Motion Sensor false positives

Just installed monitoring system. Motion sensor was working fine when system is was not armed but when I set it in the “away” mode yesterday for the first time, it senses motion that is not there. It starts the cam in that area which confirms no motion. Sensor is 7 feet up as directed. This is not from a shadow, insect etc because it happens exactly every two hours. Yesterday’s alerts were exactly at 1:00, 3:00, 5:00, 7:00, 9:00, 11:00. I’m now 1800 miles away so support said they cannot troubleshoot. What would that be anyway, delete, remove/reinstall battery and re-add the device? At least I know Noonlight works. They called and texted but the “safe word” prevented a call to the police. Seems like only option is to delete this device from home monitoring. Disappointed.

Do you have an entry sensor mounted near your motion sensor?

I believe this is being caused by EMI from that entry sensor periodically updating its status, which happens exactly every 2 hours. At least it was in my case.

So until Wyze figures out a better solution, people shouldn’t mount the v2 motion and entry sensors within a foot of each other. I know that is a particular sore spot in your case, as you are 1800 miles away!

Sorry about that.

3 Likes

Indeed it is close! Probably about a foot. It sure sounds like you have the answer. I wish there had been info about this with the directions or when I spoke to customer service yesterday. I guess I will just need to delete the motion sensor for now until I am back to this property in 6 months. Thanks a million for your response!

1 Like

Wow, this is really interesting! I will have to keep that in mind when my extra sensors come in. I am sure we will hear a lot more about this in the future from people with lots of sensors saying they’re getting false alerts.

How could Wyze resolve this? Would it be possible to program it into the algorithm to filter out alerts that are triggered exactly 2 hours from the previous sensor update unless that motion is detected continuously for longer than the amount of time it normally takes for a contact sensor to update if there was a previous update exactly 2 hrs ago?

Also, I wonder if you could shield it just enough somehow.

I might run some tests with this when my extra sensors come in. @spamoni4 would make an excellent tester for helping figure out some of this too if he hasn’t already locked all his sensors in place.

2 Likes

This product is very new, which is why support probably didn’t have an answer ready to go. I made the project aware of the issue, and asked that they relay the information to support.

@carverofchoice – I also wonder if a footprint can be identified that would allow the developers to filter this in software. Your idea of filtering any alert that is exactly 2 hours from a previous alert that isn’t over a certain duration is a good one. I wonder also if there is some other internal thing they can do with the motion sensor firmware on all alerts that might be able to filter ANY spurious signal locally. Maybe by looking at the motion sensor component itself to see if it actually saw motion before letting an alert be transmitted. The trick there is there probably isn’t an internal timekeeping clock, so they would probably have to do something like set a memory location when motion is seen, and clear that memory location some number of raw clock cycles later.

My testing indicated the motion sensor was more sensitive when the spurious transmission source was above it than it was when it was below it, probably because the motion sensor’s antenna is likely at the top (I haven’t disassembled one). I also tested on a non-metallic tabletop, but I wonder if mounting the entry sensor on something like a metal door would amplify the transmission and thus require more separation. If I get some time, I may test that!

3 Likes

The antenna is indeed on the top of this PIR but PIRs sensors are notorious for poor RF immunity.

3 Likes