Cristina Amon, University of Toronto

Biography

Cristina Amon is University Professor in Mechanical & Industrial Engineering and Dean Emerita of the Faculty of Applied Science and Engineering at the University of Toronto (UofT). She is the Scientific Director of the UofT’s Electrification Hub and Director of the ATOMS Laboratory. Prior to joining UofT in 2006, she was the Raymond J. Lane Distinguished Professor and Director of the Institute for Complex Engineered at Carnegie Mellon University. Her pioneering research spans computational fluid dynamics, multidisciplinary hierarchical modelling, concurrent design and optimization for thermo-fluid transport phenomena, with applications to renewable energy, biomedical devices, and thermal management of electronics and electric vehicles.

Professor Amon was appointed to the Order of Canada and inducted into the Canadian Academy of Engineering, the Royal Society of Canada, the Hispanic Engineer Hall of Fame, the Spanish Royal Academy, and the U.S. National Academy of Engineering. She received the highest honour for Engineers in Canada (2020 Engineers Canada Gold Medal) and Ontario (2015 PEO Gold Medal) in recognition of her outstanding public service, technical excellence and professional leadership.

Cristina Amon is the founding chair of the Global Engineering Deans Council and a member of the Board of Directors of Academics Without Borders. She has held numerous leadership roles on editorial boards, technical conferences, and advisory and review boards. She earned her Mechanical Engineering degree from Simón Bolívar University and M.S. and Sc.D. from the Massachusetts Institute of Technology.

Title: Thermal Challenges in Electric Vehicle Batteries: Multiscale Modelling, Deep Learning, and Digital Twins

Abstract

This keynote will focus on thermal challenges of lithium-ion batteries in electric vehicles (EV), beginning with an overview of key thermal issues affecting performance, safety and lifespan. I will describe our research on hierarchical multiscale EV thermal modelling from battery electrodes to cells, modules and packs, across multiple physical domains and length scales spanning up to seven orders of magnitude.

To overcome the intense computational requirements of high-fidelity multiscale simulations, I will introduce our surrogate modelling framework based on modular, hierarchical deep convolutional encoder–decoder hierarchical (DeepEDH) neural network architectures. The DeepEDH methodology will be demonstrated through the analysis and optimization of EV battery cold plates, illustrating how deep learning can accelerate design exploration while retaining high predictive accuracy.

The lecture will conclude with a forward-looking perspective on the development of digital twins for battery thermal management, and their role in integrating physics-based and data-driven multiscale models with operational data from industry-relevant battery systems to enable real-time monitoring, predictive control, and next-generation EV battery design.