Before buying a skylight cleaning robot in 2026, ask five questions: what are the tested traction limits on wet glass, where are spare cups stocked, what is the real throughput on a roof like yours, what wind and rain limits apply, and who services the machine locally. Answers to those five separate a working asset from an expensive shelf item.
Question 1: What are the wet-glass traction limits in degrees?
Every vendor quotes a slope figure. The one that matters is the wet figure, because a roof does not stay dry. A machine rated to 25 degrees on dry glass may hold only 15 degrees once rain or spray is on the pane. Ask for the number from a test report, not a brochure. If the vendor cannot produce it, assume the wet limit is well below the dry one and plan accordingly.
This single question filters out a lot of optimism. Sites in wet climates should treat the dry rating as marketing and the wet rating as the real limit.
Question 2: Where are the spare cups and brushes stocked?
Consumables decide uptime. Ask where the cups and brushes ship from, what the standard lead time is, and whether the cups are a standard size or a proprietary mould. A six-week wait for a rubber cup will idle a machine longer than any feature gap, and proprietary consumables hand the vendor pricing power for the machine’s whole life.
| Question | Good answer | Red flag |
|---|---|---|
| Wet traction limit | Number from a test report | Only a dry figure, or a shrug |
| Spare parts stock | Regional stock, days not weeks | Ships from overseas only |
| Real throughput | Site-matched figure with assumptions | Brochure number, no context |
| Weather limits | Stated wind and rain cutoffs | ‘It handles most weather’ |
| Local service | Named technician or partner | No route to repair |
Question 3: What is the real throughput on a roof like mine?
Headline coverage figures assume clean glass and no interruptions. Ask the vendor to walk through their figure with your roof size, your tank capacity and your battery runtime. A realistic answer lands 40 to 60 percent below the headline. A vendor who insists the headline number is achievable on any roof has either not visited a real site or is hoping you have not.
Question 4: What wind and rain limits apply?
Robots stop working in weather, and the limits vary by design. Ask for the rated wind speed and the rain policy. A machine with a low wind cutoff may lose whole weeks in a coastal or high-rise location, which changes the cleaning schedule and the payback. Get the numbers before you buy, not after the first storm.
Question 5: Who services the machine where I operate?
A robot is a machine with a lifespan of several years, and it will need repair at some point. Ask who the local service partner is, how fast they can attend, and whether they hold parts. A vendor with no service presence in your country is selling you a machine you will have to ship overseas for repairs, and that cost can exceed the price of the consumables you were trying to save on.
Add a sixth question if you like: ask to see a working site near you. A reference visit answers more than any spec sheet, because you can watch the machine handle real glass, real edges and real weather. Vendors who cannot arrange one are telling you something.
Who should walk away, and who should proceed?
Walk away if the vendor dodges the wet traction figure, cannot name a service route, or only has a brochure throughput number. Those answers predict a machine that disappoints within a year. Proceed if the answers are specific, the consumables are standard, and a reference site exists nearby. That combination, not the price, is what protects your investment.
For help working through these questions on your own roof, the contact page is the place to start, and an earlier checklist covers the procurement side.
What else trips buyers up in the first six months?
Three things come up again and again. The first is access. A machine that cannot fit through the service door is useless, and buyers rarely measure the whole route before ordering. Walk the path from the truck to the glass with a tape measure, including the hatch, the parapet and the corners of the plant room.
The second is water. Sites assume there is a tap on the roof and discover there is not, which turns every refill into a climb with containers. That hidden labour eats the time the robot saves and can undo the payback. Confirm a water source before you buy, or budget for tanks and pumping.
The third is training. Two operators need real hours on the machine, not a one-hour handover. Teams that skip this run into avoidable suction faults and slow routes for months. Treat the training as part of the purchase, not an optional extra.
How do you compare two quotes fairly?
Put them side by side on total cost of ownership, not on sticker price. Add the machine, the first year of consumables, the service plan, the training and any access upgrades to the roof. The cheaper machine often loses once consumables and parts lead times are counted, because downtime and shipping quietly exceed the difference in purchase price.
Score each quote against the five questions above and give honest marks. A quote that answers all five with specifics beats a cheaper one that dodges two, every time.
Does a higher price ever make sense?
Yes, when it buys uptime. A machine with regional spares, a named service partner and standard-size cups costs more up front and less over five years. If your roof is critical, if downtime matters, the premium is worth paying. If your roof is a low-rise internal ceiling cleaned twice a year, it is not, and a simpler machine will do the job just as well.
The mistake is treating price as the decision instead of the outcome of the questions. Answer the five first, then look at the number. The right machine is rarely the cheapest one.
Key Takeaways
- Ask for wet-glass traction limits from a test report, not a dry figure.
- Confirm where cups and brushes are stocked and the lead time.
- Expect real throughput 40-60 percent below headline figures.
- Get rated wind and rain limits before you commit.
- No local service route means expensive repairs later; walk away.