Those who come from vera


(Eduardo Veras) #1

i’m looking for a good and reliable smart home solution, im in love with smartThing, but there’s one thing that keeps me away from buying, reliability.From what i’ve read in the forums, the sistem when works, works great, but with a big family of non-tech-savy, i need a 100% reliabilidy, i think my only hope its wait for the local procesing to improve things. i was algo looking at the Vera Edge as an alternative and its starting to look very nice…so…question is…those who come from vera, why yo changed plataforms??


Convert everything from Vera to Smartthing
(Patrick Stuart [@pstuart]) #2

If you are looking for 100% reliability, no hub or platform will offer that. There will always be downtime, hiccups, issues, outages, problems.

It isn’t about 100% reliability in my opinion, its about response, customer service, adaptability and options for the future.

Local processing isn’t a panacea, and isn’t as scalable as cloud based processing is. What happens when your local hub is lagging because a stuck process has it waiting for a response from a device and you have to reboot it? Nothing works just like if the internet was down…


(Tim Slagle) #3

Agreed. I would like local processing for critical apps. Outside of that I would want it in the cloud because the cloud in the long run is much more robust.


(Patrick Stuart [@pstuart]) #4

No doubt there is a lot of unnecessary roundtriping of hub radio stuff that can stay local as well.

This alone will reduce the demand on ST servers greatly.


#5

While that’s true, my expectations for home automation are that it should still be a surprise when something fails, just like with home appliances like dishwashers or washing machines.

At the present time the only home automation scheduler under $2,000 I’m aware of that has made reliability a top priority is Staples Connect, and they’ve done so by NOT making openness or innovation top priorities. Or customer service, which is the equivalent of most big box stores.

They have many fewer devices, and most of the end devices they do have are among the most expensive in their categories.
You can’t write your own code, not even IFTTT recipes.
The only voice controL is IVEE, which sucks.
There’s no geofencing.

But they have demonstrated that it’s possible to make reliability and user experience a priority in a low price HA network, although the cost is a lot of the sexiness early home automation adopters are looking for.

An installation based on SC will cost more, do less, and be much stodgier than ST. But for reliability, it’s pretty good.

FWIW


(Patrick Stuart [@pstuart]) #6

Yes, but nothing is 100% reliable. Period. End of Story. Doesn’t matter how much you pay, nothing will be bulletproof reliable if it needs power and communicates with other devices.

Reliability should not be measured in percentage, it should be measured like a barometer. Is it getting better or worse.


#7

I’d have to respectfully disagree.

If I need a home respirator to breathe, percentage matters a lot to me. Going from 15% reliable to 45% reliable, while an impressive increase, won’t help me. Same is true for typically non life and death situations, like a refrigerator. Or the garage door staying closed at night.

I think setting a minimum floor for percentage reliability is valid for most households, if only as part of the ROI equation.


(Geko) #8

I used Vera Lite for two years before switching to ST and there’s little doubt that Vera was more reliable than ST. While nothing is 100% reliable (even nuclear plants fail as we all know), but I only had to reboot Vera two, maybe three times and never had issues with timers not firing or commands not executing, which is a weekly routine in ST world. It’s true that Vera has other issues - clunky UI, limited remote access, poor support, but as far as reliability goes, it was way ahead of ST.


(Patrick Stuart [@pstuart]) #9

Still to your point, anything requiring power, needs a backup source to function, remove power, it is 0% functional.

This is why I hate when LOS / TOS talk about X 9s of uptime as a measure of reliability… Past performance is not a guarantee of future success.

Edge cases aside. The question is about the reliability of a product. Moving parts, electricity, light, etc. all contribute to functional reliability…

Even a broken clock is reliable 2 times a day :smile:

The real question if ST is getting more or less reliable over time. Especially with the growing user base and changing server and hub hardware / software.


(Geko) #10

I think you’re confusing engineering with financial gibberish. In engineering, statistics do matter and have real meaning. Reliability analysis is a science, not some voodoo. There’s a difference between 5 days (ST) and 6 month (Vera) MTBF for most people, even those who hate statistics.


#11

And to add a very mundane example:

If the washing machine breaks down, there’s an easy fix: go to the laundromat. Some inconvenience, some dollar cost, but not that big a deal. As @pstuart says, off is off.

But there’s a big difference in consumer behaviour when the breakdown occurs once every two years vs once every two weeks. Off is still off, but there is a cost to dealing with the off, so that cost (even if it’s just psychological) is multiplied based on the reliability.

Most consumers will describe the once every two weeks scenario as “always breaking down,” and will start thinking about alternatives. Once every two years is annoying, but not enough to motivate most people to think about buying a new machine.

So the reliability percentage does represent the cumulative cost of the various failures, even in very ordinary situations. Some failures are to be expected, but how many does make a difference to many people.


(Patrick Stuart [@pstuart]) #12

No I love statistics and solid reliability analysis. What I hate is absolutes. There are things I can control, monitor and influence and then there are things I can’t.

Mean Time Before Failure (MTBF) would be an ideal disclosure statistically for all hub providers to publish. But averages or means are just that, nothing guaranteed in the ability to function tomorrow.

Just cause the HD I bought has a MTBF of 3,000 hours doesn’t mean that one will not fail tomorrow.

If however, I could see that the MTBF rate of the HD over the last 5 years has gone from 1500 hours to 3000 hours, it would suggest something has improved the MTBF rate and could be considered more reliable.

Getting back to HA / IoT / Hub reliability, I would love to see some sort of independent measure of Quality of Service (QOS) rating and stats on system uptime and latency.

To me, the Quality of Service and thus reliability is measured on the following attributes:
-Scheduled downtime / maintenence / security fixes / firmware updates etc.
-Unscheduled downtime
-Latency and availability
-MTBF of hubs
-Replacement rate of hubs in field
-# of customer support tickets
-Time to resolve average for support tickets

and so on, and so on… But this will probably never happen in a public manner. Instead we get people who say, ST isn’t reliable and then others will say, I have no problems…

We all want 100% reliable, but it is mutually exclusive with an open platform and many 3rd parties.


(Eduardo Veras) #13

i have a question since i havent test yet vera nor smartthings, in your personal cases, how many times a week/month/year you have to deal with trouble, and how many devices do you have?

i dont mind having trouble with the initial setup, but once i get something stable, i dont want to have to touch it anymore, or at least maximun 1 time every 5 moths or so


(Geko) #14

Unfortunately, with SmartThings it’s not possible. Expect service interruptions (big or small) every couple of weeks. See http://status.smartthings.com for historical perspective and keep in mind that not all interruptions are honored to be included it the status page. These may or may not require power cycling the hub to restore normal operation.

On top of periodic service interruptions, expect some apps to fail for no apparent reason now and then. This happened several times in the past few month when some apps or devices that worked perfectly well for months, all of a sudden started failing. Some may call it “growing pains”, but to me it sounds more like “pain in the neck” due to poor quality control and lack of testing.

SmartThings (only) biggest advantage is its open API, i.e. ability to create your own apps and integrate devices that are not officially supported. But if you’re not a programmer and cannot take advantage of it, you’ll probably be better off looking elsewhere.


#15

Eduardo_Veras,

I came from a Vera 2 that I had been running for a little over three years. It was mostly reliable but only had about 5 or so devices. Looking back I would not consider that a home automation setup because I still had to manually use my phone to get it to do what I wanted. When I decided to upgrade, I first bought the Veralite. I found its geofencing to be completely unreliable, because it would fire leaving/returning actions multiple times a night while I was sleeping.

I returned the Veralight and bought a Smartthings hub in January 2015 and have added another 5 or so devices. There was a bit of a learning curve and working between the Smartphone App and the IDE is sometimes a hassle, but by using the forums I have got Smartthings to do what I want. Now I would say my house is automated. It unlocks, turns the lights on, and heats for me when I get home without doing anything and then locks up, turns the lights off, and turns the heat down without doing anything. This is important, because my wife is not impressed when we were using Vera 2 and I told her I wanted her to use her phone everytime she left the house.

I have read several forum posts about unreliability but it seems to me that a majority of those are from the earlier days. I have a very reliable internet connection and Smartthings is now backed by one of the biggest electronics makers in the world, Samsung. Therefore, I expect reliability to not be an issue since Samsung wants to greatly expand the Internet of Things line of products. I fully expect the hub 2.0 to blow any Vera offering out of the park.


#16

Reliability issues were definitely not just the early days. The last two weeks, in particular, have had multiple overlapping major issues, all discussed in this forum, but they only affected certain devices/actions, so you may just not have run into them.

http://status.smartthings.com

These included:

  1. two or three separate issues with the Hue bridge. First, bulbs could not be added after the initial pairing. That was fixed, but then there was a problem involving the bridge itself. And then was a major problem where if the bridge was removed, it could not be re added. But if you don’t have a Hue bridge, you never saw any of these.

  2. horrendous problem involving the Minimote, the Enerwave wall mount keypad controller, and potentially some other devices that sent notifications. Things were toggling twice, so if you pressed the remote button it was processed as two presses, one after the other. So on/off or off/on. This made the remotes unusable. This went on for several days. Again, if you didn’t have that particular setup, you might not have seen the problem at all, but for those who relied on them, it was a very difficult time.

  3. sunset/sunrise triggers and some other timed actions never fired at all for some customers for a couple of days.

  4. in a separate issue, at a different time of day than the sunset one, some devices simply didn’t work for some customers. This happened twice in one day earlier this week.

See the official status page, most of these were listed there, all were discussed extensively in the forums.

Again, this is all just in the last few weeks, several at the same time. But you might have been affected by all of them or none, just depending on exactly how you have things set up at your house.

If you were lucky and felt no effects, I am glad for you. But the reliability issue is not solved yet. Things like the minimotes that worked for months suddenly stop working for days even though the household has made no changes of any kind.

I am using ST for several important convenience use cases, and I am happy with the value for what I’ve paid. (Well, happy now that the minimote is working again.) But at the present time I would not Recommend it for any health and safety use cases because of the unpredictability I see each week so far.

FWIW


(Tim Slagle) #17

I love watching these kinda of threads because they continually make me laugh. Where did we get the attitude that we should never have to get off our buts and turn off/on a light? I get it, these downtime things suck sometimes. So go outside. Go get an icecream. Relax, it will be ok :slight_smile: Sit down and right down all the things you are thankful for in your life. Remember that yesterday you didn’t touch a ight switch and “winnie” from Africa had to walk 10 miles to get 5 gallons of water from a dirty stream.

I’m not gonna sit here and say that STs shouldn’t invest in some more stable architecture, which im sure they already are, but when my stuff stops working for a couple hours i ususally just pick up a book and read, go volunteer somewhere, or go outside.

To speak to the topic, I’ve had them all, X10, Vera, Control 4, Insteon… STs is hands down the best.

#firstworldproblems… Our light switches sometimes don’t turn them selves on…

#gosolveproblemsthatmatter


#18

With all due respect, Tim, your use case is not my use case. I can’t turn on a light switch. And I am limited to the number of things I can do at all in a day before I just can’t do another one.

Could I imagine life in a world without my power wheelchair? Sure. Does that mean it’s a “firstworldproblem” if the battery dies unexpectedly and I have no way to get home? Well, technically, yes, but that doesn’t mean I should just “go get an ice cream.” (Oh, right, I can’t.)

And for the mom with a kid on the autism spectrum whose remote doesn’t work, or the CFA expecting a client for a meeting at their home office, or the guy with an aquarium tank full of dead fish, these may again technically be firstworldproblems, but they’re still real problems and they matter to the people who paid for the system.

A failure once a year or two, ok, that’s the price of living with technology. A failure every day or two may be a “firstadopterproblem,” but that’s not how this system is sold. It’s sold as a mass market problem solver.

As such, yes, I as a customer have a right to expect more reliability than we’ve been seeing.


(Tim Slagle) #19

All valid points and ones I don’t intend to argue against.

I guess i’m just on the other end of the spectrum. I don’t care enough about it to worry lol.


(Geko) #20

Sure. And if your car breaks down twice a week, why worry? Enjoy it, maybe it’s god’s will. Forget about going to work and take a day off. If your dishwasher craps out, rejoice. Throw away your dishes and silverware and use paper/plastic. Cooking range bit the dust? No problem. Have a take out instead. Whatever, dude… :smile: