As soon as I got wind of the official ST/Echo integration’s impending release, I ordered an Echo.
The existing integration is wonderful for controlling the house, but is distinctly lacking when you try to go the other direction and control Alexa. If only SmartThings could…
This parent/child SmartApp creates virtual momentary button tiles which, when activated via another SmartApp or a routine, use a text-to-speech device to issue a command to your Amazon Echo. Naturally, this relies on a Sonos or other device with capability.musicPlayer being within shouting distance of your Echo.
Here are just a few possible use(less) cases:
Want to have your favorite Prime radio station playing when you walk in the door? Create a new TTS command to “Play Station Milli Vanilli” and add the resulting virtual switch to your arrival routine.
Kids fussing about going to bed on time? Make Alexa the bad guy! Schedule a switch to “Start a ten minute timer”!
Want Alexa to tell your dog a joke while you’re away at work? “Tell me a joke”! We’ll make you a button for that, too.
Anything* you can verbally tell Alexa to do can now be done by SmartThings. If I’m being honest, I made this so I could start playing radio stations when we got home. The other suggested uses are, admittedly, pretty weak. Nevertheless, I figured I’d throw it against a wall and see if it sticks for anyone.
*Subject to Alexa being able to understand Google’s text-to-speech engine’s output voice. I couldn’t, for the life of me, get her to understand “Foo Fighters” as anything but “Food Fighters”. YMMV.
NOTE: This SmartApp should be considered beta software. Use it at your own risk, especially if you have critical or sensitive devices linked to your Echo. It could kick your dog. It might burn your house down. It almost surely will fail when you most depend upon it.
My understanding is that Alexa’s noise canceling technology specifically blocks out anything that it is playing. That’s to allow you to say “volume down” without having to shout. So I don’t think you can ever have it Talk to itself.
Eventually they may add an input command API, which would be nice. But as it is, you need a second device or Alexa won’t hear it.
Hey, i’ve told you the “car in a box” story, right?
We once bridged two noncommunicating systems in a commercial install by (shhhhh!) using a remote controlled toy car in a box initiated by system A to trigger a motion sensor on system B.
Later on we “formalized” this protocol by using a fan instead. Totally the same idea, but it did look a lot more professional.