Automatic discovery of the root causes for quality drift in high dimensionality manufacturing processes

作者：Rokach Lior^*; Hutter Dan

来源：Journal of Intelligent Manufacturing, 2012, 23(5): 1915-1930.

DOI：10.1007/s10845-011-0517-5

• Automatic root cause discovery
• Data mining
• Failure analysis
• Concept drift
• Quality control
• Fault detection
• Yield improvement

A new technique for finding the root cause for problems in a manufacturing process is presented. The new technique is designated to continuously and automatically detect quality drifts on various manufacturing processes and then induce the common root cause. The proposed technique consists of a fast, incremental algorithm that can process extremely high dimensional data and handle more than one root-cause at the same time. Application of such a methodology consists of an on-line machine learning system that investigates and monitors the behavior of manufacturing product routes.

• 出版日期2012-10