with click 3 times. The device is not far from my Aqara M3 hub or nearest Apple Home hub (In. this case an Apple TV that does Thread). I don’t know what to say. I had to try the “click 3 times” at least 6 times to get it to work.
What more information can I give?
Hello, when you quickly click 3 times to calibrate normally, there will be a flashing light effect of 2s. You can check whether the light effect has been triggered. In addition, it is also possible to judge that there is a delay in network and APP display. You can try to exit the APP device page and re-enter to see if the status has been updated in real time. Thank you for your information.
I’m aware that there should be a flashing light. That did not happen.
If this is the case, the trigger may not be successful. Please make sure to complete 3 quick clicks within 2s to see if the calibration can be triggered successfully. If it doesn’t work, it may be that the keys are in poor contact (normally, a click will have a flashing light effect, so you can test it).
I have a refrigerator where one of the doors doesn’t close completely if you’re not careful.
So I placed the sensor on the door I want to monitor with a scenario that sends a notification to my cell phone and also an announcement on Google Nest Mini to signal that the door is not closed properly.
I first tried placing the sensor on top of the door so that it would be less visible (i.e., horizontally), but unfortunately, sometimes the sensor would send the wrong status.
So I placed it vertically as recommended, and I no longer have any problems.
Received mine the other day as well. Using it as window sensor and placed it right above the handle of the window. I like the black color option as we have dark wooden window frames it blends in better than the white door/window sensors I used before.
It works perfectly fine on my end and I love how it recognizes the tilted or swinged open window. I had to recalibrate the sensor once but it’s working without any issues since.
Looking forward what Zigbee2MQTT will offer in the future but for now it’s one of my favorite sensors so far! 
Thank you for your feedback. Your recognition has encouraged our team.
Yes, please place the equipment vertically. This is the recommended installation method.
Regarding the installation in the refrigerator, if the result is not accurate, you can complete the calibration in the closed state once and adjust the sensitivity.
I have my P100 installed on my office chair, which helps to keep the lights on versus relying just on a PIR motion sensor. I have it paired via Zigbee to my Aqara M3 Hub, and then via Matter to Home Assistant. On the Home Assistant side, I had to rename the entities to properly reflect each of the P100’s detection capabilities.
Thank you for sharing. It can be said that I have seen the best guiding practice of HA at present. You are welcome to post specifically to share your detailed configuration, which can help more users with the same needs as you.
AQARA P100 vs. AQARA T1: One-Week Testing & Comparison Report
Tested Hardware:
Sensor P100 Firmware: 0.0.0_0021
Hub M3 Firmware: 4.5.40_0019
Executive Summary
The difference between these sensors is generational. While the T1 is a specialized vibration detector, the P100 (AI Multi-State Sensor) is a multi-functional spatial sensor designed to replace several types of devices at once. After a week of testing, the P100 feels significantly more robust and offers a far wider range of capabilities.
Comparison Table
| Feature | Aqara Vibration Sensor T1 (VB-S01D) | Aqara Multi-State Sensor P100 |
|---|---|---|
| Primary Technology | 3-axis Accelerometer | 9-axis System (Accel + Gyro + Compass) |
| Detection Capabilities | Vibration, Tilt, Drop | Vibration, Knocking, Angle, Door/Window Status |
| Protocol | Zigbee 3.0 | Dual: Zigbee 3.0 & Thread (Matter) |
| Battery | 2× CR2032 (~2 years) | 1× CR2450 (Higher capacity, 2–3 years in Zigbee mode) |
| IP Rating | IP20 (Indoor only) | IPX4 / IP20 (Splash resistant, indoor use) |
| Dimensions | 36 × 36 × 9 mm (Ultra-thin) | Approx. 42 × 42 × 15 mm (Bulkier) |
| Installation | Adhesive (3M) | Adhesive (No magnet/contact needed) |
Detailed Feature Breakdown
1. Battery and Longevity
T1: Uses two CR2032 batteries. Reliability is proven, but high sensitivity settings often drain the battery in less than 2 years.
P100: Features a larger CR2450 battery. Although the device is thicker, it offers much higher capacity. Thanks to the efficient Thread protocol and modern chip, it promises a lifespan of over 2-3 years even with active AI processing.
2. Durability and IP Rating
T1: No official water protection (IP20). Vulnerable to humidity in bathrooms or condensation in mailboxes.
P100: Improved build quality with better sealing. While not submersible, it handles ambient humidity better. However, it is still primarily an indoor device.
3. AI Capabilities & Spatial Awareness (The P100 Edge)
The most significant advantage of the P100 is its ability to learn and calibrate. Unlike the T1, which only reports movement, the P100 understands specific states via the Aqara app:
Closed: Saves the 0° spatial position.
Tilt/Ventilation: Recognizes a specific angle (e.g., 15°).
Fully Open: Recognizes the 90° position.
4. Build & Ergonomics
The P100 casing feels much more durable. A major improvement is the opening mechanism, which can be operated with two fingers without needing a screwdriver, unlike the T1.
User Experience & Testing Notes
The Positives (+):
The housing feels much more rugged and premium.
Significantly improved battery life expectations.
Excellent tool-less opening mechanism.
Areas for Improvement / Limitations:
Compatibility: I was only able to pair the P100 with the Hub M3. It failed to connect to the S1 Hub or G410.
Stability: During the first week, I experienced occasional disconnections from the M3 Hub where the device became unresponsive.
Environmental Protection: The device lacks IP67 rating. It is still an indoor sensor.
Market Gap: There is still a need for a larger version with an even bigger battery and true outdoor protection for gates, fences, and exterior blinds exposed to heavy rain.
Hello,thanks for sharing. I don’t know if it is due to the integration of AI tools. Here I need to point out that some information in it is distorted. For example, P100 actually uses CR2450 battery, and the actual maximum battery life is 2-3 years in zigbee mode. The normal battery life needs to be based on the actual frequency and various factors for reference.
I’ve been enjoying the p100. I have it in my main entrance below my original Aqara door sensor. My only issue is that sometimes, could be the door is closed too hard, I have it stay in the open status. That’s not too bad when you’re home but when I’m away I have to double check via camera to be sure it’s closed. Perhaps adding a way to reset status without having to physically be there(or disable vibrate) would help these occasions but otherwise solid on my end.
I’m looking at the specs and the feedback here, and one thing doesn’t add up. If the P100 truly has a 9-axis setup (accelerometer, gyro, and magnetometer), why is there a strict requirement to install it vertically?
Technically, a 9-axis sensor has all the hardware needed to calculate its own orientation regardless of the starting position. All it takes is a simple software calibration (offset) to set a “zero point” after installation. By forcing a vertical orientation, you’re basically ignoring the potential of the hardware.
Is the “vertical only” rule a hard limitation of the firmware/algorithm, or is it just a recommendation for the average user? It seems like a waste of a high-end sensor if we can’t calibrate it to work in whatever orientation the mounting surface requires.
In fact, this is not a very strict limitation. Some users have indeed encountered problems when placing the device horizontally, while there were no issues when it was placed vertically (we speculate that this might be due to magnetic components inside the sensor, such as the battery, interfering with the relative angle between the device’s orientation and the geomagnetic field, so it could indeed be related to the device’s posture. Based on our tests, this is not an issue that always occurs). For users who have stricter requirements regarding installation location or aesthetics, they can also try placing it horizontally first or reducing the sensitivity settings to see the effect.
Thank you for the explanation. However, the “magnetic interference from internal components” argument seems like a secondary issue that could be solved by a robust calibration routine.
A 9-axis sensor is designed specifically to handle orientation via a rotation matrix. Even if the battery or internal components create a constant magnetic offset (Hard Iron distortion), this is a static value. A standard calibration procedure where the user sets the “Closed” and “Open” positions should theoretically allow the algorithm to nullify those offsets and remap the axes, regardless of whether the device is mounted vertically, horizontally, or at an angle.
This leads me to believe the “vertical only” rule is a hardcoded mapping in the firmware. It appears the algorithm is optimized to look for gravity and magnetic flux on specific pre-defined axes.
Currently, for the door sensor function, users need to press the button three times while the door is closed to calibrate and confirm the door’s closed posture information. In theory, the device can be defined as closed in any posture.The posture detection function also uses this information to define the default state. Specifically, the standard posture closest to the calibration position (horizontal upward, horizontal downward, vertical, etc.) is defined as the standard for determining tilt.
As for the standard geomagnetic calibration process, the greatest hard magnetic interference comes from that button battery, and once the user replaces the battery after leaving the factory, the calibration result will become invalid. We also don’t really want users to go through more complex geomagnetic calibration steps. Perhaps in the future, with batteries that contain no magnetic materials, this type of door sensor solution could achieve higher accuracy by calibration 
