Wyze Lock Open Door Detection Flawed - How to Resolve

So after 30 minutes having a Wyze Wizard blame my app and my phone and never really addressing the problem. I started to do some experiments.

Tell me if you’ve heard this one. Wyze Lock and the Gateway will detect if your door is open (or closed). Most other manufacturers use a magnetic sensor (similar to a contact sensor for Wyze Sense), but not Wyze, they have a proprietary ‘sensor’ that they use (more on that in a minute). Many people have had this problem (there are multiple Wyze Forum & Reddit threads on this topic).

The Wyze Lock constantly or intermittently reports the door open, even when it is closed and locked. If you have the alarm on, it goes off. Sometimes recalibration works, sometimes it does not. Sometimes recalibration works but then fails again in a couple days. Wyze had some firmware update that was supposed to address it, but it evidently hasn’t. I couldn’t get my lock to calibrate other than the first attempt and I think that was a fluke. It went bad after about 10 minutes.

Now for that “sensor”. After experimentation, I believe that this sensor is in the gateway and it is a directional sensor (NOT, I repeat not a true positional sensor - ie, doesn’t understand distance, just position). Why do I think that. Well, the outlet that I originally plugged the gateway into is near where the knob/lock is when the door is in a 90 degree open position. It takes a bit of geometry to explain what I think is happening here, but if you draw a line from the closed door lock position to the sensor, the open door position of the lock does not stray very far from the line. In other words, the sensor in the gateway never recognizes that the lock has changed positions. It is my belief that the software does not understand these minute differences.

My evidence. I moved the gateway to an outlet on the same wall as the door, which means that the open and closed position of the door are about 10-15 degrees difference (as measured from the gateway). Bing bang boom, bob’s your uncle, the lock works and it understand a door open vs a door closed.

What this means is that when you position your gateway you MUST pay attention to the line of sight direction of the lock to the gateway and ensure that there is a substantial difference. Why Wyze didn’t go with a more definitive open/close sensor is beyond me. I would acuse them of trying to be too cute with their solution. They apparently didn’t think it through. All is well again though, but I could image for some, this won’t be such a great outcome.

Just my two cents (and if someone already figured this out, good on ya).

1 Like

@pthalodezin
The sensor used for the door position is a gyroscopic sensor like what is in a cell phone, I do not believe it is proprietary but I could be wrong on that part.

https://forums.wyze.com/t/wyze-lock-calibration-reset/83681/83

That doesn’t add up. If you buy that, fine, but think about it. An open close sensor needs to register two distinct locations. If you think that this sensor is monitoring position 100% of the time… Likely the gyroscope is sending motion, not position. The trigger of motion is then sent to a sensor which detects which of two positions the door is closest, open or closed. The idea of the gyroscope wouldn’t change when I relocated the gateway, but yet, that was the result. I still have issues with auto lock being triggered if the door is ajar. Pointing once again to wyze’s poor decision in open close sensors. A physical contact sensor would be the right choice

@JonathanF can verify that it is a gyroscopic sensor they use. I believe he was the PM of the Wyze Lock. The sensor is not in the Gateway.

The gyroscope makes sense. If you have the open/closed feature turned on without auto lock it says the battery life will suffer I assume since it keeps the gyroscope running. I have 1 bridge connected to 4 locks in a 3k sq foot house with all different orientations to the bridge. Are you running more than 1 bridge by chance?

This sounds like a problem with interference to me. As @s13turbo asked, do you have more than one bridge in your home?

Nope. Single lock, single gateway. When changing the gateway location, errors resolved. There appears to be a misconception here on how a gyroscopic sensor works and what it can sense. A gyroscopic sensor measures variations in angular velocity. As such, they can be used to determine either orientation or movement. You are correct in assuming that a sensor in the lock could determine if the lock is moving, but since the orientation of the sensor to the door and the lock is a constant, there can be no measure of orientation unless, … Wait for it, there is also a sensing component in the gateway. I am not arguing that there isn’t a sensor in the lock, but all evidence points to a secondary method for determining position in the gateway.