Skylight cleaning robot handover in 2026 should take two full days with your own crew on your own roof: one day on setup, safety and water systems, one day on route planning and edge handling. Anything shorter and you will be calling the supplier within a fortnight.
Suppliers like to treat training as a formality with a product brochure and a coffee. That works for a floor scrubber. It does not work for a machine that lifts itself onto a glass roof inches from a 20-metre drop.
What must the first training day cover?
Day one is about safety and basics on the machine, not on a screen. Your people should physically roof-lift the unit, connect water, check the vacuum seal on a live panel, and practise the emergency stop until it is reflex.
- Roof access and securing protocol for the machine itself.
- Vacuum seal check and what a failing seal feels and sounds like.
- Water fill, filtration and draining, including end-of-shift drain.
- Battery charging rules and the cold-weather variation.
- Emergency stop and the recovery procedure after a stall.
What should day two cover that nobody plans for?
Route planning on your actual glass. Every roof has its awkward zones: the strip near the gutter, the panel that sits 8 mm proud, the corner where the atrium frame rises. A Lingkong K3 will handle these, but only if the operator has learned to plan the path around them and slow down at the edges rather than charging through.
Day two should also cover spotting diagnosis. When a client complains about streaks, the operator needs to know whether it is brush pressure, water quality, or a missed pass. That diagnostic habit is what separates a well-run site from one that keeps blaming the machine.
How long until a crew is actually competent?
Plan for four to six full roof cleans before the team is fast. The first two will be slow and will use more water than expected. Do not judge the machine on those runs. Sites that track their first six cycles and compare them to cycle twelve are the ones who keep good records and can defend the investment.
| Cycle | Typical time vs benchmark | Common issue |
|---|---|---|
| 1-2 | 150-200 percent | Slow setup, seal checks repeated |
| 3-6 | 110-140 percent | Edge handling still hesitant |
| 7+ | 100 percent | Stable; now optimise water and schedule |
Who should not run the machine?
Sites that rotate cleaning staff every month should not buy. The knowledge walks out with the agency worker. The same applies if the roof is only cleaned by an outside contractor who has never touched this hardware. Training is only worth paying for if the crew is stable and the machine is theirs to keep.
What does a good handover checklist look like?
Ask for it in writing before the machine arrives, and check that it names your roof, not a generic site. The list should cover pre-start checks, in-run monitoring, shutdown and storage. If the document could apply to any building in any country, it is a brochure, not a handover plan.
| Stage | Checks |
|---|---|
| Pre-start | Vacuum gauge, cup condition, battery level, water quality |
| On roof | Seal on first panel, edge probe test, route confirmation |
| In run | Water level, spot-check for streaks, watch for stalls |
| Shutdown | Drain lines, clean brushes, charge indoors, log the clean |
How do you keep skills from walking out the door?
Two people trained, not one. If a single operator holds all the knowledge and leaves, you effectively lose the machine until someone is retrained. Pair the machine with a short internal reference: photos of the route, the water quality used, and the fault codes seen so far. Fifteen minutes of documentation beats a twelve-page manual nobody reads.
Refresh the training annually, or after any incident. Roof conditions change, gaskets age, and an operator who learned the machine on day one has drifted into habits that may not be the safest ones. A half-day refresher with the supplier every year is cheap insurance on hardware that sits above a public space.
What should be in the logbook?
Date, operator, glass area cleaned, water used, weather, hours of roof access, and any fault or streak complaint. Six fields. Sites that keep this log can spot a failing suction cup or a filter that needs changing before it causes a bad clean, and they have the record a client or insurer may ask for later. It is the least exciting part of owning a skylight robot and the one that saves the most money over three years.
What does the first month actually look like?
Realistically, slower than the sales meeting suggested. Week one, the crew is learning the machine on their own roof and everything takes longer than planned. Week two, they hit their first awkward zone and have to re-plan a route. Week three, a suction cup needs attention and someone has to work out why. Week four, they start to see the rhythm, and the clean that felt impossible on day one is routine.
Budget for that month in labour hours and do not schedule a client-facing deadline in it. Sites that expect instant speed spend the first month frustrated; sites that expect a learning curve spend it building a routine they will keep for years.
Key Takeaways
- Insist on two full days of handover on your own roof, not a slide deck.
- Day two should be route planning and fault diagnosis on your actual glass.
- Expect 4-6 cycles before the crew hits benchmark speed.
- The first two cleans are always slow; do not judge the machine on them.
- Stable crews get value from training; rotating staff do not. Arrange handover through our contact page.