SmartThings Is Over Party

General Discussion
81

One more thought. While I understand the marketing reasons for the decision, the fact that SmartThings gives customers zero guidance on how to set up a network backbone does lead to a lot of frustration that shouldn’t have to exist. if people knew that

  1. they need one device that can repeat about every 40 feet (every 60 feet for Z wave plus). And it’s even better if you have two devices that can repeat.

  2. zigbee repeats only for zigbee and zwave repeats only for Z wave. The repeaters form the “backbone” of the network. So there will be one backbone for zigbee and one backbone for zwave.

  3. battery powered devices do not repeat, but most mains-powered devices do including plug-in modules, light switches, in wall relays, and plug in sensors

  4. Z wave plus has the longest range per device of the Z wave/Zigbee options

  5. The hue bridge forms its own mini network and none of the devices connected to it will repeat for any of your other devices

  6. wi-Fi is irrelevant to the performance of zigbee/Z wave devices except that very strong Wi-Fi can drown out nearby Zigbee

  7. Different architectural features can make it harder or easier for signal to get through, effectively making the range shorter. For example, foil backed insulation inside a wall might make it hard to get signal through to the next room. Signal might pass easily from one side to the other when a garage is empty, but be blocked by cars when the garage is full. These kinds of obstacles may require an additional repeater or changing the placement of devices.

  8. Z wave is limited to four total hops per message between the end device and the hub, zigbee allows for 15

  9. each end Device keeps a list of its closest neighbors. Anytime you add a new device to the network or physically relocate a device, you should run the Z wave repair utility for Z wave devices or do a zigbee heal for zigbee devices to make sure that all the neighbor tables are up-to-date. To do a zigbee heal, take the hub off power (including removing the batteries) and leave it off for at least 20 minutes. When you put the hub back on power, all the individuals zigbee devices will automatically update their neighbor tables. This process can take a while for both Z wave and zigbee, so you may not see efficiency improvements until the next day.

They would then know 95% of what they need to know to set up an efficient mesh network. One line in the hub user manual “go online for information on how to set up the most efficient communication network for your home automation devices” and a one page web explantation of those nine points and you’d save a lot of support calls.

Professionally installed systems, even Xfinity home, obviously don’t have to do that. Their installers will deal with those issues.

But I do think it makes sense to provide the backbone planning information for a DIY system.

JMO :sunglasses:

7 Likes
82

Don’t forget that ST is advertised as plug and play. It’s far from it. A lot of apps that are needed or highly desired (nest manager, askAlexa, CoRE etc) are not ST created but made by community developers here. Most buyers won’t know to come to the forum to look for apps or for help even. They expect plug and play.

2 Likes
83

Um, thanks for the Necro post? lol I can’t say my feelings reflect your feelings. In my opinion, the technology you talk about should’ve been developed years ago and it’s rather pathetic it’s coming out now, and SmartThings has had its fair share of both success AND the opportunity to make their original product even better, and they haven’t just failed at that, they’ve simply flopped. So, Frankly, the fact that they’re the best that’s out there, isn’t a plus, that’s a pathetic zero in my book. This tech is so far behind what it’s capable of it’s just ridiculous.

1 Like
84

I don’t know why, but this post surfaced on my ‘unread’ list and I didn’t look at the date. Guess they should have improved this messaging technology years ago too?? lol

We should have had solar on every rooftop years ago. We should have gotten all electric cars a decade ago. ‘They’ should have done yadaydayda and blah blah for us ages ago… sorry Shawn, but we can’t go from there. We can only go from where we are right now. And right now, that hundred dollar box of hub and sensors can’t be expected to deliver any more than what it delivers. Given what else is available, I think it does pretty good.

But you do have a choice! There are other systems available; feel free to get one in the same price range, and tell us how it stacks up. Or better yet, get out there and make Smartthings do what you know it’s capable of doing.

1 Like
85

Absolutely. Right now I’m trying to move into arduino to help make some of my own automation tools, since the once I want to get are far too expensive and I have more flexibility with arduino. Only problem is: I’m not a coder. So I’m also trying to learn a coding language or two, but I have a few mental limitations that make focusing on tedious and mundane tasks extremely difficult, so it’s been a difficult road, but I think it’ll be worth it in the end. :slight_smile: We’ll see, but I’m really hoping someone comes out with an even more inexpensive and better functioning system soon, or SmartThings listens to their customers and community and picks up the slack where it’s obviously needed.

86

@Jclendineng @JDRoberts

I am about to say something negative about the zigbee and zwave protocol but first let me say that I am happy in general with my SmartThings setup.

I have 87 devices and many automations and things work well.

AND for me, however limited local processing is in the V2 hub is, works great!!!

I spend time working around the local processing limitations to get the best performance.

I had both the V1 hub and V2 hub and in my opinion the V2 hub is much better.

-----///

Both zigbee in zwave do NOT guarantee packet delivery as part of their protocol, this in contrast to “tcp/ip” running over Ethernet or wifi, which as part of protocol guarantees end to end delivery of packets.

Also, “tcp/ip” also has guarantees about in order delivery of packets. Zigbee and zwave do not.

This is why if a device does not have a really really good wireless connection to the network it can exhibit unpredictable behavior.

Also, any devices in the network that misbehave can cause problems for all devices!!!

I had a in wall zwave switch that was silently failing for a long time that was causing problems to all items in the zwave network.

It physically worked and seemed to be work, could be controlled remotely.

But looking at the logs for it, something was seriously wrong as it was sending all sorts of messages to the network and replacing the switch with a brand new switch, improved the performance of my network.

Unfortunaly since the switch was 2 years old it was out of warranty.

Another example, when a button is hit on a battery powered device, the button sends off the message to its neighbor, but the packet may or may not arrive at the intended destination.

87

It is still early, but Wink 2 gets 4 stars. Surprising and something to watch.

1 Like
88

Mesh is mesh

No question, as the network engineering saying goes “mesh is mesh.” Mesh topology was designed for two purposes: to reduce the amount of human intervention needed to maintain the network, and to dramatically reduce the cost of operating at work, both in dollars and in energy. And that’s it.

It’s never been intended to be “better” at message delivery than a star network (Wi-Fi) or point to point (Bluetooth). Just cheaper. But that cheaper also meant much better battery life than Wi-Fi, which creates a lot of home automation options.

Most people who buy SmartThings have no idea what they’re buying, and if they’ve grown up with Wi-Fi networks, they are often disappointed at the absence of what they think of as “essential” features. Like the hub always knowing the status of every individual device.

Delivery acknowledgement

On top of that, SmartThings is just about the cheapest implementation of mesh network that you can get. It doesn’t do zigbee channel hopping. And as for proof of delivery, both zigbee and zwave do offer message received acknowledgment, but as far as I can tell, the SmartThings hub doesn’t do anything if the acknowledgement is not received. And it doesn’t give us any network mapping tools to identify repeaters that may be overburdened or have no buddy to carry the load if they do go down.

Network Backbone

As far as one device taking down the network, again, “mesh is mesh,” and if an important repeater dies you can definitely have problems. One of the biggest differences between the mesh and wifi topologies is that with mesh you have to plan the network backbone. Device placement means everything in terms of efficient message delivery.

Mesh devices are actually smarter about message routing than Wi-Fi devices because they have more options – – if a message to the hub doesn’t get through, the end device will try a different route. But it’s up to the human to make sure the alternate routes are in fact available. If a device only has one repeater to talk to, the fact that it knows that that message isn’t getting through won’t help it.

The key to avoiding that is in the design of your network backbone. I personally have two repeaters for every protocol in each room. And I check the network wellness every morning. That would be a typical industrial design for non-critical applications.

Most DIY homeowners don’t want to spend that much money on devices, and so will try to get by with one repeater every three or four Rooms, or even just pay no attention at all to the network backbone. If they have a device that seems to have range problems, they will just stick another repeating device halfway between the hub and that problem device and hope that fixes the problem.

A Chatty Failure

As far as the bad light switch, bad devices happen occasionally. They shouldn’t flood the network, though. Or at least that should be really really rare.

You didn’t mention the brand of the wall switch that failed, but just as an example I personally don’t buy GE wall switches because some things that should be “really really rare” have turned out to be a little too common with that brand for my taste. But they are very inexpensive, and if you’re willing to put up with a higher failure rate as the device ages, many people do think it’s worth it since even having to replace 10 or 15% of their switches keeps them well under their overall switch budget compared other brands.

Summary

If you buy a Mesh home automation system thinking it’s going to perform like a Wi-Fi based system, you’re going to be disappointed. Wi-Fi costs more In terms of energy use, and it delivers more for that cost.

If you buy a mesh home automation system and receive no information on how to lay out a network backbone, you may well end up disappointed again.

So you’re definitely right that there are areas where Wi-Fi is superior in message delivery to any Mesh system. But you’re going to pay for that superiority in energy use. (Including battery life for battery powered devices.)

So it’s a choice to be made. It would just be nice if users received more information about the choices before they made their device selection. JMO

5 Likes
89

Why buy a router when you can just buy a battery back up for power outages. Of course it won’t help if it’s extended or your main ISP goes down . But thankfully in my case neither is a problem.

90

the thing is that I’ve researched other things and yes, everything else runs a lot better, just not as easy to set up. homeseer is expensive, openhab works but has zigbee issues, so ill stick with smartthings hub, I only wish they would allow local user routines and apps. As a followup, the light issues I was having was fixed when I updated firmware, which caused the motion sensor to start working again. My speaker still only says “door open” after the doors shut and i dont care anymore, so def am not recommending this for the average person who doesn’t like to tinker but I love tinkering so this is right up my alley. I will wait till they get rid of those awful servers and make local everything a thing until something that is similar comes along. I’ve played with other things and the difference in processing time is night and day. The firmware update was a nice touch, but personally I would love to be able to do basic things without delay, like you can with literally every other hub out there.

91

I’m tired of waiting for SmartThings to improve. I’ve been having stability issues since I purchased it last July and it has been completely unreliable since November. When my routines failed to run again after the last update, I decided my patience had run out.

I’ve merged everything over to a Wink 2 hub. If it works reliably for the next few weeks, I’ll be selling my ST hub. The reliablily issues of this platform are simply not acceptable.

92

I can understand being frustrated if you’re having stability issues, but moving to Wink of all choices? Really?

I’ve owned my ST v2 hub since its release (August 2015?) and really have very, very few issues. On the other hand, I’ve owned Wink since November 2014 and have had several annoying issues, particularly since last November. On top of all of that, good luck waiting for Wink to support hardware people actually want. You still can’t get a humidity sensor, a vibration sensor, a light sensor, and the temp sensors are limited to the Go Control motion sensor and Leaksmart. Wink has nothing comparable to CoRE. I got tired of Wink’s excuses and jumped to ST and haven’t looked back. I did buy a Wink 2 late last November because I did like the direction they were heading, but since then, it has been nothing but more excuses and little action on the device front. Thankfully I haven’t opened it and plan on eBaying it.

Now, that’s not to say ST is utopia - it isn’t. While I’ve not had any major stability issues, I do worry about what happens if/when my hub dies - I REALLY don’t want to have to re-add all my devices. However, I’ll still take the open platform approach over Wink’s closed system with poor device support any day of the week.

3 Likes
93

None of the benefits of ST you mention do me any good if my automations fail to run. I don’t need third party support to run every IOT device on the planet. I just need the stupid thing to be reliable.

I deal in project management at work. Unless ST wants to start cutting me a paycheck, I shouldn’t be expected to troubleshoot their failings on a what should be a consumer ready product.

94

Then you’ll likely be very unhappy with Wink too. I am unaware of any IOT systems at or near this price point that are truly consumer ready. Most consider this a fairly stable but expensive - both in time and money, hobby.

95

HomeKit is reliable. It just doesn’t do very much compared to SmartThings. I often use the example of notifications.

If I want to get a push notification and have a light turn blue every time the linen closet door is open, that’s very easy to do with either HomeKit or SmartThings

But if I want to get a push notification if the closet is left open for more than three minutes, I can’t do it in HomeKit. Or if I want the blue light to turn orange after the closet is open for five minutes. But I can do both of those with SmartThings.

As of this writing, February 2017, in the low cost range the systems with more complex features also have less reliability. But if you don’t need the complex features, there are alternatives to consider.

96

[quote=“Shife, post:93, topic:63648, full:true”]
None of the benefits of ST you mention do me any good if my automations fail to run. I don’t need third party support to run every IOT device on the planet. I just need the stupid thing to be reliable.[/QUOTE]

I’m not talking about running every IoT product on the planet - I’m talking basics like multisensors and Wink has been putzing around for 2 years with no solution for humidity, light, vibration, etc. sensors. Their supported “temp sensors” are a joke. Wink’s hub hardware isn’t bad, but the company itself sucks and their model is terrible. I have no faith in them and for such an extremely limited ecosystem, you’d think they would have rock solid reliability but guess what? They don’t. I’m still a member of the Wink FB group and I cringe every time I see all the issues creeping up or when someone wants to do something that is a piece of cake on ST but is nearly impossible on Wink.

If you’re expecting 100% reliability from a $100 hub with no monthly subscription, you’re expecting too much to be honest. If you want that kind of reliability, you have to spend the big bucks.

And I’ve been in IT over 20+ years and have very few issues with ST - I don’t know what relevance our jobs have to this discussion to be honest. I’m sorry you’re having issues with ST, but many of us don’t. To be honest, I didn’t have many issues with Wink either until last November and I’ve given up on them. Their platform is just too limited and for such a small ecosystem, there is really no excuse for the issues they have.

97

When these devices are marketed as consumer ready solutions, I don’t find that as acceptable. I absolutely don’t care if Wink is less “advanced” than ST. Wink offers local control and that is where ST has failed for me. I’m not going to sit around and wait for ST to get their crap together. There’s a good chance Amazon will crush this market and force most of their competitors out anyway. Enjoy/loath ST while it’s here, because I don’t see them staying in the game, at least not with their current offerings.

98

I actually do agree that neither Wink or ST will win the HA wars - it will come down to Apple and Amazon IMO.

99

Apple won the early battle for smartphones (iOS) and Amazon for e-readers (Kindle) and we’ll soon find out if Google’s ultimately beating Apple with Android repeats itself with Google beating down Alexa.

Samsung is #1 (or close?) in Android phones and very strong in a few other product lines and is investing in IoT and smarthome on my fronts (Harman…). They can be a formidable contender if they commit to it. SmartThings has survived it’s flaws quite well; though I don’t know what the market penetration is. Thought it would be deeper by now… But you can never predict a sudden upward tick in growth.

100

I think HA, in general, will sputter along for the foreseeable future, especially DIY products. I just don’t think there is a compelling reason for the mainstream market to adopt, no “killer app”. Seems most folks in this community are technology enthusiasts who enjoy exploring what’s possible. The average consumer wouldn’t even know where to start. They won’t think about a wholistic solution. They might do some things piecemeal (i.e., use Alexa to turn on their TV), but not truly automation.

2 Likes