The most frustrating part of my setup is the constant flux / half way support of Z-Wave in Home Assistant.
I’m constantly getting stuck with devices that aren’t supported or devices where only 2 of 10 features are supported.
In addition, this constant bouncing between hacks and open source solutions that are in constant flux to get Z-wave on Home Assistant is adding to that aggravation.
I’m fine with Home Assistant being Open Source, but Z-wave support for it shouldn’t be.
So what is the best way to do get a complete Z-wave solution with Home Assistant?
I used to run HomeSeer and the z-wave functionality was excellent. That said the HomeSeer HS4 upgrade was botched in my opinion. I’m now on Home Assistant and it is far better in general. You might consider HomeSeer for z-wave and integrate with Home Assistant over MQTT.
Are you using the new zwave js integration? The ozw stagnation, ozw beta irrational exuberance debacle was pretty bad, but it seems like zwave js has some real weight behind it from both the home assistant side and zwave js project itself’s side.
The inability to set config parameters and no gui for it was a big deal to me but those have since been added.
In my personal experience with ~15 devices, the zwave js integration has better hardware support and is much faster than the deprecated integration did. E.g. My inovelli fan/light controller now properly comes through as a fan and a light without any template shenanigans.
I constantly run into devices that either aren’t recognized, aren’t really compatible or are partially working.
There’s enough hacking/configuration need with home assistant already.
Zwave support is foundational to a smart home / home automation platform and should be official and legit, not a cobbled together constant mess.
Other hubs like Homeseer and Hubitat seem to have “Official” Zwave support.
And I recognize that there’s probably a cost here, but pass that cost on to us as users if we want it. Make access to an Official Zwave Support integration part of Home Assistant Cloud subscription.
When I upgraded to 2021.4, it went from being a light (that I called light.actually_fan and passed into a template), to being a proper fan entity automatically.
I may have followed these steps when I migrated to z-wave js, but I don’t think I had to (I can’t recall, it was a few weeks ago after all ). I’m not sure if that’s because I did them a year or two ago with the deprecated integration and the files persisted, or if zwave js had native support.
I’m not really up on the state of various z-wave implementations, just wanted to make sure that you were using the alleged latest and greatest before you went off on some big hardware change.
Fibaro Lite. 90,- Euro’s. Comes with integration, does it all, been using that now and before that had a Fibaro Home center 2 (more expensive, no longer required now) and been using that for more then a year without a single hickup.
Interesting, thanks. It might be worth spending the 90€ for trying it out.
Even though zwave js is certainly an important step forward, I’d rather get rid of everything open-sourcey in that area. ZWave is just too important in my home and sadly it has a history of being super flaky in HA. There’s too much emphasis on adding new features and niche use cases rather than stabilizing and making the existing base more compatible with commonly used devices.
I used to have only fibaro home center 2, then i got introduced to home assistant a long time ago. I moved all my zwave to it (openzwave) but it was never as good as the fibaro experience. Now i am happy i combined both. Every person i spoke that has taken this approach has not given me any negative feedback.
Are we talking about the Fibaro Home Center Lite or the Fibaro Home Center 3 Lite ? For 700-series support you need the latter, apparently. I hate it when companies name their products in such confusing ways.
Anyway, I searched around a little and I’m not convinced this is a good replacement for zwavejs. It doesn’t seem to be a very popular device, at least in combination with HA, and most HA related content I could find about it were people having issues. The HA integration for it depends on the fiblary Python library. Which is, as so often, a single-maintainer project. Even though it is much simpler than a full zwave stack, you’re again putting your entire zwave setup into the hands of a single person, who could go missing from one day to the next (see what happened with OZW). That’s really what I was trying to avoid.
But it’s good to know that it works well technically.
No, dont use the home center 3 lite, since it has limited support on devices. A fibaro home center lite (2) has 255 or so.
p.s. i have not found a single solution that works better, and i am dating back from the time that homeseer with x10 was used in my house. If you have strange zwave devices on which fibaro has no support, maybe. So don’t get strange devices, if you get ali-like devices for 5 bucks don’t expect it to work fully.
There are no ‘strange’ zwave devices in the way of cheap knockoff Zigbee or Wifi stuff, they are all using the same licensed Silabs chip. That’s one of zwave’s huge advantages. And I’m not doubting the quality of the Fibaro zwave implementation. Fibaro is one of the big names in zwave and I have quite a few of their sensors. Their FGMS-001 is the best motion sensor I have ever used. No, the issue I am seeing is the integration into HA, which seems very fragile and has a bus factor of 1.
As far as i see it, that library only gives support to talk to the rest api of fibaro. The actual support for importing lights/scenes/climate etc is build into home assistant.
Not trying to convince you, but like i said i had more then a year this setup without issues.
Alternative, have not tried this: Vera - Home Assistant same idea as the fibaro. Although the added benefit is having a nice ready made iphone/android/tablet app to control something is a nice to have (wife prefers that over HA interface)
). I’m not sure if that’s because I did them a year or two ago with the deprecated integration and the files persisted, or if zwave js had native support.