A Fuzzy Logic Approach for Incorporating the Effects of Managerial Actions on Semiconductor Yield Learning

Abstract

The competition in the semiconductor industry is becoming more and more fierce, which significantly distorts the learning process of semiconductor yield improvement. For example, if the yield of a product could not reach a certain level before a given deadline, then the competitiveness of the product will disappear and capacity will be re-allocated to other products. To prevent that from happening, some managerial actions, e.g. executing a quality engineering project, quickening the speed of mass production, etc. can be taken to accelerate yield learning. After such actions, the yield learning model has to be modified. In this study, how to incorporate the effects of such managerial actions on semiconductor yield learning is investigated. Subsequently, a new fuzzy yield learning model is developed. The proposed methodology has been applied to the data of four semiconductor products. Experimental results revealed the effectiveness of the proposed methodology.

Energy-Efficient Production Planning Using a Two-Stage Fuzzy Approach

“…As a result, a crisp yield-learning model could not perfectly fit the collected yield data. As an alternative, many studies [41,42,[57][58][59][60] used fuzzy-valued parameters to model the learning process of yield to generate fuzzy yield forecasts with ranges. Since there was only one factor (i.e., time) that affected the yield, and all data related to energy consumption were given, the analytical results in this study can be full replicated in another study.…”

An online collaborative semiconductor yield forecasting system

“…After these managerial actions, the yield learning model has to be modified. For this purpose, Chen and Wang's fuzzy yield learning model was modified to incorporate the effects of such managerial actions in Chen (2007a).…”