Edge AI · Hardware · Wearables2025 – 2026
SAFESTRIDE
“Fall detection at the edge. No cloud. No latency.”
Role
Team Leader · Hardware + Software
Stack
Arduino Nano 33 BLE SenseTFLite MicroBLEC/C++
Status
Completed
Links
Overview
SafeStride is a wrist-worn edge-AI fall detection wearable using onboard IMU sensor data and TensorFlow Lite Micro for sub-5ms offline detection.
Screenshots
Key Features
→
Real-time fall detection - Onboard IMU accelerometer + gyro
→
On-device ML - TFLite Micro model runs on Arduino MCU
Tech Deep Dive
Arduino Nano 33 BLE Sense MCU running INT8 quantized TFLite Micro model with C/C++ firmware polling at 100Hz.
Next ProjectSMART MOBILITY →