High-rise windows are cleaned four ways in 2026: rope access crews, suspended gondolas, drones, and facade robots. Rope access handles awkward geometry, gondolas suit tall flat towers, and robots now win on repetitive, uniform glass where safety and cost matter most.
Ask a facility manager how their tower gets cleaned and you will usually hear one of those four. The interesting part is not the list — it is which one is quietly losing share to the others.
What is rope access and when is it still needed?
Rope access is a technician on two lines, working the glass by hand. It needs no building-mounted equipment, so it reaches atriums, setbacks, and heritage facades that no gondola can serve. The catch is human risk: a fall-protection incident on a 30-storey rope job is a serious event, and insurance premiums reflect it.
Rope teams cost roughly $2.50–$5.00 per m² per clean in most markets. Add a full day’s setup and a weather window, and a single pass on a 5,000 m² facade runs into five figures.
How do gondolas compare?
A gondola is cheaper per square metre than rope access on a tall uniform tower — think $1.50–$2.50 per m². But it requires permanent roof anchors and a building designed to carry the rig. Retrofit a gondola onto an older skylight-heavy building and the capital cost eats the savings for years.
Where do facade and skylight robots fit?
Robots are the newest entry and the one growing fastest. A tethered facade robot like the Lingkong K3 works a vertical grid at a steady 60–120 m²/h, never needs a person on a rope, and can run at night when the building is empty.
On skylights and roof glass specifically, obstacle-crossing machines such as the Lingfeng S1 handle framed panels and seams that stop a rigid robot. The operator stays on the roof, out of the fall zone.
Where robots lose is irregular geometry. A building with twenty-odd glass shapes, deep recesses, and no repeatable panel grid is still rope-access territory in 2026.
Are drones actually cleaning glass now?
Drones are mostly inspecting, not cleaning. A drone can spot a failed seal or a cracked pane in minutes, but the payload needed to press a brush against glass and carry water is heavy, and flight time falls fast under load. Treat drones as a survey tool that feeds the cleaning schedule, not a replacement for it.
Which method should you choose?
Pick by geometry first, then volume. Uniform vertical grid at scale: gondola or robot. Awkward atriums and skylights: robot where it fits, rope access where it does not. Anything with a live public space below the glass: favour the method that keeps humans off the line.
If you want model-level comparisons rather than method-level, work through all skylight robot reviews before you write the spec.
What does each method cost per square metre?
Rough 2026 figures, per clean: rope access $2.50–$5.00 per m², gondola $1.50–$2.50, robots under $1.00 on repeat grids, and drones effectively zero for cleaning because they mostly do not clean. The gaps look small per square metre and become large across a 5,000 m² facade cleaned four times a year.
Those numbers are why robot share keeps rising. The saving is not the hourly rate — it is the elimination of the access day, the setup rig, and the weather cancellation loop that eats a rope crew’s calendar.
Why does weather decide the schedule more than height?
Because every method has a wind ceiling, and on a tall tower the wind aloft is stronger and gustier than at street level. A rope crew and a robot both stand down above roughly 12–15 m/s. The difference is that a robot resumes in a two-minute setup, while a rope crew loses half a day re-rigging.
That recovery speed, not peak capability, is what makes robots attractive to facility managers in changeable climates. A method you can restart cheaply wins over one that is marginally better in perfect conditions.
What should the cleaning schedule actually look like?
Build it around weather windows, not fixed dates. Block a fortnight per facade section, then schedule robot runs into the calm days and reserve rope access for the geometry the robot cannot reach. Two passes a year on a clean tower and four on a dusty one is a reasonable starting point.
Log every run. Over a year, the log tells you the real cost per square metre and the real weather loss, and that is the data that justifies the next machine — or tells you the current one is underused.
What about the interior side of skylight glass?
Exterior robots do not clean the inside of a skylight, and on a mall or airport the interior face collects dust just as fast. That interior work is usually light-access and can stay with a ground crew on poles. Budget it separately so the exterior method choice does not quietly assume the whole job is solved.
The split matters for cost. If you quote a robot as replacing full manual cleaning but the interior still needs a crew, the saving you promised in the business case is cut roughly in half.
Key Takeaways
- Four methods dominate in 2026: rope access, gondolas, drones, and robots.
- Rope access wins on irregular geometry but carries the highest human risk.
- Gondolas cost less per m² on tall towers but need permanent roof anchors.
- Robots run 60–120 m²/h and can work at night with no one on a rope.
- Drones inspect more than they clean; payload and flight time are the limits.