Exploration of deep learning in LIBS analysis for Enhanced Steel Production Processes

Research question

The primary objective is to explore recent deep neural network-based techniques and machine learning to significantly enhance the laser-induced breakdown spectroscopy (LIBS) analysis of chemical composition in steel manufacturing processes. Quick and accurate chemical analysis is needed in autonomous systems of sorting steel scrap to go into the production line as a secondary raw material.  

Sustainability aspects

This innovation translates into accelerated production timelines, resulting in reduced energy consumption and increased circularity of high-quality steel. In a broader context, this not only enhances the competitive edge of Swedish steel producers but also reduces the carbon footprint associated with steel manufacturing—an essential step in our journey towards sustainability.