Christos Georgakis, Tufts University

Data-Driven Modeling of Batch Processes: Two Methodological Generalizations of DoE

The ever-increasing availability of process data presents us with the challenge of seriously reexamining our modeling practices. Most models can be broadly categorized into two main categories: data-driven and knowledge-driven. The present seminar focuses on the development of data-driven models. It describes two generalizations of the classical design of experiments (DoE) methodology, the long-standing data-driven modeling methodology of choice. The first generalization enables the design of experiments with time-varying inputs, called Design of Dynamic Experiments (DoDE). The second generalization enables the development of a dynamic response surface model (DRSM) when time-resolved measurements are available. We will discuss how both advances are able to contribute significantly to the modeling, optimization, and understanding of batch processes for which a knowledge-driven model is not easily at hand. Two industrial applications to a Dow batch polymerization process and a Pfizer pharmaceutical reacting system demonstrate the utility of the two generalizations.

Presentation slides


Dr. Christos Georgakis is Professor of Chemical and Biological Engineering at Tufts University where he has also been the Gordon Senior Faculty Fellow in Systems Engineering. He received his Chemical Engineering Diploma from NTU in Athens, Greece; his MS from the University of Illinois and his Ph.D. from the University of Minnesota.  He served as du Pont Assistant Professor and Edgerton Associate Professor of Chemical Engineering at MIT, as Professor of Measurement and Control at the University of Thessaloniki in Greece where he initiated the Chemical Process Engineering Research Institute. He joined Lehigh University in 1983 where he founded and directed the Chemical Process Modeling and Control Research Center. Lehigh honored him in 2001 with the Iacocca Professorship.  After two years as the Othmer Distinguished Professor at the Polytechnic University, in New York City, he moved to Tufts in 2004. His research has been recognized by a Dreyfus Foundation Teacher-Scholar Grant and the Computing Award of the CAST Division of the American Institute of Chemical Engineers.   He is a fellow of the AIChE, AAAS and IFAC. He served as the President of the American Automatic Control Council. Six years ago he initiated a new series of conferences called Future Innovation in Process Systems Engineering (FIPSE).

Professional homepage