How to properly zone an apartment for a robot vacuum cleaner Xiaomi

Modern robot vacuum cleaners from the Xiaomi brand and the Roborock ecosystem have advanced navigation functionality that allows them to navigate the space with high accuracy. However, for maximum efficiency of the device, it is not enough to just turn it on and start it at full capacity. Competent separation of living space into logical segments, or zoning, is a key stage of initial configuration, which is often ignored by beginners. It is the right map that allows the gadget to understand where the kitchen is and where the bedroom is, and apply the appropriate cleaning algorithms for each type of surface.

The process of creating a digital twin of your apartment takes time, but the result is worth it. You get the ability to send the janitor only to a specific room, for example, in the hallway after guests arrive, or to prevent him from entering under the curtains in the living room. Without first building a full map of the room, the virtual walls and zoned rooms in the Mi Home application will not work. In this article, we will discuss all the nuances of settings, so that your smart assistant works like a Swiss watch.

To get started, make sure your vacuum cleaner is fully charged and installed on a base in the middle of the open space. The Mi Home app should be installed on your smartphone, and the gadget itself has been successfully added to the list of devices. Only after these basic conditions are met, you can start creating a map. You should not rush the process, since the accuracy of the first aisle depends on the accuracy of the robot's further navigation around your house.

Principles of Lidar and Map Construction

The basis for zoning is a map created by a laser rangefinder known as a lidar. This sensor, located at the top of the housing, rotates and scans the room, measuring the distance to walls and objects. Based on the data, a two-dimensional model of the apartment is built, which is displayed on the screen of your smartphone in real time, the accuracy of this model directly affects how correctly the virtual boundaries will work later.

During the first run, the robot follows a complex trajectory, often returning to the base and starting again. This is the normal behavior of an algorithm that tries to tie different parts of the room to a single coordinate system. At this point, you can absolutely not move the base or interfere with the movement of the device. Xiaomi uses complex algorithms SLAM (Simultaneous Localization and Mapping), which require environmental stability for correct calculation.

⚠️ Attention: During the first map construction, remove all small objects, wires and clothing from the floor. Lidar may perceive black carpets or dark rapids as a cliff (fall) and will not go there, considering it a danger, or will get entangled in wires, which will lead to a card error.

If there are many mirrors or glass partitions in the room, lidar can work with errors, since the laser beam passes through the glass or gives complex glare. In such cases, it is recommended to pre-stick time stamps on the bottom of the glass doors or simply be ready to manually adjust the map after the cycle is completed. The quality of construction also depends on the light, although modern models do well in the dark.

Preparation of the premises before the first launch

The quality of the final map and the efficiency of zoning depend on the preparation of the room by 80%. A robot vacuum cleaner is not just a toy, but a complex technical device that requires certain conditions to be met for proper operation. Ignoring the preparation will lead to the map appearing “artifacts”, false walls or tears that will have to be fixed by hand for a long time.

First of all, you need to have unimpeded access to the charging base, at least 0.5 meters free on the sides and 1.5 meters free on the front, so that the robot can easily dock with contacts after cleaning or recharging in the process. If the base is in a niche or corner, make sure that the robot can turn around for a ride.

• 🧹 Remove all wires lying on the floor: they can wind on the brush and knock the orientation of the gyroscope.
• 🚪 Open all interior doors: the robot should be able to build a single map of the entire apartment, not a separate room.
• 🪑 Lift chairs and light objects: If the legs of chairs are thin, the robot may get stuck between them, thinking it's an obstacle.
• 💡 Provide good lighting: Although lidar works independently of light, camera models (with visual navigation) require bright light.

You also need to figure out where the robot should be closed, which can be areas with long-pile carpets where the device might get stuck, or places with scattered parts of the designer, and once you build the map, you can put virtual walls there, but in the scanning phase, it's better to physically restrict access by closing doors or putting temporary barriers.

💡

If you have a two-level apartment, be sure to move the charging base to another floor and build a separate map for each level.

The process of primary scanning and saving the map

Starting the first cleaning cycle is the most crucial moment. In the Mi Home app, select Quiet Cleaning or Standard, but make sure the Mop function is turned off unless you plan to wash the floors immediately. A wet mat can leave traces on the map or create a false sense of obstruction if the robot gets stuck. The process can take anywhere from 40 minutes to 2 hours depending on the area.

As you move, watch your smartphone screen. You'll see the gray fog of war dissipate as you open the contours of the walls. If the robot gets stuck, the app will report it. Once you fix the problem (like you freed the wheel), press the continuation button. It's important that the robot find its way back to the base or complete the cycle to keep the map in memory.

Once you finish cleaning and return to the base, the card will automatically save. Some Xiaomi models require you to manually click on the "Save" or "Apply" button in the application. If you do not, the next start-up, the device can start building the map again, ignoring previous data. Make sure that the status of the card has changed to "Save".

Sometimes the map is saved with distortions, for example, one room can be superimposed on another or deployed 180 degrees, in which case the map can be edited: rotate, merge the broken areas or, in extreme cases, delete and start the scan again.

Configure virtual walls and forbidden areas

Once the map is saved, there's a fine-tuning phase, and virtual walls are a powerful tool to create an invisible barrier that the robot won't cross, and this is important for protecting areas with wires, or a pet bowl, or just rooms that you don't want the janitor to go into at the moment. There are two main types of constraints: lines (walls) and squares (zones).

Limitation lines work as barriers. A robot can drive up to a line like this, pull it down, but it never crosses it. It's perfect for separating the kitchen from the living room if you don't have doors, or for enclosing an area around a computer desk with a bunch of wires on the floor. The restriction zones (squares) prevent the robot from even entering the designated area.

To set up, go to the map menu, select the Virtual Wall tool or the Zone tool. Draw a line with your finger or highlight an area. You can create many of these zones. If you move the charging base to a new location, the map will go off, and all the virtual walls will be out of place, possibly blocking the passage to another room.

⚠️ Attention: Virtual walls have no physical height, if you have pets or young children, they can easily cross that boundary, and don't use virtual walls as a substitute for real safety doors for children.

The number of virtual walls and zones in a device's memory is usually limited (often to 10 per card). Try to use them rationally. If you need to deny access to the entire room, it's better to use the Room feature and simply not include it in the cleaning plan than to paint the entire area with a virtual wall.

Dividing the map into rooms and naming

The most useful feature for owners of large apartments is to divide a single map into separate rooms. By default, the robot sees the apartment as one large space. To understand where the "bedroom" is and where the "kitchen" is, you need to use the "room division" tool. In the application, you draw the separation lines where there are doorways or conventional boundaries.

Once separated, each zone must be given a name. The Mi Home app offers a standard list: Kitchen, Living Room, Bedroom, Bathroom, Corridor, etc. Why do you need it? First, it allows you to set different suction power for different types of rooms, for example, in the kitchen you need maximum power, and in the bedroom you need quiet mode.

The separation process is as follows:

1. Open the map in the app and click on “Maps Editing”.
2. Select the "Divide Rooms" tool.
3. Draw a line through the doorway, dividing the room into two segments.
4. Click on the resulting segment and select “Room Name” from the list.
5. Save the changes.

If the automatic separation went wrong (e.g., the robot didn't notice the threshold), you can use the Unification tool to erase the extra line, or the Drawing tool to add the missing wall. The accuracy of the separation affects the cleaning statistics: you can see in the report which room was cleaned and how long it took.

Cleaning scripts and room schedules

When you divide the map into rooms, you can create complex scenarios, so you can set the schedule so that on weekdays, the robot cleans only the kitchen and the corridor, and on weekends, it does the general cleaning of the entire apartment, which saves battery life and prolongs the life of the engine.

In the schedule settings, select Repeat, then specify the days of the week and the time. In the task settings, instead of Cleaning the entire map, select Rooms. Tick the desired rooms. You can also set the operating mode (Quiet, Standard, Max) and the carpet mode (Detour, Rise, Ignor).

Another useful feature is cleaning sequences, so you can set the order in which the robot will visit the rooms, for example, it makes sense to first clean the bedrooms (where it is cleaner), and then move to the kitchen (where it is dirtier) so that you don't carry garbage around the clean rooms. Set up the order in the "room cleaning sequence" menu.

Using scripts, you can automate the process so that you can spend weeks without touching the settings, the robot knows when and where it needs to clean, and the main thing is to remember to empty the dust container and add water to the tank.

💡

Properly configured room scenarios reduce robot life by up to 40% while maintaining perfect cleanliness in priority areas.

Frequent zoning problems and their solution

Despite the advanced algorithms, users can encounter errors, and the most common problem is a map that has been taken, which happens if the robot can't find the base, or if you rearranged the furniture during cleaning, in which case the map can split into two unrelated islands, and the solution is to remove the map and start building again, following the rules of the room preparation.

Another problem is that the robot ignores virtual walls. This can happen if the line is drawn too thin or in a place where the robot cannot physically stop (for example, on a slippery floor at an angle), make sure that the virtual wall is perpendicular to the direction of the robot's movement or forms a closed loop, and also check if you have updated the firmware, as sometimes the navigation algorithms change.

If the robot is constantly losing the map (forgetting the layout of the apartment), try to clean the sensors. Especially important is to wipe the cliff sensor and bumper sensors. Dust on the sensors distorts the data, and the robot thinks that it is in an unknown place, starting to rush randomly.

⚠️ Warning: Never lift or carry the robot vacuum cleaner while working unless it's stuck.The transfer disorients the gyroscope and lidar, guaranteed to lead to card loss and the need for re-scanning.

In rare cases, the problem may lie in the floor covering itself. Glossy black surfaces or mirrors to the floor can be tricked by sensors. In such areas, it is better to install physical limiters or magnetic tapes (if the model supports), since software zoning can work unstable.

FAQ: Frequently Asked Questions