Description
Each of the presented projects covers at least one novel aspect of semiconductor manufacturing. The first project discusses the application of an automated probe station for implementing SPC on a CMOS baseline process, creation of a novel multivariate exponentially weighted moving average scheme suitable both for closed loop control as well as SPC. The second project deals with computer-based experiments. More specifically, a process simulator was connected to a statistical analysis package in order to create easy to use response surface models of process behavior. The third project deals with the application of SPC on controlling a process that has both batch as well as sequential characteristics. The technique described here employs novel data grouping and can be applied to many similarly configured microfabrication processes. The last project deals with the application of statistical techniques in modeling the behavior of plasma etching of photoresist. This project is an example on how physical understanding can be coupled with empirical observations to create a compact model of a complex process.