UBTECH U1 Pro joins the platform — it hears, speaks, and pulls faces
The UBTECH U1 Pro ultra-bionic humanoid is now supported: microphone, speaker, text-to-speech, facial expressions, head motions, both eye cameras, and switches for the firmware behaviours. The vendor wake word is turned off at startup so speech goes through the platform’s own pipeline.
What you get
Eight cards, usable on the canvas the moment the driver is up.
| Card | What it does |
|---|---|
mic | Microphone stream, normalised to the platform’s audio/pcm-16k whatever the device reports |
speaker | Plays whatever audio stream you wire into it |
tts | Speaks a line, can be interrupted, and reports when playback actually finished |
expression | Vendor facial expressions |
head | Vendor head motions |
camera_left / camera_right | Both eye cameras, as JPEG |
system_controls | Switches for the firmware’s own wake word, follow-up listening and autonomous visual behaviour |
The vendor’s own agent gets out of the way
A robot that already has a wake word and its own visual behaviour will talk over the one you are building. So the driver turns those three switches off at startup — and reads them back. If any of them refuses to turn off, the driver fails to start rather than registering cards that quietly fight the firmware. You can turn each of them back on from the system_controls card when you want them.
The two eye cameras are genuinely two cameras: a frame is only published as the left or the right eye when the stream says which eye it came from, so you never get one picture dressed up as two.