One of the issues that might come up with multiple handsfree voice interaction is microphone resource handling. We’ve seen it in a few implementation where as soon as you run a handsfree trigger, other handsfree triggers or even voice applications like video conferencing or telephone calls stop working.
Should Google Assistant, Alexa, Siri, or other AI assistants make it onto a single far field device, the wake up word trigger engines are going to have to share resources with each other. Ideally, for such an implementation, a single wake up word engine would be running. Otherwise, developers will have to create a scheme that allows multiple apps to access microphone at once.
The issue with sharing resources is that there are times where you do want to hog the mic. For example, during calls, you wouldn’t necessarily want to trigger Alexa running on that device. Likewise, in the middle of calling to Siri (again, on a hypothetical device that could run multiple services), you wouldn’t want to also trigger a music request from Google Assistant.
Mic hogging is a pain for developers. We’ll likely see some new schemes pop up over the coming year for making this easier.