2006 | OriginalPaper | Buchkapitel
A Data Mining Approach to the Joint Evaluation of Field and Manufacturing Data in Automotive Industry
verfasst von : Christian Manuel Strobel, Tomas Hrycej
Erschienen in: Knowledge Discovery in Databases: PKDD 2006
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
The manufacturing quality can be evaluated only by considering the failure behavior of the product in the field. When relating manufacturing events to failure events, the main challenge is to master the huge number of combinations of both event types, of which each is only covered by a small number of occurrences. Additionally, this leads to the problem of selection of interesting findings – the appropriateness of the selection criterion for consequent decision making is a critical point. Another challenge is the necessity of mapping the process of manufacturing tests to a vector of variables characterizing the manufacturing process. The solution presented, focuses on correct rule generation and selection in the case of combinations with low coverage. Therefore statistical and decision theory approaches were used. The multiple hypothesis aspect of the rule set has also been considered. The application field was quality control of electronic units in automotive assembly, with thousands of variables observed.