Sorry if this question is done to death but as I don’t fully know what I’m asking I couldn’t really search.
I’m very new to ST and have a combination of compatible devices and some that are not compatible such as ikea bulbs (I know there is some success with connecting these but I’ve had limited luck), Alexa, lightwaveRF, ring doorbell and hive thermostat.
I’ve seen people mention on Facebook that its possible to use a pi to somehow bridge the gap.
I wondered if anyone would kindly point me to a link that explains in simple terms what I would like to do.
The other question is I don’t have pi but I do have a nuc that stays on nearly 24/7. Could I use the nuc instead of a pi, I guess I would have to configure a dual boot so that it runs linux as well as windows? Although the the cost of a pi it may not be worth it but just trying to cut down on unnecessary items where possible.
There’s no one simple explanation because each individual integration may be done in a completely different way. You’re just going to have to look for each one one at a time to see what other people have done.
But the basic concept is pretty straightforward for most of the devices that you listed. The raspberry pi will act as a “man in the middle“ So that smartthings will talk to the pi, A program you install on the pi will convert that request to something which the receiving device’s cloud expects (Sometimes it’s the device itself, but not usually), that gets sent to the receiving device’s cloud, who then talks to the device to execute the command.
The lightwave RF integration is a good example of this. But again, it will be somewhat different for each device you are trying to integrate and you will need separate programs for each.
If you look at the quick browse lists in the community – created wiki, and then look for projects in the device class that you were interested in, such as lighting for lightwaveRF, you can see project reports that other people have posted. Anything which is tagged “requires server” is probably using this “man in the middle“ construct. The tag saying that you have to have an additional server device to run code to make the project work.
Before you start down that rabbit hole, though, be aware that anything which has an IFTTT channel can essentially work the same way without you having to do a lot of coding. For example, the newest generation of lightwave RF does have two Way status update, Which means you can use their IFTTT channel and probably get most of the functionality you would from the server based system with almost no setup time.
Here is the thread for the server based project for LightwaveRF (this is a clickable link)
And here’s the IFTTT channel for the second generation:
Again, for the others on your list you just have to go through one by one to see if you can find a project report that someone has already posted.
And again, almost all of these are going to function as a cloud to cloud integration, not hub to pi to device. But there are a few exceptions, so just be aware as you go through them whether Internet connections are required or not.
Also…Ring and Alexa both have official integrations which are quite good, you shouldn’t need a man in the middle server for those.
If you do want to get into really fancy stuff with Alexa, you won’t use a man in the middle server approach, instead you will use code that runs in the Amazon cloud. You can find those projects on the quick browse lists on the list for “voice.“ But they have become less and less important as Amazon keeps adding features to the official integrations.