Glass roof robot ROI reduces to three numbers: net cleanable area in m2, your current manual cost per clean, and realistic robot coverage per shift. Not passes. If manual cleaning costs you more than about USD 1.20 per m2 per visit, a robot usually pays back inside three years.
Vendors pitch efficiency. Finance cares about cash. Those three inputs decide everything, and two of the three are usually guessed badly.
Why is net cleanable area usually overstated?
Because people measure the building, not the glass. Subtract mullion width, vents, plant screens, service walkways and any panel the machine physically cannot reach. On a steel-trussed factory roof, net cleanable area is often 70 percent of the gross roof area, and on a complex terminal it can drop to 55 percent.
That single correction is usually what turns a good-looking business case into a marginal one. Do it before you build the spreadsheet, not after.
What does manual cleaning really cost per m2?
Add up more than wages. A rope access or MEWP team carries equipment, supervision, permits, a spotter, and often a premium rate for height work. Realistic all-in figures for high-access glass cleaning sit between USD 1.50 and USD 4.00 per m2 per visit depending on region and access difficulty.
| Site type | Manual cost per m2 per visit | Visits per year |
|---|---|---|
| Low factory skylight, walkable | USD 1.50-2.00 | 2-3 |
| Mall atrium, at height | USD 2.00-3.00 | 4-6 |
| Complex terminal, restricted access | USD 3.00-4.00 | 4-6 |
How do you set realistic robot coverage?
Take the pass rate and cut it by 40 to 50 percent. A machine rated at 288 m2/h on paper delivers 120 to 170 m2/h on a seamed roof with refills. Then subtract setup and pack-down, which on a difficult roof is 30 to 50 minutes a day no matter what the brochure says.
Our coverage rate analysis shows the working behind those numbers. Use 150 m2/h as a conservative planning figure and you will rarely be disappointed.
What does the payback calculation look like?
Take a 3,000 m2 net glass roof, four cleans a year, manual cost USD 2.50 per m2. That is USD 30,000 a year in cleaning spend. A robot with an all-in cost of USD 25,000 plus USD 3,000 a year for labour, water, pads and power changes the annual spend to roughly USD 3,000-5,000. Payback lands near 12 to 18 months, and everything after that is the return.
Now rerun it on 600 m2 cleaned twice a year. Manual spend is USD 3,000 annually. A USD 20,000 machine never pays back on cleaning alone. Sites like that buy for access, safety or presentation reasons, not ROI — and should say so honestly.
What is usually missing from the model?
- Consumables: pads, brushes and filters run 5-10 percent of machine cost per year.
- Downtime during your only access window, which is a real risk, not a rounding error.
- Water supply upgrades if the roof has no point within hose distance.
- Training days, and the productivity dip in the first month.
- Insurance and inspection changes, which can go either way.
Include those and your model will be boringly accurate, which is exactly what a finance director wants. If you want the three inputs stress-tested against your own roof, send the details through the contact page. The cost and payback article covers pricing bands in more depth.
What does the first year actually look like?
Month one is training and slow. Months two to four you learn the roof and shave setup time. By month six you have a repeatable route and a maintenance rhythm. Treat the first year as a learning curve with a cost, not as an instant saving, and the rest of the payback period will look after itself.
What return do owners actually report?
Sites we speak to that clean frequently, with real manual costs above USD 2 per m2, report the machine paid for itself within one to three years and then reduced annual spend by 70 to 85 percent. Sites that clean rarely report the machine mainly solved access and safety, with a weak cost case. Both are valid reasons to buy, but only one is an ROI story.
How do you track whether the machine is earning its keep?
Log three things every cycle: hours run, square metres covered, and water used. After six cycles the trend line tells you more than any brochure. If coverage per hour is drifting down, a consumable is failing. If water use is rising, your filtration has degraded. Owners who log these three numbers spot problems in weeks rather than discovering a damaged roof at year end.
What does a typical cleaning cycle look like?
A full cycle is more than the time the machine spends cleaning. It is the walk to the roof, the water fill, the boundary setup, the run itself, and the pack-down. On a well-mapped roof the setup is 20 minutes and the run dominates. On a difficult roof with restricted access it can be 45 minutes before the machine lays down its first pass. Budget the whole cycle, not the pass, when you plan a shift.
How does glass type change the brush and pad choice?
Coated glass, laminated units and textured or fritted panels all behave differently under a rotating brush. Fritted glass wears pads faster and needs lower brush pressure. Coated low-E glass scratches easily, so a clean pad and a fine brush are not optional. Textured or prismatic glazing traps dirt in the pattern and often needs more passes, which quietly doubles your clean time on those sections.
Key Takeaways
- Use net cleanable glass, typically 55-70 percent of gross roof area.
- All-in manual high-access cleaning runs USD 1.50-4.00 per m2 per visit.
- Plan on 150 m2/h realistic robot coverage, not the pass rate.
- Above USD 1.20 per m2 manual cost, payback usually lands inside three years.
- Small, infrequent roofs rarely pay back on cleaning alone — be honest about why you buy.