Mechanical Engineering undergrad @ SUNY Korea · MEIC Lab
I work on data-driven physics modeling — teaching machines to discover the governing equations behind physical systems from data. My long-term goal is physics-informed modeling as a foundation for robust, real-world robotics.
- Data-driven equation discovery — SINDy, DMD, weak/integral-form methods (WSINDy); model discovery under partial observation & noise
- 3D reconstruction — Structure-from-Motion, 3D Gaussian Splatting, sparse-view reconstruction
- Physics-informed ML & PHM — neural operators (FNO), prognostics, remaining-useful-life prediction
- Adaptation–detection trade-offs in data-driven equation discovery under partial observation
- On-device capture guidance for SfM + 3DGS reconstruction
- Interpretability of physics neural operators
| Project | Description |
|---|---|
| DMD-SINDy | DMD + SINDy for compact, stable governing-equation discovery from vortex-shedding flows · KSME 2026 |
| Bearing-RUL-OrderTransformer | Speed-robust bearing RUL: order tracking + time-gap-aware Transformer + causal EOL smoothing, leak-free eval · PHM Korea 2026 |
- 🔗 Lab: MEIC Lab
- 🌐 Website: hajun011103.github.io
- 🎓 Google Scholar:
- 📧 Email: hajun.jang@stonybrook.edu

