I’d like to have available in automations the device that triggered a sentence. This would allow contextual automations. My wife and I are triggering assist using the apps on our Wear OS watches. I would like to be able to do automations where “turn on lights” knows it was triggered from my watch, and my watch is in the living room, so I mean turn on the living room lights. I don’t see any way to do that right now.
I’d argue that “triggered by” (device, user, etc) is sorely lacking in all automation trigger variables, but it really stands out as a missing feature within Assist.
I would also love such a feature. Currently we use Amazon Echos and these are grouped with the room they live in and the corresponding lights, thermostat and so on.
To turn on the lights we just say ‘Licht an’ Which btw. is also way more reliable than ‘Licht an im Wohnzimmer’
I’ve only just started playing around with the m5-echo and I can instantly see this would be an exceptional feature.
As @hannemann does, we use Amazon Echos, until they get replaced with m5-echos . The ability to have them “location aware” would only accelerate the demise of Amazons devices.
For the m5s you can identify which one triggered it by checking the assist in progress sensor on them.
I am also looking for a way to identify if it has been triggered by the companion app. Any ideas?
The assist in progress sensor can be used but is not reliable enough.
If there are two persons speaking to two different devices at the same time we cannot be sure anymore about which device triggered the intent.