Chemometrics is the application of statistical and mathematical techniques to analyze chemical data such that information is transformed into decision making. Process chemometrics focuses on the chemometric techniques that are applicable to process data and manufacturing quality data. Example techniques include multivariate statistical process control (MSPC) and multivariate statistical quality control (MSQC).

Process chemometrics has been successfully implemented for various on-line and off-line applications in The Dow Chemical Company and the purpose of this poster is to provide an overview of these applications, implementation experiences, and research directions. In particular:

- Multivariate statistical quality control: The use of principal component analysis (PCA) to monitor on-line quality performance of products.

- Process troubleshooting: The use of regression techniques such as partial least squares PLS and classification techniques such as discriminant partial least squares (DPLS) and Fisher discriminant analysis (FDA) to diagnose process problems.

- Inferential sensor: The use of linear techniques such as PLS and nonlinear techniques such as genetic programming (GP) to develop on-line models for predicting quality parameters.

- Research directions: Current research at Dow to advance process chemometrics.

Extended Abstract Status: Not Uploaded