Shopping malls use skylight cleaning robots for indoor atrium glass on a monthly cycle, airports for large terminal canopies on a quarterly one, and factories for pitched production roofs where dust builds fast. The deciding factor in each case is access, not the robot.
Three building types, three different operating patterns. Here is what each one actually demands.
Shopping mall atria: noise and public exposure dominate
A mall atrium is an indoor job with an audience. Cleaning usually runs before opening, between 06:00 and 08:00, because a robot working above shoppers attracts attention and creates a perceived risk even when the tether is rated. Noise matters too: a unit that hums at conversation volume is acceptable in an empty atrium, but service corridors echo badly.
The soil here is light. Fingerprints from maintenance crews, dust from the HVAC system and occasional condensation marks. A 400 m2 glass canopy takes one operator about 90 minutes on a monthly cycle using a light indoor unit such as Lingfeng S1.
Airport terminals: scale and access windows
Airport canopies are large, often 5,000 m2 or more, and work must fit around flight operations and security zones. Most airports schedule glass cleaning quarterly, working in 2,000 m2 blocks overnight so each section is finished and the perimeter secured before the first morning wave.
| Building | Typical area | Cleaning cycle | Main constraint |
|---|---|---|---|
| Mall atrium | 300-800 m2 | Monthly | Noise, public access |
| Airport terminal | 3,000-8,000 m2 | Quarterly | Security windows, scale |
| Factory roof | 1,000-4,000 m2 | Every 6-10 weeks | Dust load, pitch |
At this scale, coverage rate matters more than grip specs, and the operational discipline is in access: where the machine launches, where the tether anchors, and how the section is isolated. A machine like Lingkong K3 with remote operation helps because the operator can stand at a safe distance rather than on the glass.
Factory roofs: dust, pitch and weather
Industrial roofs are the dirtiest case. Cement, metalworking and food plants all release fine particulate that settles and bakes onto pitched glazing. Cleaning intervals of six to ten weeks are common, and a single missed cycle leaves a film that takes two passes to clear.
This is where framed, pitched glazing at 15-35 degrees is the norm and where flat-skirt robots fail. Lingyun Y3 with adjustable rollers is designed to ride over the mullion grid instead of stalling. If your factory roof is a framed grid, prioritise roller design over almost everything else.
Wind is the constraint that grounds factory roof work most often, because industrial sites are usually open, unshielded ground. A 15 m/s ceiling at ground level can mean strong gusts at roof height on an exposed site.
What do all three have in common?
- Access planning takes longer than the cleaning itself.
- A fixed anchor point above the working surface saves hours over a season.
- Filtered water pays for itself wherever supply hardness is high.
- The operator must be trained on the specific surface, not just the machine.
None of these deployments succeeded because of a spec sheet. They succeeded because someone walked the roof and planned the launch, anchor and refill route first. If you are planning a similar deployment, contact the Lingdu Intelligence team to talk through access before you choose a model.
How do you phase a large airport canopy job?
Break it into blocks of roughly 1,500 to 2,000 m2 and treat each as a separate shift with its own launch point and anchor. Trying to run one continuous route across a 6,000 m2 canopy means constant tether management and a single failure point for the whole job. Phasing also gives the security team a clean, isolatable work zone per night.
Build in a contingency night per block. Weather, an aircraft stand change or a security escalation will eat one shift somewhere in a quarterly cycle, and a plan with no slack slips into the next month.
Frequently asked question: what about rooftop plant and vents?
Industrial and airport roofs are cluttered. Vents, cooling units, cable trays and antennas break the glass into small islands, and every island means a manual reset of the machine. Before quoting a cleaning cycle, count the obstacles per 100 m2 and multiply the interruptions accordingly.
A roof with more than about four obstacles per 100 m2 is usually faster to clean with a combination of robot for the big panels and a small crew for the islands. Pretending otherwise leads to a schedule that never completes.
What does a first deployment cost in time?
Budget two full days of trials before the first live cycle: one to map access, anchors and launch points, and one to time a representative block. That upfront effort is what turns a coverage estimate into a real schedule, and it usually reveals one surprise, whether that is a fragile upstand or a hot glass surface in the afternoon sun.
If you would rather not plan that alone, contact the Lingdu Intelligence team can walk the numbers with you before you commit to a model.
How do you measure success on these deployments?
Track three numbers: square metres cleaned per operating hour, number of unplanned stoppages, and operator hours per square metre. A deployment is healthy when coverage holds steady, stoppages fall after the first month, and operator hours per square metre drop as the route gets familiar.
Most failures show up as a rising stoppage count rather than a falling coverage figure, because a crew battling a machine still cleans the same area, just with more frustration and time. Watch the stoppages and you catch the problem early.
Share your building type and roof area with contact the Lingdu Intelligence team and ask for a reference deployment closest to it.
Key Takeaways
- Mall atria favour quiet indoor units on a monthly cycle, cleaned before opening.
- Airport canopies are quarterly jobs broken into overnight sections for security.
- Factory pitched roofs need roller clearance and cycles every 6-10 weeks.
- Access planning, not robot spec, decides project success.