Value Chain Optimization for Metal Recycling Processes through Probabilistic Modeling

Metal recycling's value chain optimization is challenged by process complexity and limited sensor technology. This study introduces a Bayesian Network framework for modeling the entire recycling process, integrating machine learning with expert knowledge. Using simulated data, we demonstrate how our probabilistic approach predicts output fractions with a confidence interval and optimizes input parameters to maximize economic efficiency. Our findings highlight the role of probabilistic modeling in advancing recycling process sustainability and efficiency.