YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
WHOOP Clone — Open Source Health Wearable Project
Complete architecture, hardware design, ML models, training datasets, and build guide for creating an open-source WHOOP-like fitness band.
What's Inside
- Full system architecture (Hardware → Firmware → Phone App → Cloud)
- Open-source hardware projects to start from (ZSWatch, HealthyPi Move, PineTime)
- Pre-trained ML models (NormWear, wav2sleep, UniMTS, ssl-wearables)
- Training datasets (PPG-DaLiA, WESAD, PAMAP2, MESA, SHHS, etc.)
- India-specific sourcing guide (Robu.in, ProtoCentral, Mouser India, JLCPCB)
- Build phases & timeline (36 weeks from breadboard to wearable)
- Bill of materials (~₹29,000 total project cost)
Key Pre-trained Models
| Model | Task | Link |
|---|---|---|
| NormWear | Foundation (PPG/ECG/IMU) | mosaic-laboratory/normwear |
| wav2sleep | Sleep staging | joncarter/wav2sleep |
| UniMTS | Activity recognition | xiyuanz/UniMTS |
Recommended Starting Hardware
- ZSWatch (nRF5340 + MAX30101 + BMI270) → github.com/jakkra/ZSWatch
- HealthyPi Move 🇮🇳 (nRF52840 + ECG + PPG) → protocentral.com
- PineTime (nRF52832 + PPG) → $27 devkit
See WHOOP_CLONE_COMPLETE_ARCHITECTURE.md for the full guide.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support