Modern cleaning devices are complex gadgets that require initial setup to work efficiently. Many Xiaomi Robot Vacuum owners face a situation where the device randomly rushes around the apartment, gets stuck in in inappropriate places or misses entire rooms, not because of a marriage, but because of the lack of primary training of the navigation system.
The process of adapting the robot to the geometry of your home is called mapping, and that's where the laser rangefinder (LDS) scans the space, creating a digital twin of your apartment, without which you can't fully use the functions of room separation and virtual no-go zones.
In this article, we will take a detailed look at how to conduct a primary launch, what conditions are necessary to accurately build a map, and how to avoid typical errors that confuse navigation. You will learn how to control the behavior of the janitor through the Mi Home application and achieve maximum autonomy of the device.
Preparation of the premises for the first launch
Before you turn on the device, you need to prepare space. Robots with LDS or vSLAM navigation are sensitive to clutter. Remove wires, socks, toys and other small objects that can wind on the brush from the floor, which will not only improve the quality of the scan, but also save the machine from breakdown.
Lighting is critical for optically navigable models, but laser systems need at least minimal light to operate the fall sensors correctly, open the curtains or turn on the lights in all rooms where cleaning is planned. Darkness can disorient the sensors.
⚠️ Warning: Do not start the learning process if there is water spilled on the floor or there are wet spots.This can lead to a short circuit of electronics or false alarm of carpet pollution sensors.
The robot must be able to drive from different sides to charge the base station for the correct connection of contacts, the distance from the base to the walls and furniture should be at least 0.5 meters on the sides and 1.5 meters in the front.
☑️ Preparation for map training
Initial connection to the Mi Home application
To manage and train, you’ll need a smartphone with the Mi Home app installed (or Xiaomi Home). Make sure your account is registered and the region is selected correctly (usually “Russia” or “China”, depending on the device version). Connect your phone to a 2.4GHz Wi-Fi network, as Xiaomi’s robot vacuum cleaners most often do not support 5GHz.
Turn on the vacuum cleaner and press two buttons on the case at the same time (usually an on button and a return button to the base) until you hear a voice message about the pairing mode. The Wi-Fi indicator should start flashing. In the application, click "Add device" and follow the instructions on the screen.
Once you've successfully connected, take your time to start a full cleanup. First, check the firmware version in your device settings. The manufacturer regularly releases navigation algorithm updates that can significantly improve the quality of the map.
- 📱 Make sure that the smartphone is on GPS Bluetooth for fast location data transfer.
- 📶 Connect only to a 2.4 GHz network, even if your router broadcasts one network with a common name (it is better to temporarily divide networks in the router settings).
- 🔋 Battery charge before first training must be at least 80% for uninterrupted work.
The process of constructing the first map
Now, to the most important thing, mapping, in the app, select Map or Map Building (sometimes called Zoning), and the robot will leave the base and move methodically along the walls, scanning the room with a laser.
At this point, it is strictly forbidden to move the base station or change the location of large furniture. If the robot encounters an obstacle, do not pick it up immediately - give it the opportunity to bypass the object. Raising the device during the scan will knock out the coordinates and lead to the appearance of "twins" of rooms on the map.
The process can take anywhere from 15 to 40 minutes depending on the area of the apartment, and the robot can return to the base several times to recharge if there is not enough space, and then continue building from the same place, which is normal behavior for SLAM algorithms.
Configuring virtual walls and zones
Once you've done that, you'll be able to edit the image, and the app will automatically divide the space into rooms, but you can adjust those boundaries, so that the robot can understand where the kitchen ends and the hallway begins.
Use Virtual Walls (invisible barriers) to protect sensitive areas, such as driving under chairs with thin legs or near a bowl of water for animals, and for long-pile carpets that can get stuck in a brush, it is also better to install a limiter.
| Type of restriction | Function | Where to apply |
|---|---|---|
| Virtual wall | Total travel ban | Children's room entrance, wired area |
| The no-go zone | Banning cleaning inside the square | Fringe carpet, animal tray |
| Zone cleaning | Cleaning only the allocated area | Kitchen after cooking, hallway |
| Consistent cleaning | Cleaning rooms one by one | First the kitchen, then the living room. |
To fine-tune the names of the rooms, tap the area on the map and select the appropriate name from the list (Kitchen, Bedroom, Living Room) which will allow you to give voice commands to the smart speaker in the future, such as, "Alice, ask the robot vacuum cleaner to clean the kitchen."
Schedule and scenario setting
When the map is ready and the zones are marked, you can automate the process. Go to the schedule settings and set the start time. It's recommended that you set the time when you're not home so that the engine noise doesn't interfere.
The Mi Home app also has automation scenarios, so you can set up a vacuum cleaner when all family members leave the house (geolocation) or after the smart sockets are finished, making simple cleaning part of the smart home ecosystem.
Don't forget to configure cleaning options for different types of coatings. If the application has carpet recognition, enable automatic suction power boost when you come on a waxy surface.
Typical errors and their solution
Even after training, the robot can behave strangely, and users often complain that the device is "lost" in the same room, which can be due to mirrors or glass coupes that the laser rangefinder does not see as an obstacle.
The solution is to put physical barriers or seal the bottom of the mirrors with opaque film, and problems arise when there are many black surfaces in the room — the drop sensors can perceive them as a cliff and refuse to drive.
⚠️ Warning: If the robot is constantly trying to drive onto the dark carpet and pulls back, use a virtual wall to deny it this zone, or seal the fall sensors (not recommended for permanent use).
Clean the laser rangefinder window regularly from dust, and the accumulation of dirt on the rotating tower leads to "noise" on the map and incorrectly plotting the route.
Maintenance of the navigation system
Maintaining high accuracy requires regular maintenance, once a week, clean the sensors with a soft dry cloth, pay special attention to the lower sensors and wheels.