Modern smart cleaning devices are no longer just randomly running drives. Owners of the Xiaomi and Roborock ecosystem often wonder how to get the appliances to clean exactly where they are dirty, ignoring clean areas. This is especially true if you just spilled crumbs in the kitchen or spilled water in the hallway, and the vacuum cleaner is quietly located on the base in the living room.
The answer lies in the proper configuration of mapping and the use of selective cleaning functions. The standard algorithm for moving a snake or spiral is effective for general cleaning, but not always convenient for point intervention. Fortunately, the Mi Home and Xiaomi Home software provides powerful tools for manual route control.
In this article, we'll look at all the ways that a robot can navigate, and you'll learn how to create virtual walls, divide a map into rooms, and use voice commands to instantly start cleaning a particular area. The key to all methods is that the model has a gyroscope or lidar to build a map, because the old chaotic models can't recognize rooms.
Equipment and software requirements
Before trying to control a route, make sure your equipment is technically capable of supporting this feature.Basic models without navigation that move randomly will not be able to understand the "clean up the bedroom" command, as they lack the concept of space. You need a model with LDS-laser or at least visual navigation.
The second important aspect is the app version. The maps are constantly updated, and older versions of the Mi Home may not display room split buttons or zone settings. Make sure that the region settings are China or Europe, as the interface may differ depending on the server.
⚠️ Warning: If a robot hasn't mapped a room before, it won't be able to move to a particular room on command. Initial calibration is mandatory.
It's also worth checking the battery power. Sending the device to a remote room, especially if it's on the other floor or at the end of a long corridor, requires a store of energy. If the charge is less than 20%, the robot may refuse to perform the task or start looking for a base instead of cleaning.
- 🤖 Robot model with support for map construction (LDS, VSLAM or Gyroscope)
- 📱 Installed and updated Mi Home / Xiaomi Home app
- 📶 Stable connection to 2.4 GHz Wi-Fi network
- 🗺️ Stored room map in device memory
Preparation of a map of the room for navigation
The foundation of accurate navigation is a well-constructed map, without which all your commands will be perceived by the device as chaotic movement, the first run should always take place in mapping mode, when the vacuum cleaner travels around the entire available area, not including a powerful suction turbine.
Once the map is saved, the app will suggest editing it, and at this point it's critical to divide the space into logical zones correctly, so if the robot sees the whole apartment as one big room, it's impossible to send it to the kitchen, because there's no kitchen as a separate object.
Use the in-app separation tool to draw lines between rooms. The system will automatically suggest options, but often they need to be adjusted manually, especially in areas of doorways or arches. The more accurate a digital copy of your apartment is, the better the robot will navigate.
Remember to rename the rooms you've created. Standard names like Room 1 or Area A are not convenient for voice control and quick identification. Call them Kitchen, Bedroom, Hallway, so that you can easily find them on the list in the future.
Use of the function "Cleaning by room"
The most civilized way to send a device to a specific location is through selective room cleaning, which is available in the app interface after the map is saved and divided, and you just see the apartment plan and you can tap into the desired room.
In some versions of the firmware, you press the Room Select button and then click on the contour of the desired room, and the robot will leave the base, proceed directly to the target area, ignoring other rooms, and begin cleaning according to the given algorithm.
☑️ Check before sending to the room
This is ideal when the contamination is localized, like when you cook in the kitchen or you come back from a walk in the hallway, and you don't have to wait for the robot to go around the perimeter, you direct it to the source of the dirt.
It's worth noting that when you pick multiple rooms, the robot will first remove one, then move on to the next one, and the order of the sequence is usually determined by the logic of the device, but sometimes it can be manually set by numbering the rooms in the application.
- 📍 Open the map in the Mi Home app
- 📍 Click the icon "Cleaning by room" or select the area
- 📍 Choose one or more rooms from the list
- 📍 Start the process with the Start button»
Configuring zones and virtual walls
If room-division is not available or not working properly, virtual walls and zones come to the rescue, a powerful tool that allows you to create artificial barriers that the robot cannot overcome, or, conversely, areas that need to be removed.
From the map menu, select the Zone tool. You can draw a rectangle in any part of the map. When you start cleaning the area, the robot will ignore the rest of the apartment and will only ply inside the frame drawn, which is actually the answer to the question of how to send a vacuum cleaner to another room without being tied to automatic separation.
⚠️ Note: Virtual walls and zones only work with a map. In "no map" mode, these functions are not available.
You can also use the No-Go Zone function, and if you want a robot to go into the living room but drive around a particular carpet or wired area, draw a red square there, and that will prevent the wires from getting stuck or being wound onto the brush.
For complex scenarios, where you want to clear a long corridor but not go into the side rooms, the combination of cleanup zones and no-go zones works best: you draw a cleaning zone along the corridor and put bans on the entrance to the rooms.
Voice control and automation scenarios
Today's smart home ecosystems allow you to control voice cleaning. If you have a column with Alice, Google Assistant or Siri, you can link them to a Xiaomi account. The Alice Tell the Kitchen Cleaner team will run the same process as the manual selection in the app.
And you have to configure the script in the smart home app, and the logic is that if you say a phrase, the smart speaker sends a signal to the Xiaomi server, which tells the robot to start cleaning the room that you've chosen, and that eliminates the need to get your phone out.
Secrets of voice commands
And you can also create automatic scenarios, like, "When I leave the house (geolocation of the phone), the robot cleans the living room," or, "At 14:00 every day, clean the hallway," which makes the process completely autonomous.
Voice control is especially convenient when hands are busy, and you can send the device to another room while you continue to do your own thing without interrupting the adjustment of the gadgets.
Comparison of robot navigation methods
To choose the best way to manage, it is worth comparing the available methods by their effectiveness and the complexity of the setup, each of which has its advantages depending on the situation.
Below is a table that will help you decide on the management method depending on your model and needs.
| Method | A map is required. | precision | Difficulty |
|---|---|---|---|
| Choosing a room | Yes. | Tall. | Low. |
| Zone cleaning | Yes. | Medium | Medium |
| Virtual walls | Yes. | Tall. | Tall. |
| Voice commands | Yes. | Depends on the name. | Low. |
As you can see from the table, choosing a room is the easiest and most reliable way to use it every day. Area cleaning takes more time to set up, but gives you flexibility. Virtual walls are a tool for permanent configuration, not for one-off tasks.
💡
Keep your maps backed up in the cloud. If you drop the robot or move, it takes seconds, not hours, to reset the room settings.
Solving navigation problems
Sometimes the robot refuses to go to the specified room or gets lost along the way, most often due to obstacles, if the path to the target room is blocked by chairs, toys or boxes, the navigation algorithm can consider the path impassable and cancel the task.
Another common problem is poor lighting or shiny surfaces, where laser sensors can not read black carpets or glass doors, which makes the robot “cannot see” the entrance to the room, in which case the installation of physical restrictive magnetic tapes, if the model supports them, or temporary closing of doors will help.
⚠️ Warning: Don't move the robot base after you've mapped it. If the base moves even 10 cm, the coordinates on the map will go wrong, and the robot won't be able to find its way to the rooms.
If the robot is constantly losing the map, try cleaning the sensors. Dust on the laser rangefinder (turret on top) or on the floor sensors can lead to positioning errors. Wipe them with dry soft cloth.
In complex layouts with many narrow aisles, you may need to manually adjust the map after each cleaning.Don't ignore the application's requests to join or divide rooms if the robot makes a mistake.
💡
The stability of navigation depends on 90% of the purity of the sensors and the absence of unnecessary objects on the floor during the construction of the first map.