SensorTile.Box Motion Alert Wearable
Budget: $250 – $750 USD
I have an ST SensorTile.Box Pro that I want to turn into a small, wearable device. When the board detects a specific pattern of motion—head shaking(like shaking, "NO")—it must play an audible alert through the on-board buzzer. Heart-rate or temperature triggers are not required at this stage; the sole focus is motion analysis.
Scope of work
• Write clean, well-documented C firmware for the STM32 inside the SensorTile.Box Pro.
• Use the accelerometer/gyroscope to recognise the chosen repeated-movement pattern and ignore normal motion.
• Drive the audio output so the alert is clear and immediate.
• Expose a few easily adjustable thresholds or pattern parameters (hard-coded constants or a simple BLE/USB interface).
• Provide a functional prototype: compiled binary, full source code, build instructions, and a short demo video or live test session showing the alert in action.
Acceptance criteria
• Repeated-movement pattern detected reliably while false positives remain low.
• Audible alert plays within 300 ms of detection.
• Firmware builds with STM32CubeIDE or a comparable open-source toolchain.
If you have previous experience with STM32, SensorTile, motion classification, or embedded audio, this should be an enjoyable project. I'm happy to answer any questions.
Scope of work
• Write clean, well-documented C firmware for the STM32 inside the SensorTile.Box Pro.
• Use the accelerometer/gyroscope to recognise the chosen repeated-movement pattern and ignore normal motion.
• Drive the audio output so the alert is clear and immediate.
• Expose a few easily adjustable thresholds or pattern parameters (hard-coded constants or a simple BLE/USB interface).
• Provide a functional prototype: compiled binary, full source code, build instructions, and a short demo video or live test session showing the alert in action.
Acceptance criteria
• Repeated-movement pattern detected reliably while false positives remain low.
• Audible alert plays within 300 ms of detection.
• Firmware builds with STM32CubeIDE or a comparable open-source toolchain.
If you have previous experience with STM32, SensorTile, motion classification, or embedded audio, this should be an enjoyable project. I'm happy to answer any questions.