Modern robot vacuum cleaners from Xiaomi and the Mijia ecosystem have advanced navigation systems, but their effectiveness depends on the quality of a digital map of your home. Map construction is a fundamental process without which you can not work in wet cleaning, separation into rooms or installation of no-go zones. It is a digital model that allows the device to navigate in space, avoiding collisions and skips.
The process of creating a map is not instantaneous and requires the user to pre-prepare the room. LDS-Scanning or vSLAM can form incorrect geometry, leading to navigation errors. Understanding the physics of the scanning process will help you avoid typical errors and get the perfect apartment layout the first time around.
In this guide, we will look at everything from floor preparation to fine-tuning virtual walls. A properly configured map turns a regular gadget into an intelligent assistant that can only clean the kitchen after breakfast or bypass the area with children's toys. The key to successful scanning is a stable Wi-Fi connection and no moving people or animals during the first pass.
Preparation of the room for the first scanning start
Before you activate the mapping mode, you need to do a space audit. The robot builds the model using fixed objects: walls, furniture, table legs. Temporary obstacles, such as scattered wires, boxes or chairs with low bars, can be perceived by the algorithm as permanent barriers.
Remove from the floor any objects that can confuse wheels or sensors. Charger cables, hanging curtains and high-pile mats are the main enemies of accurate scanning. If the carpet is dark, the elevation sensor can perceive it as a cliff or a ladder, creating a black hole on the map where the robot is afraid to go.
- 🧹 Remove all wires and cables from the floor, lift them to a height of more than 10 cm.
- 🪑 Turn chairs over to tables or remove them from rooms where cleaning is planned.
- 🚪 Leave all interior doors open to ensure a pass-through.
- 🌑 Make sure there is enough light in the room if optical navigation is used.
The base should be placed against a wall with at least 1.5 meters of free space on the sides and 2 meters in front, and if the base is moved after the map is built, the robot will have to recalibrate, which can lead to a loss of reference to the coordinates.
Navigation technologies and their impact on the map
The quality of the final circuit depends on the type of sensor installed in your Xiaomi model. Laser rangefinder (LDS) devices create the most accurate and detailed plans, working in complete darkness. The laser unit rotating on the top panel emits beams and builds a point cloud, which is then converted into a two-dimensional drawing.
Camera models (vSLAM) or gyroscopic navigation work differently. Cameras need light to recognize visual markers on ceilings and walls. Gyroscopic models build maps along their trajectory, which can lead to accumulation of errors over large areas. Understanding this difference is important: for laser models, darkness is not scary, and for chamber models, it is critical.
⚠️ Warning: Do not wipe the glass window of a laser sensor or the camera lens with aggressive chemistry while working. Divorces can distort sensor readings and lead to the construction of "double walls" or false obstacles.
When using laser models, it is important to consider the height of the scanning unit. If you have furniture with low legs (less than 9-10 cm), the laser may not "see" the whole leg, but the robot will come under the table and realize that the space is occupied, which creates the effect of "floating" objects on the map, which disappear when re-scanning.
The process of initial map construction
The map-building mode is run through the Mi Home or Xiaomi Home app, which, unlike normal cleaning, does not include a brush or suck dust (in most models), and its only goal is to travel around the perimeter and interiors, fixing geometry.
Once you start the Mapping or Mapping command, the device will start spiraling or perimetering, and don't interfere with that. If you take the robot in your arms to move it to another room, the elevation sensor will go off, and the coordinates will go wrong, the map will be torn or displaced.
The whole process takes 15 to 40 minutes depending on the area, and the robot will return to the base several times if the charge is critical, but when it builds the map, it tries to complete the cycle, and if the device is stuck, the app will notify you, and the map will have to restart.
- 🚀 Click the "Map" button or select the "Mapping" mode in the menu.
- 👀 Watch the process in real time through the smartphone screen.
- 🚫 Do not open or close the door during the scanning process.
- 🔋 Make sure the battery is at least 60% before starting.
If you accidentally hit a robot or it gets stuck in the scanning process, you better interrupt the task and start over. Trying to "finish" the map after a failure often leads to layering, when the same room is displayed twice in different places in the plan.
💡
If the robot gets lost during the scan (flashing an orange indicator), don't try to guess where it is. Pick it up, take it to the base, press the reset button (if required by the instructions), and start building again.
Editing, separating rooms and zones
Once you've completed the scan successfully, you'll get a black and white contour of your apartment, but you'll need to train the system to work properly, and the app lets you divide the contour into separate rooms, name them, and adjust the cleaning order.
The Split tool draws a virtual line where the doorway should be, and this is critical for models with wash function, because they need to know where the laminate ends and the tiles begin, and the Merge tool is needed if a robot mistakenly splits one large living room into two rooms.
Virtual walls and no-go zones are a powerful control tool, and you can draw a rectangle around a flower tray or an animal bowl, and the robot will ignore that area in any cleaning program.
| Editing function | Action description | Impact on cleaning |
|:--- |:--- |:--- |
| Split | Lines between adjacent zones | Allows different power modes to be set for different rooms |
| Reunification (Merge) | Merging of two divided areas | Simplifies navigation in large studios without partitions |
| Prohibited Zone | Invisible Wall (red outline) | Complete no-entry zone (relevant to all regimes) |
| Virtual Wall | Linear Barrier | Blocks traffic in only one direction |
| Area cleaning | Selecting a specific square | Starting cleaning only in the designated area on schedule |
⚠️ Attention: Virtual walls and no-go zones are tied to the coordinates of the map. If you move the sofa or base, these zones will "go" with the map, either in the air or on the wall.
For models with wet cleaning function, you can adjust the automatic increase in suction power on the carpets (if there is recognition) or, conversely, turn off the water supply when you enter the carpet, all of these settings are available only after the correct division into rooms.
Keeping several maps and floors
Owners of two-story homes or apartments with attics often wonder whether to buy a second robot? Modern firmware Xiaomi (especially for the models of the S, G and V) allows you to save up to 4-5 cards at a time, which means that one robot can clean both the first and second floor.
To add a new floor, you physically move the robot and base to a new location, and the app selects the option "Save current map" or "Add new floor," and the robot will re-scan and match it to the base, creating a new layer in the memory.
How does a robot know it's on the other floor?
Switching between floors is automatic or manual. When you move the base to a new floor, the robot, once in the dock, may not immediately know where it is. It takes a few minutes to scan the environment and recognize the "native" map, at which point on the smartphone screen may appear a notification about the found new map.
- 🏠 Move the base to a new floor and connect to the network.
- 🤖 Put the robot on the base manually.
- 📱 In the application, select "Save the card" or wait for the auto request.
- 🔄 Name the floor (for example, "1 floor", "2 floor", "Mansarda").
You can clean the living room on the first floor at 10:00, and the bedrooms on the second one at 14:00.
Common Errors and How to Resolve Them
Even if all the rules are followed, artifacts can occur, and users often experience a robot being lost in a familiar room, which occurs when many new objects (guests, holiday decorations) appear in the room that have changed the geometry of the scan.
Another common problem is wall double-walling, which happens if the door is opened and then closed during the scan, or if the robot has traveled twice through the hallway with different paths, and in these cases, the "Recover Map" function or the complete removal and re-creation function helps.
If the robot is stuck in one place all the time, even though there are no obstacles, check the height of the rapids or the presence of dark surfaces. Black mats or glossy black floor can blind the sensors, creating the illusion of an infinite pit.
☑️ Checklist of the perfect map
What if the robot can’t see the base after the map is built?
Can I edit the map on the computer?
How often should I update the map?
💡
A quality map is the result of the preparation of the room before the first launch, and spend 20 minutes cleaning the wires and opening the doors so that the robot can work autonomously for years.
A properly constructed and customized map is not just a picture on your phone, but a complex navigation file that ensures cleaning efficiency. Following the preparation and editing guidelines, you will turn Xiaomi’s robot vacuum cleaner into a reliable assistant that will keep clean without your intervention.