2025 AIChE Annual Meeting

(383c) Bridging Chemical Engineering and Data Science

Author

Bio


Elia is a postodctoral researcher at MIT Chemical Engineering. He earned his BS, MS, and PhD in Chemical Engineering from the University of Padova (Italy). During his PhD, he carried out research on the world's first industrial biorefinery process producing 1,4-butanediol from renewable raw materials, where he leveraged data-driven and hybrid modeling techniques to improve the process and and product quality. As a postdoc, Elia is developing a general framework for hybrid modeling. He is considering aspects such as the use of different forms of prior knowledge (e.g., qualitative vs. quantitative), automatic selection of the hybrid model structure, and reliability of hybrid models (e.g., extrapolation performance and uncertainty quantification).


Research Interests


Elia is interested in data-driven modeling and hybrid modeling applied to industrial processes. Indeed, his research has constantly been carried out in collaboration with industrial partners. Specifically, he is interested in using interpretable data-driven models to increase process understanding and drive process improvement, plus development and deployment of soft sensors and fault detection systems. He is also interested in the improvement of data-driven models from the algorithmic point of view, e.g., by automation of the model training and selection workflows, adaptation to changing process conditions, and extension of the modeling methods to tackle specific cases. Finally, Elia is interested in exploring the synergy between available process knowledge and information that can be extracted from process data. He is interested in developing a general framework for hybrid modeling and in understanding the added value of hybrid models with respect to purely data-driven or first-principles models, especially regarding extrapolation performance and uncertainty quantification.
Bio


Elia is a postodctoral researcher at MIT Chemical Engineering. He earned his BS, MS, and PhD in Chemical Engineering from the University of Padova (Italy). During his PhD, he carried out research on the world's first industrial biorefinery process producing 1,4-butanediol from renewable raw materials, where he leveraged data-driven and hybrid modeling techniques to improve the process and and product quality. As a postdoc, Elia is developing a general framework for hybrid modeling. He is considering aspects such as the use of different forms of prior knowledge (e.g., qualitative vs. quantitative), automatic selection of the hybrid model structure, and reliability of hybrid models (e.g., extrapolation performance and uncertainty quantification).


Research Interests


Elia is interested in data-driven modeling and hybrid modeling applied to industrial processes. Indeed, his research has constantly been carried out in collaboration with industrial partners. Specifically, he is interested in using interpretable data-driven models to increase process understanding and drive process improvement, plus development and deployment of soft sensors and fault detection systems. He is also interested in the improvement of data-driven models from the algorithmic point of view, e.g., by automation of the model training and selection workflows, adaptation to changing process conditions, and extension of the modeling methods to tackle specific cases. Finally, Elia is interested in exploring the synergy between available process knowledge and information that can be extracted from process data. He is interested in developing a general framework for hybrid modeling and in understanding the added value of hybrid models with respect to purely data-driven or first-principles models, especially regarding extrapolation performance and uncertainty quantification.