The situation where a robot vacuum cleaner completes a cleaning cycle but does not find its way to the charging station is familiar to many owners of smart equipment. Instead of the usual signal to finish work, you see a frozen car in the middle of the room or, worse, hear an audible alert that the device is lost, a common problem that can be caused by both banal contamination of sensors and a software failure in the navigation system.
In most cases, Xiaomi or its sub-brands (Roborock, Dreame, Mijia) have a reasonably thought-out logic of behavior that allows you to quickly diagnose a malfunction. You don't have to carry a device to a service center if the problem lies in a wrong base installation or a downed room map. Let's see what factors affect the ability of the gadget to find its way home.
The first thing that comes to mind is that the battery is dead, but the current algorithms are designed so that when a critical charge level (usually 15-20%) is reached, the machine forcibly interrupts the cleaning and tries to return. If it does not go to the base at full or medium charge, then the problem is precisely navigation or physical obstruction. The critical factor is often the state of the charging contacts on the bottom of the base and the robot itself: oxidation or sticky dust creates high resistance, which makes the device not “hear” the base even in the immediate vicinity.
Pollution of sensors and charging contacts
The most common, but most common, reason for the device not to return to charging is dirt. The sensors that dock with the base are located on the bumper or the bottom of the body. If they are covered with a layer of dust, animal hair or pile, the infrared signal simply does not read correctly. The robot can circle the base, blind and disoriented.
Pay special attention to charging pins. Over time, metal contacts form an oxide film or stick hard-to-remove dirt that interferes with the transmission of current. Even if the device is on base, it may not start charging and after a while again leave for power until the battery sits at zero.
To clean, use dry soft cloth or cotton swab lightly moistened with alcohol. Do not use aggressive chemicals that can damage plastic or rubber elements.
- 🧹 Wipe the IR sensors on the bumper and bottom panel with a dry microfiber.
- ⚡ Clean metal contacts on the base and the robot with a strip or alcohol.
- 👀 Check if there is any small debris stuck in the crevices around the sensors.
- 🔄 Turn the device over and make sure the wheels rotate freely.
⚠️ Warning: Never use wet cloth to wipe sensors if it is very wet.Drowning moisture into the case through sensor holes can cause the board to short circuit.
Regular maintenance of sensors is the key to a long life for your smart assistant, and if you haven't cleaned them since you bought them, you shouldn't be surprised by the navigation problems. The mud accumulates gradually, and the robot becomes more and more blunt.
Problems with room map and navigation
Modern models using lidar (LDS) or visual navigation (VSLAM) build a detailed map of an apartment. If a map is lost or artifacts appear in it, the robot no longer knows where it is, a phenomenon called "loss of localization." In the app, you can see that the device is displayed in the center of the room or behind the wall, although it is physically located elsewhere.
Often, the problem is when you rearrange the furniture, and if you move the sofa or the table, and the robot is trying to drive the old route, it may run into an obstacle and not be able to rebuild the path to the base, and in these cases, the path-building algorithm fails, and the device starts to move chaoticly or just stops.
The solution might be to force the map to be updated, and try to start cleaning again, removing wires and small items that could have disrupted navigation from the floor, and if the problem persists, you may have to remove the old map and create a new one.
How to Recreate the Map in Mi Home App?
It is important to understand the difference between gyroscope models and lidar models. Gyroscope models (often cheaper) are less orientated and more often lost if accidentally kicked while working. Lidar models are more accurate but sensitive to transparent objects (glass doors, mirrors to the floor) that they can't see.
Obstacles and features of the interior
The interior of your apartment can hide many robot traps. Black carpets, dark floors, and glossy surfaces are the enemies of optical sensors. A cliff sensor can perceive a black surface as a cliff and refuse to go in that direction, blocking the path to the base.
It's also worth checking the area around the charging station. The infrared transceiver requires free space to work properly. If the base is in a niche, between the legs of the chairs or next to a dark object, the robot may simply not see the docking signal.
Here is a list of common obstacles that prevent returns:
- 🚧 Hanging curtains or wires that a robot can hook with a wheel.
- 🪞 Mirrors in the floor, creating a false reflection of space.
- ⚫ Black carpets with high pile perceived as a cliff.
- 🚪 Burns more than 2 cm high that the device cannot overcome.
💡
The ideal place for a base is in an open space against a wall, with a free radius of 0.5 meters left and right, and 1.5 meters in front. Don't put the base in deep niches.
If the base is on a long-pile carpet, it can stagger on contact, which causes the contacts to separate and the robot to leave, glue the base to the floor or put a solid base (linoleum, plywood) under it.
Software failures and software updates
Sometimes it's not hardware, it's software. There may be bugs in the firmware that cause the navigation module to freeze. Check the version of the software in the Mi Home app. If you have an update, be sure to install it. Developers often release patches that improve the algorithms for returning to the base.
A Wi-Fi failure can also indirectly affect work, and although a robot can clean up without the Internet, losing connection while building a map or sending telemetry sometimes leads to logic errors. Try restarting the router and the robot itself.
To reset the program state, perform a full reboot:
- Turn off the robot power button on the body.
- Wait 10-15 seconds.
- Turn the device on again.
- Run the “Find My Robot” command in the app to check the connection.
| Symptoms. | Possible cause | Action. |
|---|---|---|
| He's at the base, but he's not up. | Dirty contacts or base shifted | Wipe the pins, fix the base. |
| It's spinning in one place. | A bumper or lidar sticker | Check bumper mobility |
| Stops and stays silent. | Battery or software failure | Put it on manual charging. |
| Squeaking and blinking | Fall sensor error | Wipe down the bottom sensors. |
⚠️ Warning: If the robot makes strange noises when trying to get to the base (screaming, knocking), immediately stop trying.
Failure of the battery
If your device is several years old, the problem may be the wear and tear of the battery. Over time, the capacity drops, and the voltage can drop dramatically under the load. The robot “thinks” that there is still a lot of charge, but when you try to get to the base (which requires the engines to work), the voltage drops below critical, and the device turns off or goes into sleep mode.
You can check the battery status in the app if the model supports this feature, or by paying attention to the time of operation. If the robot used to remove 100 minutes, and now barely reaches 40, it's time to change the battery.
In some cases, battery calibration helps: completely discharge the robot (until it turns off itself), and then charge continuously for 10-12 hours, which can help the power controller reassess the actual capacity of the cells.
☑️ Diagnostics of the battery
Mechanical damage to wheels and suspension
Finally, it's worth checking the chassis. If one of the driving wheels is wedging or rotating with force, the robot will move in an arc, constantly moving from a straight trajectory to the base, and the hair wound on the axis of the wheel is the classic reason for this behavior.
Turn the device over and spin each wheel with your hand. They should rotate easily and with the same force. If one wheel is tight, disassemble it (usually you need to unscrew one screw) and clean the hair and dust.
Also check the side brush. If it has wrapped a lot of hair on it, it creates additional resistance and can lead the body aside, preventing accurate positioning at the base.
💡
90% of the problems with returning to base are solved by simply cleaning the sensors, contacts and wheels from dust and hair.
In conclusion, if none of the above methods worked, and the robot continues to ignore the base, it may have failed the control board or the IR receiver itself, in which case you should contact an authorized service center for professional diagnostics.