Zwave device moiton stuck on

Using: docker Zwavejs2mqtt with a ZOOZ S2 USB Stick Controller.

I’ve been having a problem with several types of zwave devices getting stuck with motion ‘detected’. They will stay that way for hours until they detect motion again. My network contains about 100 devices. Getting stuck seems to be random across several different types of devices. Using the “refresh command” doesn’t seem to fix the problem.

Any ideas how I could begin to diagnosis and fix this problem? It’s hard because it’s random and not reproducible. Maybe the network isn’t healing right?

From my limited knowledge of Z-Wave, your battery devices are supposed to work in 3 ways:

  1. Detect motion, sent report of motion.
  2. Initiate countdown of n-second, as defined in their configuration. Zooz goes with 15 seconds by default. Their output then depends on how the countdown resolves.
    a. Was motion detected within the time frame? If yes, restart countdown, don’t send report (status remains the same, i.e. “active”).
    b. If no motion was detected, stop countdown, activate Z-Wave radio, proceed to step 3; meanwhile, do computery things like looping.
  3. Report change in status being no motion (i.e. cleared), go back to step 1.

They do this because otherwise, every motion sensor on your network would send a report every 15 seconds, then there’d be an ACK report too, and possibly additional transmissions with group binding. Do you really need to tell a lightswitch to turn on every 15 seconds, even if it’s already on? More traffic!

I’m thinking your problem lies with some of your devices not properly updating their status. You might want to look at their configuration settings and see if your “motion timeout” window is abnormally long.
When I first paired some of my motion detectors for example, their awake timer defaulted to 0 seconds; endlessly pinging my controller, flooding the network, preventing any other traffic.
Once I forced it back to default (12 hours) I’ve never had a similar problem.

HI Travis,

Did you ever resolve this issue? I have the same problem on my Zwave network and its a similar size to yours - around 140 devices (about half powered and half battery).

I dont think its the result of short reporting intervals for battery devices as I have tried extending them for some and hasnt made a difference. I thought the communication traffic on my network might have been too busy, so I spent a lot of time reducing the amount of traffic on the network by turning off energy monitoring and unnecessary reporting. I have also updated the firmware to all devices where possible (a tedious process!).

I have noticed that I have a few powered devices with TX/RX errors and Timeouts reported so I wonder if this is related to the motion sensors being “stuck on”. I’m thinking the battery devices try to report their status but the powered devices receive but then dont route the signal forward.

There isnt alot of information available for troubleshooting and optimising a Zwave JS network. I would love to read more about this. Any ideas anyone?

I’d look at the disabled entities for the device in the device configuration panel, if you see one related to motion enable it and restart HA. I have a couple of detectors with issues and by enabling the disable entity they are working fine

Thanks Pete. I’ll look into this.