Management Decisions in Manufacturing using Causal Machine Learning – To Rework, or not to Rework?


Alan Turing Stage

Manufacturing & Logistics

Not only cost considerations, but even more so the increasing importance of sustainability demand for efficient manufacturing systems that produce high quality items. However, in complex value-chains, like in semiconductor manufacturing, imperfect processes lead to products that do not always meet the required quality targets. Defect compensation techniques, such as in-line rework, can affect the production yield in both ways - positive and negative. Therefore, we present a data-driven model for the estimation of optimal rework policies in manufacturing systems. To mitigate the inherent confounding and regularization bias we employ robust causal estimation methods, particularly Double/Debiased Machine Learning.