Xiaomi Vacuum Mop (or its Dreame, Viomi brand analogues) can create a map of your home on its own, but it doesnβt always go smoothly. Problems with scanning accuracy, missed rooms, chaotic routes, or the inability to save a room plan are common problems for many users. The reason is often not because the device malfunctions, but because of improper preparation or settings.
In this article, you will find step-by-step instructions for building a map with the nuances of different models (Mi Robot Vacuum-Mop 2 Pro, Dreame D9, Viomi V3, etc.), as well as ways to eliminate typical errors. We will discuss how to optimize the scanning process, edit the saved map in the Mi Home application, and why sometimes the robot βlostβ in space. If you have just unpacked the vacuum cleaner or have been struggling with its βcapricesβ for several weeks, this guide will help you set up the cleaning as efficiently as possible.
β οΈ Important: The map-building algorithms for models with LiDAR (laser scanner) and gyroscope/optical sensors differ. In this article, we will show you what steps are relevant for your type of device.
1. Preparation of the room: what to clean and what to leave
Before you first launch a robot, many users make the mistake of not removing small objects or furniture with "legs-traps." A LiDAR scanner (if you have one in your model) is able to detect even small obstacles, but optical sensors and gyroscopes often perceive them as walls, which leads to distortion of the map.
β Get off the floor:
- 𧦠Socks, slippers, wires (especially black ones, which the robot may not notice).
- πͺ Chairs with thin legs (or lift them to a height of β₯10 cm).
- π Pet bowls, toys, boxes.
- πΏ Pots with plants on the floor (often perceived as obstacles)
β Don't touch:
- ποΈ Large furniture (sofas, cabinets) β it serves as a reference point for scanning.
- πͺ Doors (their position will help the robot determine the boundaries of rooms).
- π Plinths (if not higher than 2 cm)
π‘ Tip: If there are mirrors or glass surfaces in a room (like a coffee table), cover them with a cloth. Optical sensors can get confused due to reflections, and LiDAR can perceive them as passageways in the wall.
2.The first run: how to build a map correctly
The scanning process starts automatically when you first turn on the vacuum cleaner, but there are nuances that affect the quality of the map.
πΉ Step 1 Charging to 100%
Before the first run, fully charge the robot (the base indicator should be green). Models with LiDAR (for example, Mi Robot Vacuum-Mop 2 Pro) can start scanning with a low charge, but interrupt the process for recharging, which will lead to a fragmented map.
πΉ Step 2. Installation of the base
Place the charging station so that:
- π The sides of the wall were β₯0.5 m free.
- π« It was not exposed to direct sunlight (this interferes with IR sensors).
- πͺ There were no reflective surfaces (mirrors, glass) within a radius of 1.5 m.
πΉ Step 3: Start the scan
- Press the Power button on the robot body (or in the Mi Home app).
- Select Cleanup mode (not Zone or Point mode).
- Wait until the robot goes around the room and returns to base.
βοΈ Preparation for the first scan
β οΈ Warning: If Xiaomi's robot starts to ram around the room instead of routinely traversing it, it probably won't recognize the base as "home." Reboot it (hold the Power button for 10 seconds) and repeat the scan.
3. Typical scanning errors and how to fix them
Even if you're prepared correctly, the map can be distorted, and let's look at the most common problems and their solutions.
| Problem. | Possible cause | Decision |
|---|---|---|
| The map is fragmented (not all rooms) | Doors were closed during scanning or the robot was discharged. | Run a cleanup with open doors or manually transfer the robot to a missed room and wait for it to connect the cards |
| The rooms on the map "stick together" in one | There are no clear reference points (for example, interior doors) | Add virtual walls in the Mi Home app or temporarily place an obstacle (such as a chair) at the edge of the rooms |
| The robot is lost and cannot return to base. | Weak Wi-Fi signal or base shifted | Move the base to a more open location and restart the scan. |
| Map distorted (walls curved, disproportionate rooms) | Interference from reflective surfaces or bright light | Close the mirrors, turn off the bright lights and rescan the room |
Why does a robot draw a map with its feet up?
π§ Additional advice for LiDAR models:
If your robot is equipped with a laser scanner (Mi Robot Vacuum-Mop 2 Pro, Dreame D9), but the map still turns out to be curved, check the purity of the LiDAR lens (it is under the top cover). Wipe it with a soft napkin without pile - the dust on the lens distorts the scan data.
4.Editing the map in the Mi Home app
Once you've done a successful scan, you can split the map into rooms, add virtual walls or cleanup areas.
π± Step 1. Save the card
- Open the Mi Home app β select your robot.
- Go to the Map tab (or Map in English version).
- Click Save (floppy icon or checkbox).
π± Step 2. Division into rooms
- Click Edit (pencil in the upper right corner).
- Select Add room and circle the boundaries with your finger.
- Assign a name (for example, "Kitchen", "Bedroom").
π± Step 3: Adding virtual walls and zones
- π§ Virtual Walls: Click Add β Wall and draw a line on a map, useful for blocking areas where the robot should not go (for example, near a cat bowl).
- π― Cleanup Zones: Select an Area, circle the area, and specify the number of passes (e.g., 2 times for highly contaminated areas).
π‘
If you reset the map after editing, check the firmware version of the robot. In older versions (below 3.5.8) there was a bug with save changes. Update the software through Settings β Update the firmware in the Mi Home app.
β οΈ Warning: If you use a Viomi or Dreame robot, some map editing features may only be available in proprietary apps (Viomi Home or DreameHome). Mi Home often displays stripped-down functionality for these models.
5. Optimization of the map for multi-room apartments
If you have a large apartment (3+ rooms), the robot may lose contact with the base or incorrectly connect maps of different rooms.
π Problem 1: Robot doesn't move from room to room
Decision:
- Check the height of the thresholds β they should not exceed 2 cm (otherwise the robot will not overcome them).
- Open all interior doors during the scan.
- If the door is narrow (less than 60 cm), use the Virtual Wall with the passageway in the middle so that the robot βunderstandsβ that it is a doorway.
π‘ Problem 2: The map is split (two separate maps of one house appear)
Decision:
- Move the robot manually to the "lost" room.
- Start cleaning β the robot must connect the cards automatically.
- If it doesnβt help, reset the card (Settings β Map β Reset) and rescan the entire apartment.
π Problem 3: Robots can't reach far-off rooms
Decision:
- Install an additional base (if the model supports, for example, the Dreame D10 Plus).
- Use the function Continue cleaning after charging in the settings (included in the Schedule β Additional).
π‘
For apartments with an area of more than 100 m2, it is recommended to use robots with support for multi-card memory (for example, Xiaomi Mi Robot Vacuum-Mop 2 Pro or Dreame D9 Pro). They can save up to 3-5 floors and automatically switch between cards.
6.Save and transfer the map to the new robot
If you bought a new Xiaomi robot vacuum cleaner or reset your old one, you can move the map so you don't scan the room again.
π² Method 1: Export/import via Mi Home (for firmware 3.5.9+)
- Open the map in the Mi Home app.
- Click Settings (βοΈ) β Backup card.
- Save the file to your phone (it will be in.map format).
- Connect the new robot, go to its map settings and select Restore.
π₯οΈ Method 2: Manual file transfer (for advanced users)
If there is no backup function:
- Connect to the robot using SSH (see 4PDA forums for instructions for your model).
- Copy the map file by path /mnt/default/rockrobo/rrmap (for Mi Robot) or /mnt/UDISK/map (for Dreame).
- Transfer the file to the new robot in the same directory and reboot it.
β οΈ Warning: Moving the map between different models (e.g., from Mi Robot Vacuum 1S to Dreame D9) can lead to navigational errors, in which case it is better to build the map again.