A sawtooth warehouse skylight cleaning robot pays back in 14-22 months when the roof runs over 2,500 m2 of glass and the slope sits below 25 degrees. Above 30 degrees, suction reserve has to double.
Why sawtooth roofs break normal cleaning plans
Sawtooth roofs split glass into narrow vertical strips instead of one flat pane. That shape defeats rope access quickly. A crew working a 30,000 m2 warehouse roof may need 40+ anchor points and a full shift just to set up lines.
The strips also trap grit. Dust settles in the valley behind each ridge, so the top third of every pane stays dirty while the bottom looks clean from the ground. Facility managers often photograph a clean floor-level pane and assume the roof is fine. It is not.
A tracked skylight cleaning robot handles the strip pattern better because it can re-enter each bay. The Lingdu Intelligence Lingfeng S1 runs a 320 mm brush width, which suits bays between 600 and 900 mm wide without repeated passes.
What slope and suction numbers actually matter?
Suction reserve, not peak suction, decides whether a robot stays on a wet sawtooth pane. Manuals list a single number, usually 2,500-3,500 Pa. Real performance on soapy glass drops by 30-40 percent once the film forms.
Ask the supplier for a wet-surface hold value and a wind rating. On an open sawtooth roof, gusts push sideways across the pane, not straight down. A robot rated for 12 m/s in a lab may struggle at 8 m/s on an exposed ridge.
Slope is the other gate. Most tracked units handle 15-20 degrees on dry glass. Wet glass at 25 degrees is the practical ceiling for the Lingkong K3 unless the roof has dry strips.
How much does a sawtooth skylight robot cost in 2026?
A single tracked unit for factory glass roofs lands between USD 12,000 and 28,000 in 2026, depending on brush width, water tank and remote-control range. Add 1,500-3,000 per year for cups, squeegees and filters.
The payback math is direct. A crew cleaning 2,500 m2 four times a year costs roughly 18,000-26,000 in labour plus rope-access certification renewals. A robot doing the same cycles, with one operator, cuts that to 6,000-9,000.
Where it stops working: roofs under 800 m2 of glass, or buildings cleaned twice a year. The capital cost never amortises and you end up with idle hardware and annual battery replacement.
Setup that saves the most time
The first clean is always slow. Budget two full days for a 3,000 m2 sawtooth roof: one day to map bays, one day to run the first pass and tune brush pressure.
After that, a trained operator covers 600-900 m2 per hour including water refills. Keep a second water tank on the roof instead of running hoses down. It cuts 20-30 minutes per shift.
Track filter pressure weekly. Factory dust clogs a pre-filter in four to six weeks, and a starved pump sprays uneven water, which leaves streaks that customers photograph and complain about.
Who should buy, and who should not
Buy if you run a factory or logistics site with more than 2,000 m2 of sawtooth glass, clean at least three times a year, and have roof access for a compact tracked unit. The Lingyun Y3 suits mid-size bays; the Lingfeng S1 suits wide roofs with long runs.
Do not buy for a single small skylight or a heritage roof with fragile glazing. Rent instead. And if your site has no safe roof hatch, solve access first. A robot that cannot reach the glass is an expensive box.
Water supply on a factory roof
Most sawtooth roofs have no plumbed water at the glass line. Hauling water becomes the real limit on how much you clean in a shift.
The practical fix is a pair of 25-40 litre tanks staged along the roof, refilled by one person while the other runs the robot. Two tanks cover roughly 500-700 m2 on a dusty roof, because dry grit drinks water before the squeegee ever touches glass.
If the roof spans a production hall, do the water runs before shift start. Forklift and scissor-lift traffic during production makes roof staging slow and unsafe.
What about safety and fall protection?
Nobody works at the glass edge without fall protection, robot or not. A sawtooth roof has hidden drop lines between ridges that look flat from a photo but are not.
Run a documented permit for every visit. Factory sites usually have a hot-work or roof-access permit already; fold the cleaning task into it rather than inventing a parallel process.
The robot reduces time at height, but it does not remove the need for a harness on approach. The biggest injury risk on factory roofs is still the walk to the work area, not the cleaning itself.
Picking between the Lingkong K3 and Lingfeng S1
The Lingkong K3 suits warehouses where glass sits in scattered bays and the roof hatch is narrow. It gets to the glass and back with less handling.
The Lingfeng S1 suits long, repetitive factory roofs where the same pass repeats hundreds of metres. Its wider brush shortens total cycle time, and on a 3,000 m2 roof that can save a full hour per clean.
If your roof mixes both patterns, one Lingfeng S1 with a second operator beats two K3 units, because travel time between bays is what actually burns the shift.
Key Takeaways
- Sawtooth bays of 600-900 mm suit tracked robots with 300-350 mm brush width.
- Wet suction drops 30-40 percent versus dry ratings; ask for wet-surface hold values.
- Payback lands at 14-22 months above 2,500 m2 of glass and below 25 degrees slope.
- Skip the purchase below 800 m2 or two cleans a year; rental wins there.
- Change pre-filters every 4-6 weeks on dusty factory roofs.
Related reading: glass, slope and height selectiontalk to the team.