New paper accepted in Management Science
Our paper “Using explainable artificial intelligence to improve process quality: Evidence from semiconductor manufacturing” has been accepted for publication in Management Science. The paper is authored by Julian Senoner, Prof. Netland, and Prof. Feuerriegel (Chair of MIS).
In the paper, we present a novel explainable artificial intelligence (XAI) method that helps improve production yield. We propose the use of nonlinear modeling with SHAP values to infer how a set of production parameters and the process quality of a manufacturing system are related. We tested the method in a field experiment in a semiconductor factory of Hitachi ABB in Lenzburg, Switzerland. Compared to the average yield in the sample, the experiment reduced yield loss by 21.7%. Our method is now regularly used in Hitachi ABB.
This is one of the first papers that demonstrates how XAI is useful for quality management. We thank Hitachi ABB, especially Dr. Peter Kaspar and Dr. Patric Strasser, for the excellent cooperation. A postprint version of the paper is available for download Download here (PDF, 915 KB).