About
Second-year EE master's student studying automated optical inspection (AOI). I mostly write Python.
Research background
My thesis is about 3D depth inspection: a Keyence LJ-X8000 series laser profiler (LJ-X8020 head, LJ-X8000A controller) with a three-axis motion controller, forming a system that scans a surface and finds defects.
For the next two months I'll get comfortable with three things: acquiring and exporting sensor data, three-axis control and synchronization, and the processing and defect detection workflow for 3D point clouds or height maps.
Tools I use
- Python (NumPy, SciPy, Open3D and more)
- Keyence LJ-X8000 series
- A three-axis motion controller
- AI development tools such as Claude Code
- Git and config files, to stay reproducible
How I learn
Principles first, then hands-on
I understand laser triangulation and the coordinate relationships before writing code. Code without that understanding can only be debugged by guessing.
Keep it reproducible
Any result can be rerun with the same command. Environment, parameters and data versions are all recorded.
A verifiable result every week
Each week ends with something I can open, run and show, not just notes.
Why investing and education side projects
The data is easy to get, which makes these topics good for repeated practice in data cleaning, automation and agent design. They're my training ground, and what works comes back into the thesis code.