litre leave visible spotting when the film dries, robot or not. If your site has hard water, budget for a small filtration or DI setup regardless of which machine you pick.
Who should buy which?
Buy a robot if the roof light sits over a void, above 4 metres, over public space, or if you have more than 20 square metres of glazing. Buy a pole kit if your roof lights are reachable from a walkway or a flat roof, if you have fewer than four of them, or if the glass condition is uncertain.
Buy neither, and call a rope access team, if the glazing is original single-pane from before 1990, if the frame shows movement, or if the fall zone is a public circulation route and you have no method statement. No consumer machine makes that safe, whatever the listing says.
If you want to compare the machines in the middle of that range, our full skylight robot review set lists them by slope rating, glass thickness and edge behaviour, which are the three numbers that actually decide whether one works on your roof.
How do you decide reach vs cleaning quality?
Write down the height, the pitch and what is underneath. Height sets the access method. Pitch sets whether a robot can hold. What is underneath sets the safety case and often the insurance answer. Those three facts eliminate most of the market before you look at a price.
What should you not assume?
- Robot for unreachable glass over voids; pole kit for reachable roof lights.
- Poles reach edges; robots leave a 2 to 6 cm band.
- Pole kit about 1,150 USD over five years; robot 1,700 to 2,000 USD at home.
- Commercial labour savings flip the maths towards robots.
- Neither is safe on pre-1990 single glazing above public space.