A Study on Data-based Wafer Slope Adjustment Strategies for Yield Improvement in Customized Wafer Manufacturing

With advancements in technology, the importance of customized wafer manufacturing has been increasing, particularly as the demand for precise manufacturing requirements, such as wafer slope, continues to expand. However, relying solely on simple statistical analysis or empirical approaches has limitations in effectively improving the yield of customized production. Therefore, manufacturers must adopt a data-based approach to respond effectively to these evolving requirements. This study explores the key factors influencing wafer slope adjustment for yield improvement by effectively responding to wafer slope requirements through a predictive model based on manufacturing process data. To achieve this, multiple linear regression and deep learning models were developed to control the manufacturing process for precise wafer slope implementation, and their performance was comparatively analyzed. The experimental results indicate that the key factors influencing wafer slope and their effects vary depending on wafer slope requirements and product type. Additionally, it was found that effective data collection and utilization, along with selecting an appropriate model, are essential for accurately representing product-specific requirements. Through this study, various considerations regarding data utilization for demand-driven quality management were examined, and the findings are expected to contribute to the development of strategies for improving the yield of related manufacturing processes.