Predicting Quality Issues in Manufactured Goods by means of Process Mining in Enterprise Resource Planning Systems

Process mining techniques have proven effective in analyzing and improving complex business processes, especially in addressing bottlenecks. However, current research predominantly focuses on historical data analysis. This paper proposes an approach that combines process mining with machine learning techniques to explore predictive analysis in real-data environments. The methods use historical data from an Enterprise Resource Planning system to cluster event traces by their process features and conformance to identify patterns associated with high and low-quality manufactured goods.