Wearable System Recognizes Human Activities in Real Time
This project presents an end-to-end smart wearable system that recognizes and classifies human physical activities in real time. It combines an RT-Thread RT-Spark (STM32) device, Mango Cloud backend, a lightweight temporal machine-learning model for live inference, and a Progressive Web App for visualization and diagnostics. The design emphasizes low-latency streaming, timestamped logging, and UI transparency through motion-intensity and window-accumulation metrics.
Key Points
- 1Implements RT-Thread RT-Spark (STM32) with onboard accelerometer/gyroscope to stream multi-axis motion data
- 2Trains lightweight temporal ML model in backend to perform live inference distinguishing walking, running, and resting
- 3Provides PWA showing motion-intensity, window-accumulation and live signals to improve debugging and user trust
Scoring Rationale
Practical, reproducible wearable architecture with real-time ML and transparent UI + limited novelty and single-source project description
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems