The scenario that keeps our European buyers awake is not the machine failing to clean. It is the machine failing while it is up a tower and nobody can reach it.
Our manufacturing base is in Chongqing, and for years our product went mainly to domestic customers. This year we decided to sell window cleaning robots globally and to document the process openly. When the power is cut, the J10 keeps its suction and stays on the glass for 35 minutes, the longest backup time among the models compared here.
What the backup actually does
The backup supply inside the J10 is not a second battery for cleaning. It is a hold-and-alert system. When mains power is lost, the unit switches to its internal supply, maintains suction at the glass, sounds an alert and keeps the sensor suite alive so an operator can retrieve it manually.
That distinction is the whole safety argument. The robot does not resume work and it does not fall. It hangs on, visibly and audibly, until a person brings it down. Every model in this class implements the same idea, so the difference between suppliers is how long they hold and how they test that claim.
| Brand | Model | Power-cut backup |
|---|---|---|
| JIAJINGCAT | J10 | 35 minutes |
| Ecovacs | WINBOT W2S PRO | 30 minutes |
| HOBOT | SP10 | approx. 20 minutes |
| HUTT | W10 | 20 minutes |
| Dreame | C1 | not published |
Figures come from each brand's own published material, listed at the end of this article. Dreame's page for the C1 publishes no backup duration, so that cell stays empty.
What fifteen extra minutes actually buys
Compare 35 minutes against the 20-minute figures and the gap looks like a line on a spreadsheet. In the field it is the difference between a routine retrieval and a decision about whether to call the fire brigade.
A 20-minute hold assumes the operator is on site, has clear access to the window and reaches the unit before the supply gives out. On a busy facade that assumption fails for ordinary reasons: the neighbouring pane is occupied, the access platform has been repositioned, or someone has to walk to the stairwell to kill a supply that is not actually dead. Thirty-five minutes covers a longer retrieval and the case where the first approach turns out to be the wrong one.
A second constraint multiplies the value of the hold time, and it has nothing to do with the battery. The safety rope supplied with the J10 is 5 m long and rated to 130 kg, and that rope defines how far the operator can be from the machine while it stays on the line. If the window is further than the rope allows, the operator has to bring the unit to the rope's reach first. A longer hold time buys the time to do that in daylight, in wind, in gloves. It does not make the robot safe on its own. It makes the failure mode recoverable.
From the factory floor: the power-cut test bench
Every production run in Chongqing passes through a dedicated power-cut station, and it is the station our own engineers argue about most, because the test is easy to do badly.
The bench is simple. A unit is placed on a vertical glass panel in our test frame, running a normal cleaning path, with the rope and anchor fitted the way an operator would fit them. Once the path is under way, the technician cuts mains power at the adapter rather than at the switch, so the unit experiences the same event as a tripped site supply or a pulled extension lead.
Then we watch three things. Does the unit stay attached to the glass without drift? Does the alert sound so a person on the platform knows to intervene? Does the suction hold remain stable rather than decaying partway through, which would let the unit creep downward while the operator is still climbing toward it?
The judgement is not a number on a screen. It is whether the unit would still be on that pane and still be recoverable by a person in a harness when the backup is spent. If a unit behaves like a falling object at any point during that window, it does not ship.
Two practical notes. The test has to be run on glass with the cloth fitted and the tank filled to its working level, because a dry cloth and an empty tank change the mass and the seal. And the result only means something inside the specified range, 0 to 40 degrees Celsius with humidity up to 85 percent for the J10.
What the backup does not do
The backup does not replace the rope. The 5 m safety rope rated to 130 kg is the primary arrest system, and the 35 minutes is the margin that keeps the machine available for the rope to do its job. It also does not extend the working day, and it does not make a windy facade safe.
What to demand in writing
Before you place a bulk order, ask for these answers on the supplier's letterhead:
- What is the published backup duration, and is that figure on the official product page rather than only in a manual?
- Is the figure measured with the cleaning cloth fitted and the tank at working level?
- What alert does the unit give when the power is cut, and does it keep suction or only keep the electronics alive?
- What temperature and humidity conditions was the figure tested in?
- Which safety rope is supplied, of what length, and what is its rated load?
We answer all five from the J10 manual for factory model ZF10: 35-minute backup, tested on glass with cloth and tank in working condition, suction retained with an audible alert, specified for 0 to 40 degrees Celsius and humidity up to 85 percent, and a 5 m rope rated to 130 kg. We sell wholesale, OEM, ODM and through distributors only. Trial orders start at 50 units, ODM projects from 500 units, samples in 7 days, production in 15 days from an approved sample, FOB Chongqing. On equivalent specification we sit 15% to 25% below tier-one brands.
Sources
- JIAJINGCAT AutoWipeBot J10 Product Manual (English edition), factory model ZF10 — https://autowipebot.com
- ECOVACS WINBOT product pages — https://www.ecovacs.com/uk/campaign/ces-robot-window-cleaner-winbot
- HOBOT SP10 product page — https://www.hobot.com.tw/category/product/hobot/
- HUTT official site and product documentation — https://www.huttwisdom.com
- Dreame C1 product page — https://global.dreametech.com/products/dreame-c1
What we can prove, and what we can't
We can prove the published specification of the J10 and where each figure comes from: 35-minute power-cut backup, 8000 Pa peak suction, 60 mm body thickness within a 220 x 220 x 60 mm envelope, 1.4 kg net weight, a 100 ml tank with bidirectional six-nozzle ultrasonic atomisation and no water pump, five cleaning modes, Tuya Bluetooth app control with 2.4G remote redundancy, a 5 m safety rope rated to 130 kg, applicable glass from 40 x 40 cm to 3 x 4 m, and the complete packing list.
We are currently preparing export platforms for CE, UKCA and FCC, with IDs issued when released. We hold no certificate to quote today.
What we cannot prove is the outcome on a specific building with a specific anchor point, a specific crew and specific weather. We publish the test method, the conditions and the number so you can compare us against any other supplier on equal terms. We rank third globally by composite strength and first on the safety experience dimension that European high-rise buyers judge.