Mino — Credent’s voice device — can now turn speech into real-world action. Say “turn left” and the robot turns left. Emrys is the brain; your hardware is the body.
We’ve extended the Mino firmware so a child can speak a command and watch the world respond. Emrys — Credent’s agentic AI tutor — hears the request, understands it, and tells the device exactly what to do. The result is hands-on robotics that starts with a sentence, not a screen. Everything runs on Credent’s own voice engine, with Emrys as the only AI brain.
The firmware now ships with voice‑driven movement —
forward, backward, left, right,
and stop — built on a simple two‑motor design. Emrys decides
what to do from natural speech; the firmware drives the motors. The
same pattern powers set_led for lights, so anything you wire up,
you can control by voice.
This is a foundation, not a black box. Download the firmware source, set your
motor pins, flip one switch (MOTORS_ENABLED), and flash. Add your
own actions — an arm, a gripper, a buzzer — by copying one example. It’s the
Credent way to teach robotics: real hardware, guided by an AI teacher, starting
from a child’s own voice.
Voice detection, speech‑to‑text, and the spoken reply all run on Credent’s own engine — not a third‑party cloud. Emrys is the brain. That means privacy you control, and a path to local languages the big assistants ignore.
Explore how it works, or grab the firmware and make something move.
Voice‑controlled movement requires an ESP32‑S3 device with motors wired by the builder, and the Credent voice engine running on a reachable host. Motor support is off by default and enabled in the firmware config. This is a developer/maker capability — designed to be built with, not a finished consumer product.
Firmware source available now · one‑click browser flashing coming soon.