Let me start with big thank you to the team that has dedicated so much time in getting voice assistants up and operational in HA - this project has been awesome and @synesthesiam and others get all of the kudos, this post isn’t intended to undercut any of that work.
That made clear, as an early adopter (and feedback provider!) of Wyoming Satellite (built units on RPI4s with Seeed mics), I’m very confused on the roadmap for voice features overall. It appears all of the work has been shifted to focus on the ESPHome-based VPE hardware they’re selling. Is the expectation that if you’re a Wyoming Satellite user who build satellites during last year’s “Year of the Voice” you should just eat the sunk cost and shift to VPE? Or should we hang on and just wait for certain features (pre-announce, start conversation, continue conversation, etc.) to flow to the Wyoming Satellite architecture as well. If so, is that a couple of releases, or a couple of years?
I’m all good with selling hardware, and even making that hardware the priority, I just want to understand if I’ve wandered down a dead-end road (development-wise) and need to reverse course, or just be patient? Any insights? TIA
I would also like to know this, as I’ve built several rpi based satellites for my system. I’d really like for my project to end up as a bunch of paperweights…
I have 4 Rpi zero 2 w Sitting in a drawer, there is no new documentation, and they never really worked i could never controll home assitant with them, it was always a hit and miss, I wish there was a update and a little better / easier way to do it.
Is there like a better walk through, tutorial out there where we can get to the point of getting rid of our alexas and googles…???
Mike from the future here. Wyoming satellites are being “replaced” with the (poorly named) Linux Voice Assistant. This runs on a Raspberry Pi but uses the ESPHome protocol underneath, which makes it much easier to keep it up to date with the other ESPHome-based voice assistants
Is that why we have heard so little from you ? But the results are definitely worth letting you get on with what you do so well.
After (almost) completing my greenhouse project I have bought a HA Voice PE and coming back to Voice Assist. I have removed my RasPi & Rhasspy satellites and done a clean install of HA and Voice for the VPE only.
Next stage is to bring my Rhasspy satellites (RasPi 3A and Zero W with reSpeaker HATs) into the current voice ecosystem … or have they fallen so far behind that I should just focus on ESP32-S3 (which I’m using for my greenhouse) ?
Thank you for your post, indicating that RasPi still has a future here - it comes at just the right time for me ! I shall try skipping Wyoming, and jump straight into Linux Voice Assist.
I was going to ask about hardware - specifically microphones and DSP. I’ll start a new thread with an appropriate title.
