Gabriel D. Patrón joined the Department of Chemical Engineering & Applied Chemistry (ChemE) as an assistant professor on August 1, 2026.
Returning to the department where he began his own engineering journey, Patrón is excited to help shape the next generation of chemical engineers while advancing research at the intersection of artificial intelligence, mathematical modelling and chemical engineering.
His research focuses on developing AI and optimization methods to make chemical manufacturing and energy systems more efficient, sustainable and adaptable. By combining machine learning with traditional process systems engineering approaches, Patrón aims to create tools that can help engineers make better decisions — while ensuring those decisions remain transparent and understandable.
Growing up in Oakville, Patrón chose to study chemical engineering at the University of Toronto because of its reputation for innovation and academic rigour. During his undergraduate studies, he developed a strong interest in process modelling, control, sustainability and energy, completing U of T Engineering’s Sustainable Energy minor. These experiences shaped his goal of using mathematical and computational approaches to address some of the biggest challenges facing the chemical engineering field.
We spoke with Patrón about his path back to U of T, the future of AI in chemical engineering, and his approach to teaching and mentorship.
What initially sparked your interest in chemical engineering and your field of research?
I chose chemical engineering because of my love for mathematics and my desire to apply it to practical problems.
During my undergraduate studies at U of T, I was especially drawn to courses related to process modelling and control, thanks to some great instructors, including Professor Krishna Mahadevan (ChemE). At the same time, I became increasingly interested in sustainability and energy — two of the biggest challenges facing humanity.
These interests led me to complete U of T Engineering’s Sustainable Energy minor and shaped my goal of taking a mathematical and computational approach to sustainable engineering.
How have your research and teaching experiences shaped your career?
The AI boom began during my time as a graduate student at the University of Waterloo, and it has strongly influenced the approaches I have used — and will continue to use — in my research group.
With the abundance of data and computing power now available, AI has the potential to make chemical manufacturing safer and more efficient. However, my teaching experiences have reinforced that these tools must also be understandable.
Chemical engineers and plant operators need to be able to trust and interpret AI outputs without needing to become AI specialists themselves.
Together, these experiences have shaped my focus on developing practical, interpretable AI methods for chemical process systems.
What excites you most about returning to U of T and joining the ChemE community?
I am very excited to be part of such a diverse department, with research spanning areas such as bioprocessing, electrochemistry and nuclear engineering.
Having such a wide range of potential applications within the department is incredibly exciting. U of T also has a strong AI ecosystem, and I hope to contribute through initiatives involving the Vector Institute and the Acceleration Consortium.
Beyond U of T’s world-class research, I am also excited to work with faculty members who taught me the fundamentals of chemical engineering.
How would you describe your research? What problems are you trying to solve?
My research is about using mathematics and AI to produce everyday chemicals and energy more efficiently and sustainably.
Machine-learning methods can help identify the conditions under which production systems work best, while mathematical optimization helps translate those insights into decisions about how to operate a plant or energy system.
A central part of my work is understanding the decisions made by AI. How can we, as humans who do not think in 1s and 0s, understand why these systems make certain recommendations?
This transparency will be essential to the broader adoption of AI in chemical manufacturing.
On an industrial level, I aim to make production processes more cost-efficient. On a societal level, I hope these efficiency improvements result in more sustainable and less energy-intensive processes.
My work may also help inform policy and investment decisions by using machine learning to identify the conditions and strategies that have supported successful clean technology adoption, and then exploring how those lessons can apply to emerging technologies.
What research directions are you most excited to pursue at U of T?
In chemical engineering, modelling real-world systems is challenging because mathematical models inevitably differ from the plants they represent.
One exciting application of AI is using data to learn these mismatches and express them symbolically as equations. This could help uncover previously unknown chemical phenomena and improve our understanding of complex systems.
Another area I am excited to explore is using neural networks to represent electricity price uncertainty in energy systems optimization. These approaches could lead to efficient optimization models for battery operation that can be solved in real time.
How would you describe your teaching philosophy?
My teaching philosophy is centred on active, application-driven learning.
Students learn technical material most effectively when they can connect theory to a physical system, work through problems themselves and receive immediate feedback on their reasoning.
I aim to combine clear explanations of fundamentals with live problem-solving, discussion and computational exercises. My goal is not only to help students obtain the right answer, but also to give them the confidence and tools to approach unfamiliar engineering problems independently.
What strategies do you use to engage students in the classroom and support their learning?
I use several teaching tools — such as the whiteboard, slides, demonstrations and video — within a lecture. Switching between them can help sustain students’ attention and vary the pace of the class.
I am also a strong advocate of learning by doing. I like to pair lectures with live example problems that invite class participation, allowing students to observe the thought processes of both their peers and the instructor.
For mathematical subjects, I supplement classroom teaching with computer laboratories where students can put theory into practice. I also introduce new concepts through familiar, real-world examples — for instance, connecting control systems to applications such as automotive cruise control or temperature regulation in the human body.
I also aim to adapt to my students’ needs by adjusting my teaching style in response to class feedback.
What advice would you give students interested in pursuing computational research?
Computational research is more accessible than ever, and students should not wait until they feel like expert programmers to begin. Choose a small question that genuinely interests you, build the simplest model you can, and then test, refine and document it. Online resources and generative AI can be useful learning aids, but they are not substitutes for understanding the fundamentals or validating your results.
The hardest part is often deciding to start; once you do, your coding skills and research judgement will grow with each project.
What do you enjoy doing outside of work?
I am an avid squash player and a close follower of Premier League football (soccer) and the Raptors. When I am not playing squash or watching sports, you can probably find me back in Oakville visiting my family or watching movies and TV.
What is one thing students might be surprised to learn about you?
I really dislike watery vegetables and fruits, such as cucumber and watermelon — I’d rather just drink a glass of water!
