Decision-Based Virtual Metrology Technology for Advanced Process Control
Summary This technology aims to automatically extract and define features from a data-driven framework of data engineering, machine learning, and ensemble learning algorithms to analyzing the big data collected from equipment sensors and quality characteristic measurement results. This technology provides prediction results of unmeasured products quality characteristics and its confident score base on the extracted feature for advanced process control.

In the past, when using device sensing data for virtual metrology, it was necessary to incorporate domain knowledge to define key intervals and statistics of the profile in advance for analysis. This technology integrates statistics, machine learning, and decision analysis methods, automatically define key intervals and convert them into interpretable features, combines unsupervised clustering methods to reduce errors, constructs predictive models, and calculates confidence scores for predicted values.

In the case of high-tech manufacturing such as semiconductor and TFT-LCD manufacturing, this technology can be introduced first to extract key features automatically from a large amount of equipment data without domain knowledge and experience, and further provides the virtual metrology results and confident indicators for advanced process control. In addition, it also benefits that can be quickly extended to different products and processes. Other manufacturing industries can also apply this technology after completed infrastructure and information systems.
Technical Film
Keyword Huge amount of data Semiconductor equipment Machine tool Interdisciplinary integration
Research Project
Research Team
More like this
Provide the latest information of AI research centers and applied industries