Reliability of Z-wave. Can it be used for serious tasks or just for fun?

[Kdiamond]

Hi.

I'm using Z-wave for 2 years now. have about 50 devices in my 3 floor house. The road so far has been very interesting but also stressful and certainly not set and forget job. The conclusion is I would never dare to install Z-wave to any non system admin person that has no ability to act instantly when **** happens. It only takes one device for fail.

Other thing is I have heating and power management done using HS. So failure of a system will lead to:
1. people freeze in winter days
2. people have saunas in their rooms
3. Peak protection (turning off non-priority devices) fail could overload house input and provider will cut of the supply.

Those are serious issues. Is the design of Z-wave up to the task even when everything set up properly. Does the protocol allow for that kind of reliability?

So the main question that comes to my mind are the following.

- Communication between device and host . Many times values are not updated properly or actions missed.Battery devices will send out signal on state change. Then is sleeps to save the battery lifetime. So what happens when the signal does not reach the host. Is there any bidirectional communication for host to approve the signal received? How can it happen quote often that HS devices are not updated with real state of the devices. Where is the catch? In the device, host, communication data???

- MESH
Mesh seems like a good option to extend the range, but is really OK? What happens is one device in MESH fails? Will that cut all the devices that is using it as a gateway. Would it be better to set multiple receivers (let's say one for each floor or how many it needs) and have it communicate

- ? Device Icon
Many times I get ? device icon. It always scares the **** out of me. Sometimes it corrects on next update, sometimes it does not.

So after 2 years of dealing with HS, I can really say it's a great platform, but really not reliable and makes me scared of using it in demanding operations. Needless to say Logs are great, Z-Seer+ test on all devices are 100%, but **** happens anyway.

So the question is. How to maximize Z-wave reliability for reliable serious tasks with bad consequences on failure, where status update of the device must never be missed? Is Z-wave even up to it?

Thank you
Br,
Dali

[ServiceXp]

IMO.. Nope.

Sent from my LM-G820 using Tapatalk

[Rupp]

My Z-Wave installation has been rock solid for years. This includes all lighting and 5 ceiling fans in my 2800 sq ft home. All switches are HS-Wx100's and HS-Wx200 switches. I wouldn't hesitate to recommend these switch and Z-Wave in general but only certain Z-Wave devices as many aren't of the same quality.

[Mark S.]

I have run a mix of Z-Wave and X-10 for several years - I would definitely say that Z-Wave is unreliable compared to ancient X-10 technology. But it seems that the Z-wave failures I see are related to the Z-wave controller and corruption of Z-wave related HS files - probably a few times per year. So I'm not sure the inherent technology is unreliable, but rather the Z-Stick and HS execution, and I'm sure the frailty of Windows plays a big role. A restore of a backup of my HS directory usually fixes my problems, but that has gotten old. I have never had to do that for X-10. Maybe I'm in the minority regarding my experiences, but because of all the failures I've had I will never trust any mission-critical task to Z-wave with HS.

I will watch this thread for suggestions on how to improve reliability.

[jvm]

I'd say that anything before Z-Wave Plus is a bit questionable, but newer devices (i.e., those supporting Z-Wave Plus, which includes most made / sold in the last several years) are reliable.

[simplextech]

It's not the protocol that has the problem it's the implementation and mixed matched stew starting from hardware all the way to software. The very few professional systems that provide z-wave only support an extremely limited set of devices from vetted vendors with specific firmware versions.

[dmiller]

Don't use HA/ z-wave for "reliable serious tasks with bad consequences". For example, change comfort presets on a thermostat, but don't directly manipulate temperature. I do have homeseer/z-wave control secondary HVAC devices like humidifiers, de-humidiers and fans where the consequences of a missed update are unimportant. But I leave the safety features of my thermostats intact. I also never have homeseer switch between modes on thermostats either (heat/cool/auto).

This morning I noticed z-wave failed to run a ceiling fan last night that I use to distribute humidity. But it didn't matter that one command was missed and the cycle continued normally in the following hour.

I also have temp and humidity exception reporting set up on secondary z-wave devices that have those sensors. I assume parts of the system will fail from time to time.

I have owned z-wave thermostats and I was not comfortable with the reliability. The mainstream wifi thermostats are safest IMO.

[ksum]

I have 60 z-wave devices. Yes, I have had to replace 2 of my Evolve brand light switches but otherwise it has been running with no issues. Been running z-wave for 7 or more years and never had a missed communication.

[sparkman]

My z-wave install consists of powered non-plus z-wave devices (Leviton switches) and then a mix of other devices some of which are plus and some which are not. The network is very reliable overall, and way more reliable than X-10 ever was and that was with phase bridges, filters, etc. I do get the occasional glitch where one device starts blocking the network, but a power off/on will fix that. Seen the same issues with battery powered X-10 devices that if the battery becomes low, they can start continually sending data, and thus flooding the network. Z-Wave thermostats have been very reliable (non-plus). They have a minimum temp set on them, so never an issue with temps getting too cold. I'll be at over 130 z-wave nodes later this week, with a few devices to convert still from other technologies. Once I'm done, I should be around 160 or so devices, all on one z-net.

[AllHailJ]

My z-wave network has 91 nodes on a single net. I built my system starting 6 years ago and have built it slowly over time. I am not a fan of evolve switches as have had them fail more than once. To be fair I have also had a ws-100 and a Jasco switch fail. I have water sensors in all wet areas and have had them trip and shut off the water when needed. It has not failed me in this regard. Most of my problems have been between the keyboard and chair due to poor assumptions or ignorance on how the system functions. When these are corrected the system works. Is HS perfect, HE!! NO! Is it robust enough to allow most automation to be done reliably, Yes. Is there a good support community - Yes. Is there good hardware support - Yes. No system is bulletproof. There are always weaknesses and it is the responsibility of the person between the chair and the keyboard has to be aware of the limitations and address as necessary.

HS is not a real-time battle hardened system. If it was, nobody would use it as the level of knowledge would preclude all but IT professionals and very savvy hobbyists. I respect all the opinions to the contrary that have shown up in the discussion so far. From an engineering standpoint it is good enough to get the job done reliably 99.9% of the time.

Hope this helps

[Guest]

I have 60 z-wave devices. Yes, I have had to replace 2 of my Evolve brand light switches but otherwise it has been running with no issues. Been running z-wave for 7 or more years and never had a missed communication.

I got 4 of those evolve brand switches, 3 of them failed after a year or two.. the dimmers was more reliable.
Any how, I changed everything with Wx200 (20+ devices) from homeseer, hope they are reliable!

Sent from my Pixel 3 XL using Tapatalk

[ksum]

I changed the second Evolve unit to the HomeSeer switch and if/when others need replaced I will do the same. Mine are all dimmers or companions.

Also note that I do not use battery power on any devices except when I Use a motion sensor in a temporary solution such as holiday lights.

[macromark]

Older Z-Wave technology was pretty finicky. However, modern Z-Wave networks (based on Z-Wave Plus devices) are very solid.

[mikee123]

Older Z-Wave technology was pretty finicky. However, modern Z-Wave networks (based on Z-Wave Plus devices) are very solid.

I thought so too. I split my network, and one part of the house created a new network, all zwave plus devices (21 nodes now) with the exception of 1 battery non plus garage door sensor. All devices are fairly close to each other and the controller. First of all, everything was instant. Within a few days problems started again. I get zwave errors failed sending command... the devices still switch, so I can ignore the errors. But very often my motion sensor triggered lights (mains powered zwave plus sensor) triggers, but the lights can take anything up to 30 seconds to turn on. Other times they are still instant. The lights are on a Qubino RGB zwave plus module. I have turned polling off on the network to see if that makes any difference. It doesnt. I have been trying to find the gremlin for a few weeks now without success.

So my conclusion is its great for a hobbyist, but would I sell HS3 systems (I have thought about it, I would love to do that) - Unfortunately no.

[langenet]

Then there's this.... https://www.theverge.com/2019/12/19/21029661/zwave-open-standard-radios-smart-home-multiple-vendors-silicon-labs