Why should our devices disable when SmartThings is down?

Kind of a newbie question but can anyone explain to me, in light of last night’s SmartThings service failure, our local devices should cease to function when “the mothership” is down? Why aren’t our systems independent, but able to communicate with ST as long as it’s up? Finally, forgive me for asking THIS, do other smart hubs have this limitation?

Smartthings is primarily a cloud-based platform. Almost all of the code, including routines and all smart apps other than some features of SmartLighting, run on servers operated by SmartThings. In that sense, it’s sort of like when you Google something. The search doesn’t take place on your own phone/tablet/laptop. It takes place in the cloud, and you just get served up the answer.

The hub at your house is basically just the radios that send messages back-and-forth to the individual devices, and then send those messages up to the code in the cloud for processing.

So no cloud, no processing.

With the introduction of the V2 model hub a few things, and it is very few, can “run locally,” that is, run even if the cloud is not available. This is basically a few sensors and a few switches. So those motion sensors can still make those switches turn on if the cloud is not available.

But the mobile app can only talk to the hub via the cloud, so you can’t do anything with it if cloud/hub communications are not working. And a lot of devices are not available for local operation. For example, the smartthings/Hue integration will not run if the smartthings cloud is not available. And you can’t arm or disarm smart home monitor, because that status is kept in the cloud.


For more Details and any follow-up questions about local processing, see the following active thread:

As to why SmartThings is designed this way, The short answer is just that they thought it would be easier to do lots of integrations. ( some people also believe that the company intends to eventually do data mining on what its customers do, but that’s never been definite.)

Some of the competing home automation systems in this price range are also cloud-based, like Lowes iris. Some run almost entirely local, but then they can’t do some of the cool cloud integrations.

Some, like vera, do the initial set up via the cloud and after that run locally. But then they don’t give you IFTTT integration or echo integration or LIFX integration or most of the other cloud to cloud options because vera doesn’t have a cloud.

Every system has pluses and minuses. The following thread discusses these:

But the short answer is that smartthings is a cloud-based architecture. No cloud, no operations. Or rather, only a very few operations.

Which devices run locally?

And here’s the list of device type handlers which have been verified to run locally:

So those will work with smartlighting if they are the only devices in your automation. As soon as you add a device, like a Hue bulb, which is not eligible to run locally, the whole automation has to run in the cloud.


Wow, thank you for such an incredibly in-depth and helpful response!! :slight_smile: I really appreciate it.

1 Like

If you use smart light switches (zwave GE Dimmer for example) they will always work like a normal light switch, regardless of the cloud or smartthings.

When I designed my automation, I always have a manual backup. Even if everything ran 100% local, there is a chance the hardware will fail.

They will only work like a normal light switch if they are a load control switch. Some are, some aren’t. It’s definitely something to be aware of when you plan device selection.

(Smart bulbs always act like a normal bulb in the sense that they turn on when they get power and they turn off when they don’t.)

Well, good question. Smart things always might not work properly. If you have more technical problems, better you contact a technician.