They are, but also Z-Wave has gotten a lot better over the years and keeps getting better.
But the silver lining is that once you weed out the bad apples you can enjoy stability for years to come. And when one device goes bad, you know where it is, replace, and move on. Hard to get started and reach stability, but easier to maintain in the long run.
something else to maybe check and make sure they used the screw terminals properly - they are not like regular traditional switches ...this might be obvious to some but it wasn't to me at first 
Well, I feel like I'm wasting my time here. Maybe Hubitat Z-Wave groups just don't work as well as I hoped.
Replaced 3 Zooz dimmers that seemed like they were the most persnickety of the bunch. But brand new devices installed at those same locations did not perform any better.
Refocused my attention on the remaining "good" devices. Maybe they aren't so good, after all. Maybe they just behaved a little better when the overall deployment was smaller.
My test case: A tiny Hubitat group of a single Hue Bulb + a single Zooz Z-Wave dimmer. I can sit here and toggle the group rapidly (via hub). No matter how fast I go, the Hue bulbs keep up. The Z-Wave device keeps up at first, but after a while I goes bonkers -- stops responding, lights toggle erratically, logs are flooded with events over the next 30 seconds that don't correspond with any real (observed) changes. It's like an overloaded buffer or something.
The Hue stays responsive and reports accurate events the whole time while this is happening.
I can single out any of my ("good") Z-Wave devices in this manner and it's the same result. Expanding a the group to include several Z-Wave devices just exacerbates the problem -- it "overloads" the system almost immediately.
Maybe I'm expecting too much of Z-Wave lighting here. Maybe the Hubitat group app is too flaky and can't operate Z-Wave devicds reliably. I don't know. But I'm all out of ideas.
What does your mesh look like. So many people don't have issues like this that run exclusively z-wave. Post your z-wave details page in its entirety for us to look at
Is this a common use case in your household?
Same model, or different? As mentioned a few weeks ago, it might just be the device is unable to handle a flurry of consecutive commands as the Hue bulb. If it was the app, you'd see the same behavior on both the bulb and the Z-Wave device.
Does your mesh look as good as it was before you reset it, a few weeks ago?
I'm so tired of this response.
It's illustrative of the problem. When group operations are issued rapidly in my environment, the system seizes up and becomes unstable. The larger the group, the faster the system fails. But even a "group" of a single device can make the system unstable.
And yes, it is a common use case in my household, because users of physical lighting controls expect near-instantaneous response times. When someone presses a button and the lights don't respond instantly, they assume they mis-pressed and will try again. With a modest group of five Z-Wave devices, those two button presses can be enough to overload the system. So it means I cannot use my scene controller for group lighting actions.
I've been using Zooz devices this whole time. I selected them because I've heard good things, and for usability/aesthetics we like that they resemble a conventional paddle switch.
I do wonder if there's something about the Hubitat group app in particular that's flaky. i can operate these Zooz devices from the hub directly (via the device view) and issue rapid on/off commands and the system seems stable. But when I establish a "group" consisting of just that single device, on/off commands of the group device is unreliable.
For context, before diving into Zooz/Hubitat, our primary smart lighting solution was Lutron Caseta. It had stellar performance: Instant response times, excellent dimming range. We moved to Hubitat/Zooz hoping to achieve similar performance but with a different aesthetic.
The absence of near instantaneous response times is an indication of an inefficient or poorly functional mesh. Repeating the process rapidly is going to make the situation much worse. Why you ask? Because z-wave radios are serial devices that can operate at speeds as low as 9.6 kbps. Repeated commands (and the associated noise) at that bandwidth increases the likelihood of gumming up the radio.
Fixing your z-wave mesh will eliminate this being a common use case in your household.
It is always possible that individual setups expose issues with apps. On the whole, over the last 3.5 years, I haven't experienced this particular issue.
It sounds like either the system is getting overloaded with your testing or the mesh is having communication issues (which are more noticeable with higher message rates).
Can you provide more details on how many commands you are sending over a fixed period of time? The groups app/device are event driven, and would add some (relatively small) additional processing (latency), so it would seem to reason that whatever the maximum throughput of the system is (for reliable and observable activity) using devices directly maybe lower when going through a group device (and, of course, there must be such a maximum as we don't have infinite resources anywhere in the system).
Also, can you provide a screenshot of the z-wave details page again (I don't think you have done this since you indicated that you reset the system), as well as identifying which device you used for this single device group test?
I'm starting to reach the conclusion that my Z-Wave mesh is not fixable, for whatever reason. I've rebuilt my mesh from scratch a couple of times now, with different mix of devices in different locations, and can't achieve the performance or stability I require from the system.
Maybe there's some RF interference in my environment that I don't know about. I wouldn't know how to diagnose such a thing. Or maybe Zooz devices aren't as good as I'd hoped, and are too susceptible to being overloaded or too slow to process Z-Wave commands or raise status events that are required for reliable Hubitat group operation. Or maybe my Hubitat (/Z-Wave radio) is just a lemon.
I don't know how to test any of those theories without throwing more time and money at it, and I've already sunk countless hours and $500. Might be time to cut losses here.
I've had much better success with Lutron Caseta and Hue lighting controls in my environment, which I believe operate on different RF spectrums. I'd go back to Caseta in a heatbeat if my wife didn't hate the control layouts so much. Or I'd go RA2, if it wasn't so damned expensive.
Difficult to say. If you wish to troubleshoot your z-wave mesh, I would second @tony.fleisher's suggestion to share a screenshot of your current z-wave details page.
One other suggestion. Try metering with ~100-200 ms in your z-wave groups.
Edit - one more suggestion. If you are handy with a solder iron, consider adding an external z-wave antenna.
Hah! My gut says that there's a good chance something like this would help in my case, but alas, no, I'm not handy with a solder iron, and that's a level of DIY beyond what I was prepared to bite off in the Hubitat system.
(I'm a software guy and am not intimidated by rolling up my sleeves a bit on the programming side of things; But didn't expect to be so stymied by the hardware / RF protocol side of things.)
I would agree. My z-wave mesh (~60 devices) was reasonably performant even without the antenna; I had no complaints at any rate. However, I have noticed that it is extremely performant with the antenna. Yet, ~50% of my devices are z-wave (not z-wave+), and I even have one 100-series z-wave outlet.
There are some community members who have kindly added antennas for others, like me, who are not handy with a soldering iron.
Thanks, I'll gather my Z-Wave details just for one more crack at making this work, but if I can't, I'll be looking to sell these Zooz devices and start over with something else. Innovelli Blue (ZigBee 3.0) might be worth a try.
I'm a zwave fan and have a good working stable mesh. But everyone's use case is different. We don't use motion Lighting or any "one button for many lights" type control though. Our smart light control is all via voice (Alexa) so we wouldn't notice lag others might.
Some people have good luck with ZigBee. Others are good with zwave. Some (like us) use both with what we consider success. Everyone has a different environment. We have both zooz and inovelli zwave switches and they're perfect for our needs.
That being said you can't go wrong with Lutron. You get what you pay for with RA2 Select (and now RA3). Yes it's relatively expensive but what's your WAF worth? We can't stand the physical UI on Caseta but we know some people love them (and if you're a motion lighting home it's less of a concern.)
This group is here to help support your integration to HE with whatever you might choose. It's not always easy and that's why some families choose a complete turn key integration like Control4 or Crestron. But if you're doing it yourself taking parts and pieces from different vendors, there will always be a pain curve. We're here to help minimize it!
Yup, I totally get this. Very different expectations/tolerance for responsiveness of voice triggers vs physical buttons.
I'd probably be fine with our Z-Wave performance if I was only doing voice control, but alas, we've got this fancy scene controller device that we were hoping to use for group lighting controls, and it just isn't working out.
Z-Wave mesh details. This was after (yet another) Z-Wave repair a bit ago, so I don't know how much data has been collected, but certainly plenty of red in that table.
Many of those signals are kind of low.
Can I ask how long your hub had been running before you took this screenshot? There are a lot of route changes, which is also an indication of a mesh that bears strengthening.
By way of comparison, my hub has been up for 3 days. One of devices has 9 route changes. Two have 3. The rest have 1 or 0.
Download the Hubitat app
