For my thesis, I worked on a portable spectrophotometer and the software to analyze its readings. The aim was to estimate concentrations in water samples using a low-cost instrument and machine learning.
2025 / Hardware & ML
Building a low-cost spectrophotometer
The software takes absorbance data, fits calibration models, and lets you evaluate their predictions. It has a PyQt6 desktop interface, with scikit-learn, PyTorch, and XGBoost handling the modeling.
The project was recognized as Best Thesis in our program and won the Academe Expo 2025 project demonstration. The code and paper are linked below if you’d like a closer look.