Improve Z-Wave response speed?

[rccoleman]

On 2/11/2019 at 1:01 PM, rccoleman said:

I'm still investigating (I have a few new Z-Wave Plus motion sensors coming today), but wanted to see if any of the improvements that I mentioned earlier (this post) are planned.  

@Chris Jahn, @Michel Kohanim

Maybe at least the second one?  Can we create associations between the ISY and other responders as we can with Z-Wave devices within a scene?  I think that would help improve performance, especially when an Insteon device controls a set of Z-Wave devices. That’s a very common scenario in my house. 

[lilyoyo1]

1 hour ago, rccoleman said:

Maybe at least the second one?  Can we create associations between the ISY and other responders as we can with Z-Wave devices within a scene?  I think that would help improve performance, especially when an Insteon device controls a set of Z-Wave devices. That’s a very common scenario in my house. 

Insteon is not zwave based so it cannot control zwave devices. UDI can't create something in a switch that the protocol does not support. 

The ISY simply acts as a middle man when insteon and zwave devices are in a scene together

[rccoleman]

11 minutes ago, lilyoyo1 said:

Insteon is not zwave based so it cannot control zwave devices. UDI can't create something in a switch that the protocol does not support. 

The ISY simply acts as a middle man when insteon and zwave devices are in a scene together

Yes, I understand that Insteon and Z-Wave are different protocols.  Please ignore Insteon for now.

Creating an association between a Z-Wave switch and other Z-Wave modules improves the speed at which those devices respond compared to controlling all of them individually from the ISY without associations. It seemed almost twice as fast, and there was less variation in the timing, especially when triggering multiple lights in a room. I don’t know how much autonomy the Z-Wave board in the ISY has, but I wonder if an association can be made between the Z-Wave board in the ISY and other Z-Wave devices, allowing the communication to bypass the ISY firmware and maybe improve the performance.  Perhaps it has no autonomy and all Z-Wave communication involves the ISY firmware, even communications within an association group,  but that’s why I’m asking here.

As it is, there’s sometimes a delay of a second or more when turning on two lights in the same room via an ISY scene and I’m looking for some way to improve that. 

[lilyoyo1]

7 hours ago, rccoleman said:

Yes, I understand that Insteon and Z-Wave are different protocols.  Please ignore Insteon for now.

Creating an association between a Z-Wave switch and other Z-Wave modules improves the speed at which those devices respond compared to controlling all of them individually from the ISY without associations. It seemed almost twice as fast, and there was less variation in the timing, especially when triggering multiple lights in a room. I don’t know how much autonomy the Z-Wave board in the ISY has, but I wonder if an association can be made between the Z-Wave board in the ISY and other Z-Wave devices, allowing the communication to bypass the ISY firmware and maybe improve the performance.  Perhaps it has no autonomy and all Z-Wave communication involves the ISY firmware, even communications within an association group,  but that’s why I’m asking here.

As it is, there’s sometimes a delay of a second or more when turning on two lights in the same room via an ISY scene and I’m looking for some way to improve that. 

Associations was designed for direct device to device communication. Unfortunately the popcorn effect is how the protocol is at this time rather than it being a controller issue. Hopefully this improves in future revisions of the protocol.

There is an all on/off association which I don't know if the ISY supports (not all devices support this association either) which could potentially help. However I'm not sure it would work for multiple rooms since they would all use the same parameters. 

[rccoleman]

The other problem that I’ve seen with associations is that they only seem to work if you manually control the device, at least if I try to control an associated device from the ISY. I suppose that makes sense because you’re letting the devices talk to each other without the intervention of the hub, but that also makes it unsuitable for programming.  I’m hoping that it’s a matter of sending on/off/etc *to* the association group in the associated device, or something like that, in order to have it work remotely. And the associated devices don’t report the status change, regardless of their being Z-Wave plus and supporting instant status, presumably also because they’re communicating independent of the hub, but I could work around that.

I poked around on the web and found some discussion about associating the responding devices, allowing you to turn one on and have the others come on automatically and presumably faster. That’s sounds intriguing, and I can create an association between two on/off modules in the ISY with one as a controller, but turning on one via the ISY or locally doesn’t affect the other linked device. I don’t know whether that’s a limitation of the device, protocol, or how the ISY creates associations, but I saw that Chris mentioned that there are some issues with associations in 5.0.14 that perhaps are related. As I mentioned above, I *am* able to get a Homeseer switch to directly control two other modules via associations created in the ISY, so they do work sometimes. 

Edited February 14, 2019 by rccoleman

[mwester]

One of the features I'd *REALLY* like to see, and one I think would help UDI as well, would be the addition of an "Association Editor" for advanced usage.  Hide this behind a "we're not responsible for what you do with this" Advanced dialog, if necessary.  Such a tool would be used to allow us to create/edit/delete associations in all or any of the groups, or all or any of the Z-Wave devices.  Closely related would be the ability to enable a program to be triggered on ANY message from a z-wave device -- perhaps allowing the use to specify some simple form of regular expression matching to limit the messages of interest.  This doesn't have to be fancy, or even efficient -- just enough for experimenters to take actions for devices or messages that are otherwise not-yet-supported.

The point would be to allow experimenters to see how best to set up and use Z-Wave devices -- once we establish mechanisms that work well, UDI could add those to the core support.

[rccoleman]

I’m considering setting up a HomeAssistant test environment specifically for that.

It still would be helpful to know what UDI is planning to do here. The last post that I could find from Michel said that there were no plans to support advanced association editing or to stop the ISY from destroying associations created elsewhere. The thread just ended at that point, and I hope that’s changed. Maybe I’ll actually like HomeAssistant, but I’m not replacing my Insteon switches and will still need something to tie the two technologies together. 

[Michel Kohanim]

@mwester,

We are working on something that, although not full editor, will make all these types of activities a lot easier and will not require firmware changes. 

With kind regards,
Michel

[rccoleman]

I've been playing with Home Assistant (hass.io) using an Aeotec Z-Stick 5 for a couple of days now and I'm seeing some really positive results.  Even without setting up associations, turning Z-Wave devices on and off is much, much faster than going through the ISY.  It's about as fast as Insteon, and I'm just controlling two GE/Jasco plugin modules sequentially in a set of HA "automations".

With a SwitchLinc controlling two lights and two GE/Jasco "Enbrighten" Z-Wave Plus plugin modules, all of the lights go on and off almost simultaneously when I use the SwitchLinc manually, when I use either of two RemoteLincs, or when I turn a scene on or off in the ISY.  Much of this is made possible with the built-in ISY support in HA that exposes the ISY devices and scenes directly in HA, making it easy to detect scene activation or manually flipping switches that the ISY controls.  The only thing I'm missing is seeing the status of the HA-controlled Z-Wave devices in the ISY, but I can do that via REST & variables fairly easily.  A Polyglot node server would be an even better way to expose HA entities to the ISY.

At least for now, I'm finding Home Assistant to be much less user-friendly to program than the ISY, and getting the syntax and names/types of entity IDs right is tedious at best.  There are ways to improve this by using the Configurator and Node Red, but it's far behind the ISY admin console in terms of usability.  It feels very "fiddly" for lack of a better term, and some of it is due to the free-form text entry of HA vs. drop-down list UI of the ISY.  It's much faster to edit in HA, but also much easier to get something wrong.  At this point, I don't intend to move any significant functionality from the ISY to HA, but instead to use it where Z-Wave performance is important and to experiment.

This obviously adds a whole separate component to my home automation system that needs to be maintained and could fail, so it's not for everyone.  I could also just go back to Insteon plug-in modules to get the instant control that I'm looking for, but I've gotten tired of scenes randomly not working and arduous process of identifying and filtering interference.  It worked properly most of the time, but the random failures were frustrating.

I started using HA on an RPI3 and found it to be very, very slow.  It's much better in a Virtualbox VM on my always-on Mac Mini, and not too hard to set up.  I first tried native MacOS and Docker installations on the same Mac Mini and ran into problems both times, so I recommend going straight to a VM if you're using a Mac.  Virtualbox is free and works fine.

Edit: A few more thoughts:

• With the ISY managing some Z-Wave devices and HA managing others, I basically need to maintain two different Z-Wave networks, each with a solid mesh.  As I pull line-powered Z-Wave devices away from the ISY, I'm degrading the ISY mesh to some degree.
• My HA Z-Wave network is currently only Z-Wave Plus, which technically means that it should be able to run twice as fast as the other network that includes regular Z-Wave locks a couple of motion sensors.  I wonder if that's somehow contributing to the boost in performance that I'm seeing.  I can swap out all of the motion sensors, but upgrading all three locks to new Z-Wave Plus versions would require several $100 that I'd prefer not to spend.

Edited February 18, 2019 by rccoleman

[jhimmel]

Interesting topic. I just purchased some Z-Wave devices to play around with. My house is all Insteon - every switch in the house. It is extremely reliable in function. What sent me looking at Z-Wave is two things. First, because I see questions about Insteon survival, but what really pushed me was the failure rate of the hardware. I have replaced a lot of these switches over the years - they just start malfunctioning at some point and it's annoying to keep having to replace them. They have gotten better though. I just set a Z-Wave range extender on each floor and set up a couple of lamp dimmer plug-in modules, and I have some switches to try. The couple I have in now are working fine, but not as responsive as Insteon, and not as elegant. Now I'm not sure if I should just continue buying Insteon, or keep digging into Z-Wave some more. As someone here mentioned - it's expensive and time consuming. Don't want to bet on the wrong horse.

[simplextech]

I don't think Insteon is going anywhere anytime soon.  

As for the hardware failures I see that with z-wave devices as well.  So I don't think it's a Insteon only thing about quality control or anything.  I had a HomeSeer switch just fail intermittently and it was only a couple months old.  Depending on devices, QA, build issues at the time of production run etc etc there are many things and as these are electrical devices + moving parts + exposed to surges, brown outs, etc (excluding those with whole house system protections) I think it's quite amazing that the devices last for years in general terms.  If anything I think Insteon needs to do a better QA to prevent DOA devices as I've received a DOA fanlinc out of an order of 3 which isn't a great ratio.  But again I've also received DOA components out of z-wave orders.....

[larryllix]

21 minutes ago, simplextech said:

I don't think Insteon is going anywhere anytime soon.  

As for the hardware failures I see that with z-wave devices as well.  So I don't think it's a Insteon only thing about quality control or anything.  I had a HomeSeer switch just fail intermittently and it was only a couple months old.  Depending on devices, QA, build issues at the time of production run etc etc there are many things and as these are electrical devices + moving parts + exposed to surges, brown outs, etc (excluding those with whole house system protections) I think it's quite amazing that the devices last for years in general terms.  If anything I think Insteon needs to do a better QA to prevent DOA devices as I've received a DOA fanlinc out of an order of 3 which isn't a great ratio.  But again I've also received DOA components out of z-wave orders.....

There is no better QA.
Quality Assurance is quality control by the end user. Most users will not bother with the hassle of the investigation, complaint, proof, arguing with some idiot that doesn't even understand what ISY is , return shipping etc. and QA  means the manufacturer doesn't do anything other than visual inspection while packaging, or at the most, sample one device out of each production run, to see if it even works at all.

Perhaps you meant QC (Quality Control) but from the few products I have had to eat directly from SmartHome, they definitely have no QC. I will never order directly from SmartHome again. Being in Canada, the two way shipping costs for a defective product is higher than the savings from  any sale they have ever offered. Then there is the reclaiming of the taxes from Canada Customs on straight refunds.

[asbril]

1 hour ago, jhimmel said:

Insteon, or keep digging into Z-Wave

I am big proponent of Zwave but ZW devices fail as well.  I have more than 60 ZW devices in my home and I would say that on average 4 or 5 fail every year. My preferred ZW switches are Homeseer but these are more expensive. I also use GE/Jasco, Inovelli and Zooz.

[jhimmel]

20 hours ago, asbril said:

I am big proponent of Zwave but ZW devices fail as well.  I have more than 60 ZW devices in my home and I would say that on average 4 or 5 fail every year. My preferred ZW switches are Homeseer but these are more expensive. I also use GE/Jasco, Inovelli and Zooz.

I see. Thanks for the input. I have about the same number of Insteon devices as you have ZW. I am seeing about the same failure rate as you. I guess I should have just continued replacing Insteon devices as necessary instead of complicating my setup by adding Z-Wave. It doesn't look like there is much to be gained by the move - if anything. I also utilize a couple of Keypadlincs, and I'm not seeing a real comparable Z-Wave alternative. I like the LED indicator of Dim level, and the responsiveness of the Insteon devices. If Insteon is not in danger of going away (as some have suggested), and the Z-Wave device failure rate is not much better, I guess I should just sit tight for now.

[asbril]

2 hours ago, jhimmel said:

I see. Thanks for the input. I have about the same number of Insteon devices as you have ZW. I am seeing about the same failure rate as you. I guess I should have just continued replacing Insteon devices as necessary instead of complicating my setup by adding Z-Wave. It doesn't look like there is much to be gained by the move - if anything. I also utilize a couple of Keypadlincs, and I'm not seeing a real comparable Z-Wave alternative. I like the LED indicator of Dim level, and the responsiveness of the Insteon devices. If Insteon is not in danger of going away (as some have suggested), and the Z-Wave device failure rate is not much better, I guess I should just sit tight for now.

It is a matter of "taste" and it also depends one's view of the future. I dropped my Insteon devices and am a big believer in Zwave because many more systems use Zwave, and also because Insteon depends on one sole manufacturer, who rarely brings new devices. Lilyoyo1 has a different take and sees the sole manufacturer as a plus. Emotions can run high when favoring Zwave over Insteon, or the reverse. The good thing of ISY is that it works well with both. As long as one is happy with what one has, everybody is a winner.

Edited April 25, 2019 by asbril

[lilyoyo1]

1 hour ago, jhimmel said:

 If Insteon is not in danger of going away (as some have suggested), and the Z-Wave device failure rate is not much better, I guess I should just sit tight for now.

People have been saying insteon will go out of business for the last 10 years. While anything is possible I don't see it happening soon. Even if it does, your devices would still work

[Bumbershoot]

5 hours ago, jhimmel said:

If Insteon is not in danger of going away (as some have suggested), and the Z-Wave device failure rate is not much better, I guess I should just sit tight for now.

Off topic for a Z-Wave thread, but read this post from @paulbates regarding a phone call he had with @SteveL at Insteon.  He begins to shed a little light on what Insteon is going through. 

 

Edited April 25, 2019 by Bumbershoot

[paulbates]

17 hours ago, Bumbershoot said:

Off topic for a Z-Wave thread, but read this post from @paulbates regarding a phone call he had with @SteveL at Insteon.  He begins to shed a little light on what Insteon is going through. 

Yes, thanks.  That's still where it is... given the lack of defined messaging from SmartLabs, opinions are all over the place. Nothing against anyone's opinions, but that's all that they are - opinions.. mine included. I don't believe in universal, singular, one-size-fits-all answers.       

@Bumbershoot- thanks for the mention. I had already clicked "ignore this thread" when I saw it going down this rabbit hole again :D

Paul

 

[asbril]

Going through this topic, I realize that I have a lag of several seconds in my programs that turn on a light when one of my sensors detects motion.

I am not really familiar how I can/should "associate"  one Zwave device with another. From memory I believe that I could add the sensor as controller in a scene and the light as responder, but I am not sure that this works. And even if this works, how would I have that light switch off after say 1 minute ?

[aaronburnlab]

My needs are pretty simple, turn the bathroom Fan on, if a light is on in the bathroom.

I am suffering from the performance of the Zwave / ISY combination. 

All these switches are ZWave.  Some of these switches are in Scenes to link them to each other for toggles (links)

Im not able to put all the Zwave switches in a scene and query it, eg: Scene is > 0% (a light in my bathroom is on)

So I had to make this program, now switching any light in this program on/off takes about 5-7 seconds for the actual light to respond and turn on/off. 

Ideas.. 

[lilyoyo1]

1 hour ago, aaronburnlab said:

My needs are pretty simple, turn the bathroom Fan on, if a light is on in the bathroom.

I am suffering from the performance of the Zwave / ISY combination. 

All these switches are ZWave.  Some of these switches are in Scenes to link them to each other for toggles (links)

Im not able to put all the Zwave switches in a scene and query it, eg: Scene is > 0% (a light in my bathroom is on)

So I had to make this program, now switching any light in this program on/off takes about 5-7 seconds for the actual light to respond and turn on/off. 

Ideas.. 

How long does it take for your devices to update in the isy? Are they newer devices with support for instant status?

One thing you can do is use control instead of status since you want the fan to come on with any switch.

To turn off you would want to use the status of all being off to turn off the bathroom light. While there will be a delay with the fan turning off, its not as big of a deal since you'd be leaving the room anyway

[larryllix]

Off-topic post was posted elsewhere??

Edited March 4, 2021 by larryllix
somehow jumped threads from posted thread

[oberkc]

On 3/2/2021 at 2:55 PM, lilyoyo1 said:

One thing you can do is use control instead of status since you want the fan to come on with any switch.

I have not had luck with this approach.  It may depend on the brand or version of switch one is using.  It may also require the creation of a "button press node" and using that as a control condition.  I am still not satisfied with z-wave switches as much as I am with Insteon.

I also experience measurable delays in zwave devices.  I just live with it.  Most of my zwave responders are outlets, and they turn off at the end of the day.  Delays do not bother me that much.

Edited March 4, 2021 by oberkc

[asbril]

On 3/2/2021 at 1:26 PM, aaronburnlab said:

My needs are pretty simple, turn the bathroom Fan on, if a light is on in the bathroom.

I am suffering from the performance of the Zwave / ISY combination. 

All these switches are ZWave.  Some of these switches are in Scenes to link them to each other for toggles (links)

Im not able to put all the Zwave switches in a scene and query it, eg: Scene is > 0% (a light in my bathroom is on)

So I had to make this program, now switching any light in this program on/off takes about 5-7 seconds for the actual light to respond and turn on/off. 

Ideas.. 

I have a similar program in our master bathroom and the fan hardly shows a delay. I have some 80 Zwave devices in our home and the only annoying delay that I have is with the motion sensor programs. In these programs it may indeed take 5 to 10 seconds for lights (or Alexa announcement) to react.

[oberkc]

I do not claim to fully understand the relationship between insteon, z-wave, and the ISY-994.  I have a scene with almost my entire interior devices that I use at the end of the day.  Most are insteon, but more than a few are z-wave.  I just added a new z-wave outlet the other day and added it to this scene.  I noticed that, when I turn off this scene, the newly-added device is the last device to turn off, and it is likely 5-10 seconds before doing so.  It makes me suspect that the ISY runs through each of the zwave devices within the scene and turns each off individually.  This one is the last added, and is alphabetically the last zwave device in the list.  I cannot help but wonder if re-ordering the devices somehow would result in response times being different for various devices.

Edited March 4, 2021 by oberkc