Scaffolding costs USD 8-20 per m ² to erect and dismantle, adds days of disruption and still needs cleaners on it. A skylight cleaning robot works the same glass for the price of electricity and labour hours. Rope access sits between the two. For roofs cleaned more than twice a year, the robot wins outright.
What does each method really cost?
Prices below are 2026 order-of-magnitude figures for a 2,000 m ² sloping glass roof in a mid-cost market. They include labour, access and disposal, but not glass repair.
Method Cost per m ² per visit Setup time Disruption Scaffolding + manual USD 9-18 1-3 days High Rope access USD 1.4-3.2 2-6 hours Medium Skylight cleaning robot USD 0.10-0.35 30-60 min Low
The robot figure assumes amortised machine cost over five years plus one operator. It ignores the purchase price entirely, which is the point – the ongoing cost is tiny once the machine exists.
When is scaffolding still the right call?
When the glass is being replaced, when the frame needs repainting, or when the roof has failed and you are inspecting as you clean. Scaffolding gives safe access to seals, gutters and fixings, which a robot cannot inspect. If your roof needs repair this year, bundle the two jobs.
What is the hidden cost of scaffolding?
Business interruption. A scaffolded mall entrance or hospital wing can lose footfall and parking for a week. Retail tenants on ground rent often have turnover clauses; a three-day scaffold over their frontage is a commercial argument, not just a safety one. Rope access cuts the disruption to hours, and a robot to minutes.
Which should a facility manager choose?
Run the numbers on frequency. Once a year on a small roof: rope access. Twice a year or more above 1,500 m ²: buy a robot. Roof repairs pending: scaffolding, and clean while it is up. Buying a robot for a roof cleaned once a year on a five-year old building with no service corridor usually disappoints.
There is a training overhead too. Someone has to be competent to deploy and recover the machine on a pitched roof, and to notice a suction cup going soft before it lets go. Budget half a day of training per operator and a refresher each year.
How do you compare methods fairly?
Use cost per m ² per clean, not headline day rates. A cheap rope crew that returns twice because of a missed gutter is not cheap. Add access disruption and the value of the space below, then compare over a two to three year horizon. That is where the robot’s low marginal cost per visit shows up most clearly.
What if the roof needs repair soon?
Wait. If the glazing seals, gutters or fixings need work within eighteen months, do the repair first and clean while access is already up. Buying a robot weeks before a re-glaze is a wasted training cycle. Sequence capital projects so access is shared.
Does method choice change insurance?
It can. Rope access and scaffolding carry well-defined safety profiles insurers understand. A robot is newer equipment and some policies list it separately, especially where it operates above the public. Tell the insurer which method you use, confirm the machine’s CE marking and IP rating, and keep maintenance records. Undisclosed roof machinery is a common reason claims get contested.
What does a fair business case include?
Cost per m ² per clean, access disruption, safety exposure and the value of the space underneath. A scaffolded retail frontage loses trade; a rope crew on a hospital wing disturbs patients; a robot does neither, which has a real money value even if it never appears on an invoice. Put those lines in the business case or the comparison is misleading.
How do you phase the switch from rope to robot?
Run them side by side for a season. Keep the rope crew for the awkward areas and steep pitches, and put the robot on the bulk of the roof. Compare water, time and quality on the same glass. Most sites find the robot covers 60-80% of the area, and the rope team’s job shrinks rather than disappearing entirely.
What is the wrong reason to switch?
Switching purely because a supplier promised a payback of eighteen months. That number rarely holds unless utilisation is high and staff are already trained. Switch when the frequency and area support it, and let the payback follow; forcing the timeline usually means a machine parked in a store room a year later.
One last angle: talk to the people who clean the building now. They know which panes streak, where water pools and which access route wastes an hour. A robot plan built without that knowledge usually repeats the same mistakes faster. The technology is new; the geometry is not, and the existing crew already mapped it.
Do the trial before you commit. Pick one roof, run both methods, and compare water, time and finish on the same panes. A season of side-by-side evidence settles arguments that no spreadsheet can, and it gives the team confidence in whichever method you choose. The trial costs one cleaning cycle and removes most of the risk from the decision.
And check the small print on access. Some buildings restrict roof work to certain days, and a robot that can only go up on a Sunday changes the payback again. Confirm the allowed windows before the business case is signed off, not after the machine has arrived.
Key Takeaways
- Scaffolding runs USD 9-18/m ² per visit and blocks access for days.
- Rope access lands near USD 1.4-3.2/m ²; robots near USD 0.10-0.35/m ².
- Pair scaffolding with any planned glass or seal repair, never use it just to clean.
- Choose the method by cleaning frequency and roof size, not by habit.
- Add operator training and cup inspection to any robot business case.
Compare methods on your own roof measurements via the contact page. Method comparisons also appear in the news archive.