Biography
Andrea Serani is a Research Scientist at the Institute of Marine Engineering of the National Research Council of Italy (CNR-INM), where he works on computational engineering design. His research addresses simulation-based design optimization, design-space learning and dimensionality reduction, physics-aware machine learning, surrogate and multi-fidelity modelling, optimization under uncertainty, computational fluid dynamics, and multidisciplinary vehicle design.
He is the lead developer of Parametric Model Embedding (PME), a methodology for constructing reduced representations of parametric design spaces while retaining a mapping to the original design variables. This research line has developed from the original PME formulation toward physics-informed and physics-driven dimensionality reduction, nonlinear reduced representations, and design-manifold learning.
Serani has contributed to the NATO Science and Technology Organization Applied Vehicle Technology Panel since 2013 and serves as Co-Chair of the AVT-404 Research Task Group on Machine Learning and Artificial Intelligence for Vehicle Design. In 2026, he received the NATO STO AVT Young Contributor Award for his contributions to advanced vehicle-design methodologies, including simulation-based optimization, multi-fidelity modelling, uncertainty quantification, reduced representations, and machine learning.
He is a Visiting Scholar at the University of Michigan, collaborating on computational methods for marine vehicle design, and an Adjunct Professor at the University of Bologna, where he teaches Computer-Aided Yacht Design. His academic activity also includes mentoring graduate students and early-career researchers in computational design, optimization, machine learning, and marine engineering.
Research interests
- Simulation-based design optimization
- Design-space learning and dimensionality reduction
- Parametric Model Embedding
- Physics-aware machine learning
- Surrogate and multi-fidelity modelling
- Optimization under uncertainty
- Computational fluid dynamics
- Multidisciplinary engineering design
Current research
Parametric Model Embedding and design-space learning
PME supports dimensionality reduction in parametric shape design while preserving an explicit relationship with the original parameterization. Current work extends this foundation through physics-informed and physics-driven formulations and investigates nonlinear reduced representations and design-manifold structure.
Simulation-based and multi-fidelity optimization
This work develops surrogate and adaptive methods that combine simulations at different levels of cost and accuracy. It addresses computationally expensive multidisciplinary design problems, including optimization under uncertain operating conditions and noisy computational outputs.
Scientific machine learning for engineering design
The research integrates geometry, physical information, simulations, and data-driven methods to construct reduced and predictive representations for engineering analysis and design. Applications include marine, naval, underwater, and aerospace vehicles.
Selected projects
Selected publications
Physics-informed dimensionality reduction for propeller shape optimization
S. Gaggero and A. Serani · Applied Ocean Research, 166, 104932
Extending Parametric Model Embedding with Physical Information for Design-Space Dimensionality Reduction in Shape Optimization
A. Serani et al. · Engineering with Computers
A Survey on Design-Space Dimensionality Reduction Methods for Shape Optimization
A. Serani and M. Diez · Archives of Computational Methods in Engineering
Parametric model embedding
A. Serani and M. Diez · Computer Methods in Applied Mechanics and Engineering, 404, 115776
Honors, leadership & professional activities
- Current
- Co-Chair, NATO STO AVT-404 Research Task Group on Machine Learning and Artificial Intelligence for Vehicle Design
- Since 2013
- Contributor to international research activities of the NATO STO Applied Vehicle Technology Panel
Teaching & mentoring
University of Bologna teaching
Adjunct Professor for Computer-Aided Yacht Design in the MSc programme in Nautical Engineering.
Research mentoring
Mentoring and co-supervision of postdoctoral researchers, research fellows, visiting scholars, PhD students, and MSc and BSc thesis work in computational design, optimization, machine learning, and marine engineering.