Notification if something is wrong with a z-wave device?

Configuration Z-Wave
1

What are people here using for notification if something wrong with a z-wave device?
I use an automation if time since last update on a device is over a certain time-value.
But that does not work on a siren I have (Nexa) that is always on, since it doesn`t give a new last updated value. The same goes for my Aeotec usb stick controller.
Does anyone here have a solution for this problem?

2

Interesting, this would be useful. Could you share your current solution?

3

I’m using the following to detect when any device hasn’t sent an update in more than 10 minutes, and if so run a test then a heal. Just add as many duplicates of the BETA line as you need. The ALPHA and OMEGA lines are different, so if you’ve only one device it’ll need some tweaking.

alias: 'ZWave check'
initial_state: 'on'
trigger:
  platform: time
  minutes: '/5'
  seconds: 00
condition:
  condition: template
  value_template: > {{ ((as_timestamp(now()) | int) - [ as_timestamp(strptime(states.zwave.ALPHA.attributes.receivedTS | string | truncate(19,True,'',0),'%Y-%m-%d %H:%M:%S:%f')), 
        as_timestamp(strptime(states.zwave.BETA.attributes.receivedTS | string | truncate(19,True,'',0),'%Y-%m-%d %H:%M:%S:%f')), 
        ...
        as_timestamp(strptime(states.zwave.OMEGA.attributes.receivedTS | string | truncate(19,True,'',0),'%Y-%m-%d %H:%M:%S:%f')) ] | max | int) > 600 }}

	  action:
  - service: notify.WHOEVER
    data:
      message: "Z-Wave network has stalled, testing and healing"
  - service: zwave.test_network
  - delay: 00:01:00
  - service: zwave.heal_network
2 Likes
4

Thank you for your reply.
If I understand your automation correctly, it does the same as mine, just a different approch.
I have one automation per unit, so I easily can get a notification which device it is something wrong with.

- alias: timestamp fire detector kitchen
  trigger:
    platform: time
    minutes: '/60'
    seconds: 00
  condition: 
    condition:  template
    value_template: '{{ (as_timestamp(now())-as_timestamp(states.zwave.brannvarsler_kjokken_5.last_updated)) > 21600 }}'
  action:
    service: notify.pushover_einar
    data_template:
      message: 'Something wrong with fire detector kitchen'

I have a battery operated Fibaro fire detector, but even though I have a wake up interval to 1800, it does not get a new update every 1800 seconds, therefor I have so many seconds before I say it fails. I too have a fibaro water detector, and that updates every 1800 seconds. Why it`s different, I do not know.
But as mentioned earlier my siren and aeotec zwave controller does not get a last updated value (just when homeassistant restarts, but that is not a real last updated value). Is there any ways of detecting if they fails?

I do not have a test and heal network service, that could be smart to use.
But the test network service, what does it do? Could you get any output if the test fails?

1 Like
5

See here for details on what test and heal do.

I use the received timestamp field, because that always seems to have an update. Worth checking your devices to see if there are any other timestamp fields that are reliable.

6

Thank you. Sadly receivedTS is the same value as last updated. Sent TS is the same or older.

7

The problem with the battery powered devices is now fixed. The problem was that I have used the config parameters in hass UI. Seems like it just adjusted the values in the zwave config file, but did not actually change anything in the unit. So I changed to use open zwave control panel, and now the wake up interval is correct.
Still no change for the siren yet thou.