Selecting the best defect reduction methodology

Defect rates in the range of 10 parts per million, unimaginable a few years ago, have become the standard of world‐class quality. To reduce defects, companies are aggressively implementing various quality methodologies, such as statistical quality control,^1,2 Motorola's six sigma³ or Shingo's pokayoke.^4,5 Although each quality methodology reduces defects, selection has been based on an intuitive sense without understanding their relative effectiveness in each application. A missing link in developing superior defect reduction strategies has been a lack of a general defect model that clarifies the unique focus of each method. Toward the goal of efficient defect reduction, we have developed an event tree which addresses a broad spectrum of quality factors and two defect sources, namely mistakes and variation. The quality control tree (QCT) predictions are more consistent with production experience than those obtained by the other methodologies considered independently. The QCT demonstrates that world‐class defect rates cannot be achieved through focusing on a single defect source or quality control factor, a common weakness of many methodologies. We have shown that the most efficient defect reduction strategy depend on the relative strengths and weaknesses of each organization. The QCT can help each organization identify the most promising defect reduction opportunities for achieving its goals.