The hang on presence is a problem regardless of Home Assistant. Mine was problematic before and after I got my HA Green.
I’m also experiencing the FP300 being stuck on presence true.
Using Home Assistant and ZigBee with Z2MQTT.
I moved the sensor to various rooms in the house and eventually see the same thing, permanently stuck on presence until the fp300 is restarted, at which point it correctly identifies no presence. The at some point over the next few days it correctly detects presence and then stays stuck.
I’m going to swap the FP300 for another one and see if he issue remains.
But right now he only way I can get the FP300 to work is to regularly send a restart command from home assistant.
Disappointed with the FP300 so far.
after raising a support ticket I was told
“However, we need to explain that our Aqara Zigbee devices do not support being added to the HA platform through direct connection methods such as 2mqtt or ZHA. They need to be used with an Aqara hub”
so I have to buy another product to fix the misbehaving products I have already bought!
I have similar issue, it usually keeps Presense for 2 hours even no one in the room
I tried to change angle, restart, sesitive level, but still not work.
REALLY BAD PRODUCT!!!
I have tested the FP300 with the current thread firmware 1.0.7.3 with the following matter ecosystems and had the same issue with it stuck on occupancy, Apple Home, Aqara (M100), SmartThings and Home Assistant.
Matter ecosystems which are not Aqara should support firmware updates.
The tin foil method did not cause the FP300 to go back to absence.
I have heard nothing from Aqara about a fix.
……
If you have the M100 you’re just making yourself miserable. Switch it to zigbee.
I can’t believe they just launched thread only devices given the issues with just recently released products.
None of the Aqara products I owned would operate well in thread. T2’s, P2, FP300. None of them were useful in thread mode.
I don’t agree with your opinion.
I have a U200 working fine with thread!
reply from support 7/1/26
Any news on update?
We’re sorry, but the new firmware has not been released yet. It is still under testing, and we will inform you as soon as there is any progress.
We sincerely apologize for the inconvenience this has caused you.
Sincerely,
Isabel
Aqara Technical Support
Because the tinfoil method would work with a classical PIR sensor, but of course not with a mmWave sensor…
I switched to zigbee. Sensor is situated in livingroom, sensitivity medium. No one is in the room but every minute the sensor reports infrared detection. After exactly 1 minute it is switched off because no person is detected. Really annoying. Presence detection is stuck.
I’ve managed to update my firmware from 5234 to 5841 by swapping to Thread and then back to Zigbee in the Aqara app. Z2MQTT now showing a the new firmware version. am really hoping that battery level will update in time as they are only about 90 days old.
New thread firmware 1.1.0.1.
In the Aqara app switch to zigbee and then back to thread. Or wait for the firmware through the Aqara app.
Thanks to @andreas_roedl who found this.
I heard nothing from Aqara.
I’m on Zigbee and its in the living room with a ceiling fan. Which obviously is going to cause interference. Now there’s no way to fix this even with the AI learning and calibrating the sensor distance or placement. I guess this sensor is only limited to a few places such as bathroom or hallways.
now two out of my four FP300 (Zigbee) get stuck on presence true, even when moved to other rooms.
With thread firmware 1.1.0.1, standard settings, low sensitivity, 10 seconds absence delay, I now have absence working AFTER 22 minutes!
I will leave it over the weekend to see if it ‘learns’ and improves.
It didn’t. Stuck on presence again so did a factory reset.
Now I have the FP300 connected through SmartThings, Apple Home and Aqara Home.
It detects Presence but only goes to Absence after 1 hour (see image).
I have read this forum and others and kept giving it a chance, unfortunately I keep seeing too many ghosts and the presence sensor is not accurate. I am returning my sensor, but I hope Aqara can fix this because it’s actually a nice product… if it works.
Given your configuration — low sensitivity with a 10-second absence delay, the sensor still takes as long as one hour to switch to the “absence” state.
My guess is that when someone enters the space and turns on a certain device, the radar identifies that device as an interference source.
Specifically, the sensor has already learned and adapted to an interference-free environment. Once someone enters and triggers a new scenario (or introduces an additional interference source), the sensor will struggle to revert to the “absence” status.
You can verify the following: Will devices like exhaust fans be turned on when someone enters the bathroom? If you manually turn off the exhaust fan, can the sensor then correctly switch to the “absence” state?
We look forward to your further feedback. Thank you.


