I'm using Amazon's Echo Dot, and I have several Echo devices. Let's assume this is the setup:
- Room 1 (with
Echo Dot 1) ->Light 1 - Room 2 (with
Echo Dot 2) ->Light 2
I've setup a Smart Home Skill to turn on lights, and I can turn on/off the lights by saying things like these:
- "Alexa, turn on light 1"
- "Alexa, turn on light 2"
So far, so good. Now, I would like to not tell Alexa "light 1" or "light 2", but detect which Echo device (Echo Dot 1 or Echo Dot 2) detected the utterance and triggerd the skill, so I would be able to say something like:
- "Alexa, turn on the light"
.. then decide if we are on Room 1 or Room 2 by seeing which Echo Dot activated, and turn on/off the matching light.
Since Devices have to be discovered upfront, and if no device found by the specified name the skill won't be triggered, I assume this would involve to return sort of a "virtual" device in the Discover endpoint, to model both Light 1 and Light 2, and decide which one is it by using the ID of the Echo Dot that triggered the event.
However, when debugging the invoked AWS Lambda function, I cannot find any indication about which particular Echo device was activated.
Is this possible? If not, do you know of any other way to achieve the same UX?