Leviton says max 30-35 devices for reliable ViziaRF Z-Wave network.

My abridged ViziaRF story...

New Installation
I recently replaced my buggy original ViziaRF remote with a newer ViziaRF+ remote (thanks for the help AO), factory reset all my devices and reinstalled the z-wave network. The installation went much smoother than may prior installs and all the bugs I found in the old remote have been fixed although I did find a new very minor bug. I also had a few devices that took several attempts to include in the network, and when updating routes some nodes would fail to update on the first attempt.

Less than desireable performance
I have around 60 ViziaRF devices and my network has never been reliable even after multiple full factory resets and reinstalls. I use a lot of controller switches as 3way switches and it can take up to 45 seconds for some controller switches to actually notify the load switch to turn on the light. A few are instantaneous and some are not... and sometimes one that is normally quick will be very slow. I also have several 4 button controllers and some of them suffer the same performance issues. When I use the handheld primary controller remote to control lights the performance is instantaneous as long as I am somewhat near the destination switch.

Node Rediscovery issues
In the past I have tried using the node-rediscovery feature but it would often times fail to find all the nodes or the remote would lock up. I have not tried this with the new remote, but none of my devices have moved (nor had they in the past) so I'm not sure this is necessary.

Emails with Leviton
I have talked to Leviton several times in the past over the phone. I have also had email correspondence with Ben, an Applications Engineer at Leviton who told me to update my buggy remote with the latest firmware. My original remote was introduced before the remotes were flashable so this was not an option. Ben told me that I would have to buy a new remote or send in my remote to get it updated even though it was still under warranty. I agreed to send it in and requested several times for instructions on how to do so (via emails) and never received a response. During a phone call with Ben he also indicated that he could replace my controller switches with newer ones since they apparently knew of some issues in the older controllers. After following up on this via several emails I never received a response. During a phone call I also let Ben know that I offer a free ViziaRF software developer library online for .Net developers. Based on this and the # of devices I have he then asked if I would be interested in working with them testing there product and I said that was fine. I have not heard from them regarding this, however I don't mind since my primary concern is getting my ViziaRF lighting network stable.

Bug reports to Leviton
As an early adopter of ViziaRf and developer of software using the RS232 device I found many bugs in the primary remote controller, in the RS232 device, and incorrect documentation of the rs232 communication protocol. I documented all the issues I found and sent Ben an email. I was happy when Ben passed the email on to the engineers and Ben sent me their responses. Most of the bugs that I documented were reported to have been fixed, and I can confirm that the new ViziaRF+ remote that I have does not have any of the bugs that I reported. I was notified that a couple of the issues that I reported were by design, however IMHO I would say that in the particular case of those issues the design leads to user confusion.

The Latest Phone Call
Today I spoke to Barry, a technical support representative at Leviton. Barry told me that the slow performance that I was seeing between the controller devices and the destination switches was because I have too many devices for the network to perform well. I have about 60 ViziaRF devices. I then asked how many devices a network should have to operate quickly and reliably. He said 30-35. I double checked this with him and he confirmed it so there was no misunderstanding. He put me on hold several times to get more information from someone/somewhere as well. He said that when I use the controller switch the command was bouncing around the network trying to find the best route and since I have about 60 nodes it would take longer than if I had less nodes.

This was news to me and quite disappointing. Leviton (as other Z-Wave manufactures do) claim the network can support 232 devices.

I asked if there were any settings I could change to improve the performance. Barry said no, the only way to improve the performance would be to reduce the number of nodes in my z-wave network.

A call to one of our dealer friends
I called one of our dealer friends after speaking with Leviton and he said that the Leviton statement regarding the number of devices and performance was not true. He recommended I get the Think Essentials software and it could optimize the network. I have ordered the software and will give it a try.

Conclusion
My overall satisfaction with ViziaRF and Leviton is less than - less than desirable. I've been trying to deal with these issues for a couple years and my family is now at the point where it has to be fixed or it needs to be removed. Of course after investing $$ in many ViziaRF devices I would prefer to get it working as advertised.

At this point if someone were to ask me if I recommend ViziaRF, based on my experience with the product and the Leviton support I would have to say strongly - no. My level of disappointment with ViziaRF and Leviton support is fairly high due to the performance problems, refusal to replace a very buggy early primary remote, and the lack of followup emails from Leviton regarding replacing some of my devices. I do appreciate that I was able to get some details from the engineers which is fairly rare, but unfortunately it didn't solve my issues.

John

That is so disappointing. I feel for you tremendously. I have been watching and waiting on Vizia RF for a couple of years now. I also hear of mountains of problems with UPB and Insteon. In the mean time, my X-10 is still working fine and sounds like X-10 is still more reliable than any of the new more expensive technologies.

I guess I be waiting a few years more.

Thanks for the really good information John!

For me it was a toss up between the z-wave ViziaRF and Insteon. I decided to try Insteon. I was concerned about z-wave because the 'zwave alliance' seems to be implemented a little different across different manufacturers. This doesn't make sense to me as a standard shoud be just that; A standard. So in this situation, is it a fault of the 'standard'? or the company deviating from the standard?

hmmm.
kev

Sandpiper said:

I also hear of mountains of problems with UPB and Insteon. In the mean time, my X-10 is still working fine and sounds like X-10 is still more reliable than any of the new more expensive technologies.

Click to expand...

I don't want to go off track here but I have to say something about this since many people make decisions on reading threads like this. This statement is just flat out false. Yes, absolutely every technology has issues including UPB, but to say that x10 is still more reliable is just plain ludicrous. If x10 works for you I guarantee UPB will work. I can confidently say there is not a single person on this forum that would agree x10 is more reliable than UPB. Also saying 'mountains' of problems is also a bit unrealistic. The only thing that would make UPB unusable is a high level of noise on the powerline and all powerline technology especially x10 is susceptible to that. With UPB Gen II you can probably still use it with 10x the noise level where x10 would die. So if x10 works for you and you like it, that's awesome but let be fair to people reading and trying to choose a technology.

Ok, back to Vizia...

Steve said:

Sandpiper said:

I also hear of mountains of problems with UPB and Insteon. In the mean time, my X-10 is still working fine and sounds like X-10 is still more reliable than any of the new more expensive technologies.

Click to expand...

I don't want to go off track here but I have to say something about this since many people make decisions on reading threads like this. This statement is just flat out false. Yes, absolutely every technology has issues including UPB, but to say that x10 is still more reliable is just plain ludicrous. If x10 works for you I guarantee UPB will work. I can confidently say there is not a single person on this forum that would agree x10 is more reliable than UPB. Also saying 'mountains' of problems is also a bit unrealistic. The only thing that would make UPB unusable is a high level of noise on the powerline and all powerline technology especially x10 is susceptible to that. With UPB Gen II you can probably still use it with 10x the noise level where x10 would die. So if x10 works for you and you like it, that's awesome but let be fair to people reading and trying to choose a technology.

Ok, back to Vizia...

Click to expand...

I agree that X-10 can work just fine in many applications. I also agree that if X-10 works then UPB which has about 8 times the signal strength will pretty much certianly work.

Insteon had tremendous issues in the past and apparently still has some issues, however, you dont hear of as many recalls as you used to. That being said either the products are improving (and I think they are to a degree), they are denying the problems (possibly to a degree from what is on their forums and then deleted by the moderator), or not as many people are buying it (with the way they push their sales I think their sales are down somewhat but that is just the economy in general).

I am surprised Leviton is not as helpful as they could be. I just recently ran into some old friends that work there and they said Leviton seems to be doing pretty well all things considered.

Well, thanks J9. You could have posted that late last week!

Just this afternoon $1000+ of Vizia RF+ modules (plus $400 of other zwave stuff) arrived in support of my phase II build-out. This will bring me up to 20 units, so it sounds like I'll still be okay. I think my "max" implementation would only be about 40 zwave nodes, so if the ~35 optimum is true then I might not be too far out of it.

My experience with Vizia RF+ has been superb. The WAF is 200% higher than the Insteon system..she asked me when I was going to put those 'nice new switches' in the rest of the house. But that was a very limited 'test' set of only six nodes. I'll be doing my install this weekend and will try to remember to report out.

I'm sure your experience is valid and true, but I find it hard to believe that Vizia RF+ has those kinds of problems consistently and we haven't heard a hue and cry. I have heard that ControlThink is a life-saver that can identify the 'one bad node' that is bottle-necking the system. I hope for your (and my) sake that it is something as simple as a bad node!

Keep us informed!

Redclay said:

My experience with Vizia RF+ has been superb. The WAF is 200% higher than the Insteon system..she asked me when I was going to put those 'nice new switches' in the rest of the house. But that was a very limited 'test' set of only six nodes. I'll be doing my install this weekend and will try to remember to report out.

Click to expand...

Great Leviton experience and WAF as well, Redclay. Compared to the switches I had before, these things are superfast (Due to all being 40kbs trans?) and act instantly in conjunction with Homeseer Pro, or on their own.

Monk said:

Great Leviton experience and WAF as well, Redclay. Compared to the switches I had before, these things are superfast (Due to all being 40kbs trans?) and act instantly in conjunction with Homeseer Pro, or on their own.

Click to expand...

But how many nodes do you have and over how many square feet?

johnnynine said:

Monk said:

Great Leviton experience and WAF as well, Redclay. Compared to the switches I had before, these things are superfast (Due to all being 40kbs trans?) and act instantly in conjunction with Homeseer Pro, or on their own.

Click to expand...

But how many nodes do you have and over how many square feet?

Click to expand...

I think maybe 20 over 2000 sq ft. I was mostly commenting on them being better than what I had - which I really considered not-so-good. I noticed polling was superfast compared to the slower (9.6) switches I had. (As observed from within Homeseer's interface)

Monk said:

I think maybe 20 over 2000 sq ft. I was mostly commenting on them being better than what I had - which I really considered not-so-good. I noticed polling was superfast compared to the slower (9.6) switches I had. (As observed from within Homeseer's interface)

Click to expand...

Some of my controllers work instantly which is great... others do not and can sometimes take 30+ seconds to turn on a light.

I'd be interested if anyone else has 50+ Z-Wave devices (any brand) and their response time from controllers.

Johnny

johnnynine said:

Monk said:

I think maybe 20 over 2000 sq ft. I was mostly commenting on them being better than what I had - which I really considered not-so-good. I noticed polling was superfast compared to the slower (9.6) switches I had. (As observed from within Homeseer's interface)

Click to expand...

Some of my controllers work instantly which is great... others do not and can sometimes take 30+ seconds to turn on a light.

I'd be interested if anyone else has 50+ Z-Wave devices (any brand) and their response time from controllers.

Johnny

Click to expand...

Right! I'm interested in that as well...trust me. I'll have about 40 nodes max over about 4000 square feet (three floors, so about 1400 sq ft / floor). I'll be interested to see how that works out with the "3d" nature of the network. After watching their training videos it seems like I'll need to be very careful to ensure that I space out my controllers appropriately so that there is never more than five hops to any dimmable load.

So like I said Johnny, I don't doubt that your experience has been very bad...I'm just hoping that it has something to do with a peculularity of your setup and isn't a generic problem. I do agree that they should not advertise 128 nodes (which is all the Vizia RF+ nodes they claim) when in practical terms it is 1/4 that number.

johnny, i have documented my journey with zwave over the past years since i got in very early. My quick synopsis is that at the end of the day for entry level HA it is decent. i have personally run about 70 nodes in a single installation with a 3000 sq foot range. i also run about a 45 node network for testing purposes in a demo unit (switches, dimmers, hvac, receptacles, outdoor receptacle, window shade, etc.)

from a communication standpoint this is my experience:
1. nodes should be installed in a circumference from the center out like rings on water.
2. nodes have a 4 hop limitation so on larger networks they can bounce around alot to stay within this range.
3. one bad module can screw up the whole network. i don't care what zensys or anyone else says. rediscovery and all the other bs doesn't help.
4. there is not enough transparency on the backend of zwave. for example different classes force changes on what functions are supported on a certain firmware revision since there is only so much space on a controller. this is typically not publically documented and it should be. prime example was when schlage wrote the security class. it was so large that they removed previous functionality from the firmware to accommodate it but didn't tell anyone publically.
5. the real range of each node is about 1/2 - 3/4 the written rating.
6. for the longest time there was only 1 handheld remote to control hvac and it was a dinosaur. this was the main reason i wanted to do an app also. it drove me crazy that the zwave business model was to just make a product instead of a solution. i.e. there were a few hvac units on the market but none of them would make a zwave remote for it. they just assumed someone would come along and make the remotes. other than the one dinosaur this did not happen until they remotes were forced by zensys to include basic on/off control of thermostat in revision v2 of the hvac class.
7. i know the limit for zwave is 232 modules but i have an old design doc from leviton stating no more than 128 modules should be used.


i have some old docs for zwave from various sources. if you want to them PM me and i will get them over to you. just keep in mind the chips are different now so some may not apply anymore but some of the docs are good reading.

what i would really like to see from zwave

1. remote inclusion/exclusion including ability to create a new master remote from this method.
2. tcp/ip bridge
3. a zwave garage door opener that is actually zwave and not a sales gimmick
4. a regular physical light switch and dimmer. drives me crazy that no one would just make a regular light switch/dimmer. not sure if this was legal issues or something but the physical design of zwave lights are horrible IMO. just want your basic decora style toggle rocker switch with dimmer toggle/slider on the side.
5. pool controller - intermatic had the p350 or something like that but it was nothing but vaporware.
6. free analysis tools. how are we supposed to integrate networks and have no real tools to know what we are doing? these tools should have been free from zensys from day 1. they have always had the zniffer. put out a basic version for integrators to use.
7. public documentation of which chipsets/firmware revisions, etc. support what functions. this should not be hidden. i am not talking about codebase just functionality supported so when i go to buy a zwave product i don't have to buy every piece on the market to know which supports what and if they will work properly together for my purposes.

in the end i think zwave definitely has it's place in the HA market but it was supposed to catch like wildfire and never did for a reason. it is kind of like the position Palm is in right now on the mobile phone side. it is a great product but too many issues with hardware, marketing, execution, etc. drag it down.

then there was the lawsuit against zwave manufacturers. this is why i think intermatic left. i never followed the whole thing but it named quite a few zwave manufacturers so we'll see how this all plays out.

This is very interesting reading. I would like to share some information, but first off let me clear the air - I work for HomeSeer, and for the past year to a year and a half, have been doing almost all of our coding involved in Z-Wave. I am not going to try to sell you on Z-Wave or on HomeSeer, thought some of the information may make it appear so, it is not my intention. I just have a lot of experience and can explain some of this - we have been involved in Z-Wave since it first hit the U.S. some 7 years ago.


First, a word on the other technologies discussed here. I have been involved in home automation for almost 20 years now. I can tell you that the technology somebody chooses, due large in part to the fact that they spent a lot of money, can become a religion for them to the point where they defend it like a jihad. However, my experience tells me that the reliability of a technology goes like this: The worst is powerline, then RF, then hardwired. Yes, each can end up being perfect or even super reliable in one's personal home, but overall in different environments, the brutality of the powerline makes it the most unreliable - plug-in a new device and you could mess up everything. RF is next - move an object in the home or get a new piece of furniture, and all of a sudden something that worked fine before no longer does. The only technology with a true reliable record are those which are hardwired, where the quality of the "brain" or controller is the only point of reliability failure.

OK, so about this Z-Wave business and Leviton. Having not used the whole standard Leviton package to set up a network, I cannot comment on all of their firmware issues and controller software issues, I can only comment on what I know about Z-Wave, since there are facts about it that they cannot change.

From the very beginning they had their controllers set limits on the number of nodes you can have in a network that is vastly different than that of the Zensys standard - I believe it might have been 100 or 130 nodes or something like that, this thread has an even different number, but the fact is that a Z-Wave network can support 232 nodes. Leviton did not do this just to be obnoxious - they truly believed from their research that limiting the size of the network would make it easier to maintain and thus more reliable. I believe that is misguided, but it is not harmful.

They also believe that "rediscovery" is not needed, and this is the biggest problem. First of all, rediscovery is really a Z-Wave primary controller command called "Request Node Neighbor Update" (RNNU). In HomeSeer, we call it Optimization, but we also do more than just RNNU. RNNU of a node causes that node to send a signal out that asks all of the nodes that can hear that signal, to respond. It allows the node to update the list of nodes that are its neighbors, and this information is then forwarded to the controller to store, because most of the transmissions in a network originate with the controller - in the case of a HomeSeer system, this is the controller that HomeSeer uses. This is very important to do because most people who are using Z-Wave do not realize that they cannot move their plug-in modules without doing RNNU so that the neighbor list can be updated - even if the move is minor. In Leviton parlance, they would have you uninstall and reinstall the node since they (did not) do rediscovery (and now apparently have it). They would tell you this because when you add a node to the network, that is part of what happens - it does RNNU. Here is the rest of the story though... When the controller is trying to send a command out, there are actually 4 entry points, or nodes that it keeps track of that it can send the frame to in order to reach the desired node. If you only do one RNNU, then only one of those 4 slots has decent information - the rest may be out-of-date or invalid. This is why we recommend to our users that they do an optimization repeatedly until it starts "succeeding", and then keep doing it until you get 4 successful optimizations - each successful optimization results in a previous entry out of those 4 getting bumped, and replaced with a good entry. Now keep in mind that those 4 route starting points are only used if the controller cannot reach the node directly, and that may be another part of Leviton's formula - make sure everything is close enough to be reached directly or at least in a single hop. Unfortunately, they cannot know what a home looks like and whether that is possible in all cases. More on this in a minute...

Another thing that needs to be done is to update the other controllers in the network. A wall scene or zone controller is technically no different from a handheld controller, or a controller used on a HomeSeer system that connects via RS-232 or USB. Controllers can NOT be updated using RNNU. The only way a controller can get updated information about the nodes in the network is through replication, and it receives this information from a primary controller. When you add a zone or scene controller to the network, replication happens automatically. What you may not realize though is that if you remove a node from the network, or add a new node to the network, those scene and zone controllers have to be updated! This is a big mistake that is often made - a new scene or zone controller works fine because it just got the most recent network information, but install a new lamp module, then try to program that scene or zone controller to talk to it - it won't work because it needs to know it exists first by being updated. A controller does not get updated through RNNU because a controller makes its own decisions on routing. It keeps track of nodes that are frequently used successfully to route, and it maintains that table of which nodes are neighbors to every node to help it make decisions on how to route a frame.

The third part of this involves the node itself. If you take a relatively "dumb" or featureless node such as one of the very early ACT switches, you can see in our software that it is listed as a SLAVE type, but the Leviton nodes will be a ROUTING SLAVE. Both participate in the network as repeaters, but the difference is that if a node supports the Association command class or the Scene command classes, then it has to be able to originate a Z-Wave transmission (in other words, it has to be able to transmit) and so it needs to have some intelligence on routing. When that node is commanded by HomeSeer to turn on, the ACK message that it received the command is usually sent back to HomeSeer using the same path that the command came in on (the path that the HomeSeer controller determined through its routing information). If the node is associated with another node and needs to send a message to that node, then the node has to route the signal much like a controller does, which means it has to establish some routing information itself. This is where the third part comes in to play - HomeSeer, as part of the optimization, will also issue an "add return route" command. What this does is to tell the node to establish and save a route back to HomeSeer, so that transmissions to HomeSeer will be reliable. When HomeSeer is used to tell node A to add an association to node B, it also tells node A to "add return route" so that there too a route can be established between A and B. If direct communication fails, the node now has an established route that it will use.

So here is why our optimization is different and we think allows larger networks to flourish - we do the RNNU, we do the "add return route", and then if the node being optimized is a controller, which cannot do RNNU, we at least send a few Null Operation commands (we call it light optimization) to the node because every time you send a signal to another node, you are giving the controller data that it uses to establish frequently used and successful routes, to make it more reliable in the future.

Now with all of that said, here is my experience and some other tricks that we use: I have 116 nodes in my network, and other than the occasional failure, it works flawlessly. The occasional failures are with polls, which is a feature in our software where it periodically can query a node for its status. We have this because not all nodes support "instant status", which is where when the node is activated locally, it automatically sends out a signal. The times when a command (turn on, turn off, etc.) fails are very rare, and they only happen when I am transmitting some 20 commands at a time and the network is very congested - in fact most of the time it is the ACK that fails to get through the traffic and I find that the device did indeed turn on/off. With more and more of my devices supporting the scene commands, this will also become a thing of the past because I will be able to operate all of the devices with a single command, thus reducing the chances for collisions destroying any of the frames.

In our PRO edition of the software, we also allow you to have multiple controllers connected to the HomeSeer computer (they need to be Installer library controllers). By doing this, I am reducing the routing that has to take place, which in turn reduces traffic on the network, which in turn makes it more reliable. So, although I can control a device on the upper floor of the house from my HomeSeer machine in the basement through routing, it takes much longer and does have the higher possibility of failure. By putting another controller elsewhere in the house that is closer to the upper floor, I can now tell HomeSeer to use that controller for the nodes that are on the upper floor, and I am now able to send commands to many of those nodes directly, which means no routing, and everything is faster.

So I cannot fault Leviton for doing some of these things - they cannot somehow link together multiple handheld controllers and implement a feature like I just described, and so that handheld controller has to be able to reach everything in the network. If you start having to route a lot because your house is large, or you have metal studs in the walls of your house, then the time to send one command is longer, and the odds of a collision or other problem killing the transmission (so it has to start over with a new route) is higher. Their solution is to reduce the size of the network - fine - I think the original post describes how effective that is, but I cannot understand why more of these other maintenance functions are not done by their controller, or at least made possible. If you had a maintenance function such as our optimization, then it might work very well. On the other hand, I think it also demonstrates that a network with 232 nodes, even thoroughly meshed, is also not going to work worth a Tinker's damn (I can use that phrase!) with the default Z-Wave architecture and model - I think 232 nodes would only work with a system like we have implemented over the years.

There are, of course, many more tidbits of information and facets to the network to consider, such as the differences in how a static controller versus an installer controller will route, but they are probably unnecessary to make my point here.


Again, Leviton may or may not have legitimate issues with their firmware - I do not know - but I do know that there are some aspects of their system "out of the box" that are not going to lend themselves to work in A: a large network or B: a large home where routing, instead of direct communication, is used a lot. You can, however, take some of this information and use it to see if you can make things work... You can go through the process of "adding" all of your zone and scene controllers to the network again - even though they are ALREADY part of the network, doing the "add" part again causes replication to take place, so they will get updated with the latest information on the nodes in the network. You can, as another gentlemen mentioned, completely erase and rebuild your network, and when you do that add nodes by jumping all over the house, not grouped together and working your way out. I tell people to NEVER rebuild their network because all of that routing and frequently used nodes stuff is lost, but if the network was built up over a long time, then it may be a good reason to re-do it as I described above - why? Well, your handheld controller is an installer controller, and the way that they route if they cannot communicate directly is to start with node 1 and start sending to the the nodes in ascending order until the frame arrives at its destination. If you are in the master bedroom and all of your low numbered nodes are in the basement, then you are not going to have much success reaching nodes that you cannot send a signal to directly. A zone or scene controller installed in the wall, however, is a static controller and uses an entirely different method for routing (so there, I did spill more of those beans I alluded to earlier). So in some cases, rebuilding the network can help, but if the network was built up as I described and you are still having problems, then some sort of tool to do the functions as I described in our optimization may be needed.

So in conclusion, the Leviton nodes/devices are not a problem, and for the price vs features are pretty good. There are features in the Leviton devices that a Z-Wave system cannot get to without having some manufacturer proprietary information, and that is so that they can use those features with their controllers. However, to get those features without that restriction means going with Cooper Wiring devices - they have all of the advanced features wide open, but are much, much more expensive. When used in a network with HomeSeer (I have 4 Leviton switches and a couple of Leviton wall controllers) they work very well - they are a part of a 116 node network and I do not have the problems you are experiencing.

I hope this helps arm you with information that might help you get your network functioning well as it is NOT a bad network. If you cannot rewire your house to go with a hardwired lighting control system, then it is the best thing going. (I can rant about Zigbee too if you want - the potential it had, and the market it will likely not get unless manufacturers make some changes.)


Regards,

Tink

Quite the article Tink. As info, and to add to my previous "good performance" post - I would like to add that ALL of my nodes except for 2 can reach my controller directly.

If you have the Zseer software, you can easily see which nodes do not have a direct path and/or for some reason have less "neighbors" than other nodes. Pretty cool...

Tink,

Thanks for the great info. I seem to remember much of this being discussed in the recent HS webinar as well. Most of it is way over my head at this point.

So from the end user perspective, is the following true?

- Z-wave is well implemented in HomeSeer. That's the good news.
- Z-wave may not have the "advanced - for lack of a better word" implementation in other software systems. This would seem to be a disadvantage to the end users. Is z-wave really *that picky* about things? If so, this would seemingly stunt growth.
- Z-wave looks to be pretty complicated at the user end, if one wants to really get things "just so". Maintenance also seems to be tricky for the novice. This would be another shortcoming of the technology. Gee, isn't there anything we can just plug in for it work without all of this pain (rhetorical).

In the end, this really has me more wondering if z-wave is a good technology to invest into.