Improvement of Assemble-to-Order Model Processes With Process Mining: Dynamic Analysis and Hybrid Approaches

This article explores the application of process mining techniques to dynamically observe operations and enhance production planning efficiency in assemble-to-order (ATO) manufacturing systems using a hybrid approach. The study addresses critical challenges in managing the repetitive and complex workflows characteristic of ATO models. It highlights inefficiencies that impact productivity, lead times, and process flow. By combining direct flow graphs, inductive visual miner, and state chart workflows with a recursive algorithm, the proposed hybrid methodology identifies bottlenecks with repetitive works, analyzes variant impacts, and uncovers non-value-added activities. As a case study, the manufacturing processes of an automotive company employing an ATO strategy were analyzed. The hybrid approach achieved a fitness value of 92.5% with 77% precision, significantly improving upon the initial life cycle miner algorithm results (fitness value: 79.5%). Insights derived from the analysis informed actionable recommendations to optimize production workflows. This study demonstrates the potential of hybrid process mining techniques for understanding and improving ATO systems, offering a dynamic and scalable framework for addressing inefficiencies in similar production models.