Modern robot vacuum cleaners from Xiaomi and its sub-brands like Roborock or Viomi have impressive navigational functionality. But out of the box, the device doesn't know the layout of your apartment, and it's the initial construction of the map that becomes the foundation for all further automation. Without a properly calibrated route, even the most powerful machine will randomly rush around the rooms, skipping corners or getting stuck on carpets.
The process of creating a digital twin of your home requires user involvement, at least initially, and you need to provide the perfect scanning environment to make sure that you can do it. LDS-The sensor or laser rangefinder was able to read the geometry of the walls without distortion, and errors at this stage will lead to the application appearing "phantom" walls or gaps in the route, making it impossible to accurately zoning in the future.
In this guide, we'll take you through everything from the preparation of the room to fine-tuning the virtual constraints, learn how to manage the logic of cleaning by creating complex scenarios that save time and battery life, and make sure that the robot doesn't miss an inch of usable space.
Preparation of the premises for the first launch
Before you start scanning, it's critical to prepare the physical space. The robot builds the map by relying on the reflection of laser beams from objects, so any dynamic changes during this process can knock the algorithm down. Remove wires, clothes, toys and other small objects from the floor that can confuse the wheels or be perceived as a stationary obstacle.
Special attention should be paid to the lighting and condition of the sensors. The first scan of the map should be carried out with artificial or bright natural light on, as optical sensors (fall sensors) and cameras (if any) can not work properly in complete darkness.
Doors to rooms that need to be cleaned need to be kept open. If you plan to use the room-division function, the robot has to physically drive through the opening to lock the passage. Closed doors will be perceived as blank walls, and later you will have to manually edit the map or put up with the fact that part of the apartment will remain untidy.
- 🧹 Remove all wires and small objects from the floor around the perimeter of the apartment.
- 💡 Provide bright lighting in all rooms where cleaning is planned.
- 🚪 Leave the interior doors open for free passage.
- 🧼 Wipe the laser rangefinder and drop sensors with dry cloth.
Primary map construction and calibration
Starting the first mapping cycle is the most critical moment: in the Mi Home or Xiaomi Home app, you need to choose Quiet Cleaning mode or just start the standard cycle, but with the important condition that at this time there should be no people or animals in the apartment, walking around the robot will knock its orientation, and it will begin to rebuild the already created areas again, which will lead to the appearance of artifacts on the scheme.
The robot will move around the perimeter and then fill the center of the room. At this time, do not pick up the device and do not interrupt work. If the device is stuck, carefully release it and return it to base, but it is better to wait for the cycle to be completed, even if it takes longer than usual. Once completed, the application will appear saved map, which the system will prompt to edit.
⚠️ Warning: Never transfer the base (dock station) while building the map. The robot is tied to the coordinates of the base, and its displacement even by 10 centimeters will lead to a complete dissynchronization of the map with reality.
If you have a problem in the scanning process, you can reset the map and start over again, and then select the "Delete card" or "Reset" option in the card settings. Before you rerun, make sure that the robot is on the base and fully charged, since building the map is an energy-intensive process that requires up to 100% charge for large areas.
Editing the map and dividing into rooms
Once the map is drafted, we move on to structuring it, and the app allows you to divide the space into separate zones, so you can set different cleaning modes for the kitchen, bedroom or living room. Use the "Separate" tool, drawing lines to places where there are doorways or obvious boundaries between rooms.
The reverse is the combination. If the robot mistakenly splits one large room into two, use the "Combine" tool. You single out adjacent areas, and they become one. This is important for correctly calculating the area and the time of operation. And at this stage, you can also rename the rooms by assigning them clear names that will be displayed in voice reports and statistics.
If the separation line goes through a cupboard or a table leg, the robot can try to go where there is no physical space, or, conversely, skip the section along the wall. Virtual walls created by editing take precedence over real obstacles in the logic of the device.
- 🏠 Use the "Separate" function to allocate separate rooms.
- 🔗 Combine mistakenly divided zones to correctly calculate the area.
- 📝 Give rooms clear names for easy management.
- ✏️ Exactly draw boundaries on architectural elements (walls, rapids).
Configure virtual walls and forbidden areas
One of the most powerful features of Xiaomi's ecosystem is the ability to create invisible barriers, virtual walls and restricted areas that prevent the robot from entering certain places, and this is indispensable for protecting against entanglement in computer wires, curtains, animal beds or areas with often spilled water.
There are two main types of constraints: the first is the "No-go zone" (square), where the robot will not go under any circumstances; the second is the "Virtual Wall" (line), which acts as a barrier that cannot be crossed; the setting is made by simply drawing a finger on the smartphone screen on top of the stored map, the changes are applied instantly after saving.
☑️ Verification of virtual restrictions
It's important to understand that virtual walls are tied to the coordinates of the map, and if you move the sofa or the table, but don't update the map, the robot will still drive around the old place because it thinks there's an obstacle, so if you rearrange the furniture seriously, it's recommended that you remodel the map.
| Type of restriction | Form | Principle of action | Best application |
|---|---|---|---|
| Prohibited zone | Square. | Complete ban on entering the perimeter | Water bowls, heaps of wires, children's toys |
| Virtual wall | Line | Prohibition of crossing the line | Boundary between rooms, entrance to the carpet |
| The Invisible Wall | Line (old models) | Like a virtual wall. | Restricting access to stairs or balconies |
Cleaning scenarios and route selection
Modern algorithms offer several strategies for driving, and the standard Z-route mode is most effective for basic cleaning, as it allows for consistent lanes with minimal overlap, but other options are available for local cleaning or for working in difficult conditions.
You can set up the cleaning procedure in the app, for example, it makes sense to vacuum the living rooms first, and leave the kitchen and bathroom for last, so as not to carry moist dirt on the clean floor. There is also a Carpet Bypass feature available if you have a model that lifts the brush or enhances suction, which allows you to avoid wetting the pile during wet cleaning.
The Secret of Cartography Accuracy
For apartments with very complex layouts or lots of obstacles, you can turn on Precision Cleaning, which is speed for the sake of quality of walking along the boards, in which case the robot will make more passes and turns, but the result will be cleaner.
- 🔄 Z-shaped route – the standard for quick and high-quality cleaning.
- 🧹 Local cleaning – cleaning a selected area or room by click.
- 🚫 Bypass carpets - prevents wetting of the pile when washing floors.
- 🔁 Repeated Passage – Good for Highly Contaminated Sites.
Preserving maps for multi-storey houses
Owners of homes or two-level apartments are more fortunate: most modern Xiaomi models support saving up to 4-5 cards. This means that you do not need to reconfigure virtual walls every time you move the robot to another floor.
To add a new floor, just take the base and the robot to a new level, start building a new map and save it under a separate name (for example, "2nd floor" or "Mansard"). In subsequent launches, the robot will know where it is. If it makes a mistake, you can manually switch the active map in the application.
⚠️ Warning: When you move the robot to another floor, be sure to turn it off or put it on the base before transporting it. If you move the machine on, it will try to match the new environment with the old map and create porridge from the data.
The map recovery feature is useful if you accidentally delete data or the robot "forgot" the layout after resetting. The map history often retains a previous version that can be reactivated without having to re-scan the maps.
💡
If the robot lost the map after moving, try manually starting the map construction in the "Silent cleaning" mode without the garbage collector, so that it does not soil the floor during calibration.
Frequent problems and their solution
Even with careful adjustment, there can be situations where the robot is acting strangely, and a common problem is "losing" the robot on the map while it continues to drive, but the app does not update its position or it stands still, usually solved by cleaning the impact sensors and wheels, and checking whether it is stuck under low furniture.
Another common scenario is not wanting to enter a particular room. Check the width of the doorway: if it is less than 60-70 cm (depending on the model), the robot may physically fail or find the passage too risky. Also make sure that the map in this place is not drawn random virtual wall.
💡
Regular maintenance (cleaning sensors and wheels) is more important for route accuracy than frequent reconfiguration of software settings.
If neither method works, the last resort is to reset the map completely and create a new one from scratch. Sometimes software failures accumulate and the blank slate solves the problem of incorrect positioning. Remember to update the robot's firmware, as Xiaomi is constantly improving its navigation algorithms.