What % of the time does ST actually work?

(Wayne) #18

True, hence i only use smart bulbs for ambient and secondary lighting. All my primary and main lights are on smart switches for this reason. Also, most of my smart bulbs are connected to hue which are still accessible through a few hue switches via local connection.

(Monte Montemayor) #19

I am now convinced, due to this thread, to really understand how the switches work. I think thats the key to making this smooth. Thanks everyone.


Unless you have a one off lamp or two in the room I see no reason not to use the switches. I bought the Leviton Z wave plus ones and they work well, and it works just as a regular Decora switch, so guests do not even need to use the automation. On top of that I can always turn it on no matter if ST is up or down.


(Kyle ) #21

1% at the most

(Cody Farmer) #22

same issue here my resolution…

converted every toggle in my home into a ge zwave switches for areas I desired automation and 2 dollar ace hardware paddle switches in places or areas where I don’t need automation. so basically in my house every switch is ge but you have options.

with this I was able to set up all my routines smart lighting and core pistons with mode restrictions being very important.

so when I or we get home my automations kick in and sets to home. at this point my wife can go back to the stone age and flip whatever lights she wants without an automation being stopped. do not trigger when in home mode.

then at some point we go to bed and the night mode takes over so I know that whatever switch my wife messed with gets triggered back to its original state ready to fire our goodmorning lights.

it was a work in progress but I’ve got it to where she can do what she wants and I can rest knowing my pistons are firing.

(Monte Montemayor) #23



As you noticed, this is a never ending conversation, with people who are passionate on both sides. There are multiple reasons why someone might want to use a smart bulb for a particular use case, in particular if they want either color temperature adjustment (from cool white to warm white), if they want to create zones in a room which has multiple lights controlled by one switch, or of course if they want red/green/blue/color options. Or if they just don’t want to fuss with the wiring. Smart bulbs can also dim lower than most smart switches.

Switches with dumb bulbs have a lot of pluses as well, in particular cost if you have multiple bulbs in a room and you don’t care about zone configurations. And they don’t turn themselves back on after a power outage.

At my house we use bulbs in some places and switches in others. Both are good when matched to the right need.

Again, there are already literally dozens of threads discussing this in the forums so just search and you’ll find plenty. :sunglasses:


Back to the original question of this thread… Since there may be people who find this in the future just because of the thread title, I thought I would include a link to a post where we were discussing exactly that question a few days ago:

(Glenn Brockett) #26

I would say that the downtime is far less than 1%. Think about it, if it was any higher than that, it would amount to more than 14 minutes per day.

You just really notice the times it is down because of the inconvenience and the rest of the time you ignore it.

Last night’s outage really only lasted 6 hours from report to fix, not everybody was affected for all of that period, but the repercussions (lack of refresh) lasted until your hub received new information (in my case about 10 hours) and the other items worked fine.


“Downtime” is a transactional measurement. It’s how programmers will tend to look at it. But it’s not how consumer products are typically measured. If your thermostat fails to set once every day, it doesn’t matter if it works fine 12 other times. Or if your security camera fails for five minutes every day, from a consumer product standpoint it doesn’t matter that it works fine the other 23 hours and 55 minutes. A daily failure resets the MFOP clock, so that product has zero reliability. It never makes it through a day without a failure.

Currently in the US and Europe market standards are that consumer products should be “set and forget” for their MFOP. The microwave, the television, the motion sensor, should work until they break. And that break should be at least six months out from time of purchase, and preferably longer.

It’s not about percentage of completed transactions. It’s about the amount of time between failures.


Exactly! Well said!

At to this point ST has yet to make it through a single month without some sort of failure or outage! That’s pretty ridiculous for a so called Smart Home Platform.



well said. smartthings has been unreliable for me. far too many false alarms. far too many device dropouts. if i have to constantly monitor my smart home, then i’d much rather go back to the stone age.

(Monte Montemayor) #30

I work in the Business Intelligence space for my vocation. I would really love to see some data on this and create some interactive data visualizations (using Tableau). As with anything that is “a fairly new category”, I would expect to see outages, hiccups and bumps. Some of us lived through very unstable mobile phones, very unstable internet connections, and even PC OS and app instability (you used to have written instructions on installing software on PCs in the DOS days). So as long as it’s getting better over time, I’m good.


The thing is, home automation isn’t really a new category. Many of us selected SmartThings to begin with because it was using two well-established third-party technologies: Z wave and Zigbee. And home security also isn’t new.

Moreover, there have been major introductions into the newest spaces of these categories, most notably Amazon echo, the Phillips hue bridge, and Logitech Harmony home, all three of which have sold millions of units and have very stable reliable systems.

And of course control4 does everything SmartThings does and more, including adding support for Amazon echo, and is stable and reliable. But it’s also very expensive, typically 10% of the cost of the house plus an annual fee.

The only revolutionary part of SmartThings is the price. Obviously low price in and of itself doesn’t have to mean unreliable. But SmartThings marketing promises a control4 type experience at a relatively tiny price. So far they’ve been able to deliver the power and flexibility, but not yet the reliability.


As for whether it’s trending better – – you can look at either the official status page or the unofficial first bug reports and the number of major outages has been about the same for the last year, usually one a month. And several minor issues.



As far as my personal experience, I got my system in the early fall of 2014. My best experience in terms of the combination of features and reliability was in August 2015. In October and November 2015 reliability fell completely apart, but then there was some improvement until March/April 2016, when it fell apart again. (so much so that the company CEO posted to the forums).

In the past year, my guess would be that there has been statistical backend improvement in terms of the company’s ability to service more accounts and more complexity. There have been some additional official device partners added, like Arlo, Ring and iHome. The community-developed Core smartapp has hugely added to the sophistication of rules management for stacked conditionals without requiring that everyone write their own code. And community – created projects like AskAlexa and EchoSistant have added some amazing functionality.

So there have definitely been significant improvements in the last 10 months, but I don’t know that there’s been a trend towards improvement in system reliability for individual customers. The number of outages is about the same, as is the general flakiness that requires people to pop the batteries on a sensor or run a zwave repair or re-create a rule in order to get things back to working the way they had been the previous week.

For many people, the instability and the time it takes to address it is worth it to get the really amazing flexibility and versatility of the SmartThings system at this low a price. Other people have chosen to move on, either to more expensive systems or to ones which offer much less functionality but have more stability. And still others have chosen just to move the mission critical aspects of their systems, particularly security, to another platform while keeping SmartThings for convenience options.

There’s no one right answer. Each person has to decide for themselves what factors are the most important for their household and what they’re willing to pay to get them and then find the system or systems which is the best match to their own needs.

Certainly SmartThings has said that improving stability is one of their top company priorities. I definitely hope they succeed in that. But for now, I would evaluate the product based on what it does today, not on any hoped for trend line.


(Monte Montemayor) #32

I appreciate this post but I would still argue that home automation is a new category or at best, an early adopter category. I used to work at Microsoft and they use to showcase home automation literally 10+ years ago in their visitor center. If you map out the adoption curve of the internet, which was developed in the 60’s and used primarily but EDU, Gov and Military, it took until about mid-1990’s to see wide spread adoption due to the maturity of the WWW and the early browsers (Mosiac/Netscape). Today, every one of every age globally use the Internet (mostly) everywhere. I would say that in terms of “Category Adoption”, we are in 1995 due to maturing technologies, such as IoT, Zigbee, Z-Wave, Blue Tooth, SmartThings, SimplySafe etc. When we reach ubiquitous adoption and it “just works”, then we’re mature. Someday, a generation of people using their home automation will have no idea what Zigbee is just like the vast majority of today’s internet users have never heard of TCP/IP.

(Matthew Valeri) #33

I think it’s important to differentiate the hardware vs. server consistency. One month in and I’ve experienced two lengthy server outages which have highly impacted operation. A lot of this is due to Smartthings being so dependent on the cloud for basic functions. Hardware issues I can expect and handle, I’m surprised a company like SmartThings that is backed by a major corporation like Samsung can be so crippled for multiple hours without fallback provisions.

Vera’s primary problem was updates, usually breaking more things than fixing. You could always “downgrade” your firmware though to a stable version. Also, all logic was local.

OpenHab was a very hands on DIY, and things breaking was usually at your own doing. Still, at least you can accept that and help was very abundant on the forums to get you going. You just knew not to make major changes to your system during critical times.

Server issues should be the last thing a customer needs to worry about. After all, that’s what we pay Samsung for, to handle the server issues we can’t. Downtime can be expected but in today’s age, anything over an hour is a “failure” to the customer.

(Monte Montemayor) #34

And I would argue you are being generous. :slight_smile:

(Matthew Valeri) #35

I think you can put a context on how often also.

If you ask someone if a server outage of one hour a month would be acceptable, then you might get a consensus of yes but if you ask the same for one hour a week then it might be a resounding no.

Also I fully accept there are many applications where acceptance/failure is rated in minutes. A friend handles servers that many financial institutions use. 3 minutes is a catastrophe in his world.

(Bobby) #36

So what?! You don’t like to monitor the smart monitoring system that doesn’t monitor your home without you monitoring it?

(Monte Montemayor) #37

That’s the world I live in which is why I said you are being generous.