This project focuses on predicting welding quality based on time series data of electrical current and voltage measurements. The analysis employs various machine learning techniques to extract meaningful features from welding process data, identify patterns through clustering, and build predictive models for quality classification.

A machine learning classification approach for risk detection in battery cell manufacturing

This repository contains the implementation for our research on out-of-distribution (OOD) detection in gas metal arc welding (GMAW) quality prediction. Our work addresses critical challenges in dynamic manufacturing environments where process parameters frequently change, causing distribution shifts that degrade model performance.