Free ENBIS Webinar by Peter Scheibelhofer23 March 2015; 12:45 – 13:30; Webinar
Peter Scheibelhofer will talk about "Robust Multivariate Process Control of Multi-Way Data with Applications in Semiconductor Manufacturing". The webinar will be moderated by Shirley Coleman.
The monitoring and evaluation of the manufacturing process condition is a crucial challenge in modern semiconductor manufacturing. During production steps each production unit (wafer) records data information from multiple process variables and at multiple time points. This results in multi-way data arrays.
In this webinar we will present a generalized methodology for multivariate process control of such multi-way arrays by using multi-way principal component analysis. We will demonstrate that kernel techniques allow for capture of nonlinear relationships as frequently observed in semiconductor process data. Special attention will be paid to the statistical robustness. We will discuss two case studies in which observed changes in production processes can be detected and their root causes can be tracked down successfully. We will also present a software user interface which has been created for the multivariate monitoring of post-production checks that make the created models useful for process engineers.