Meet the recipients of the Acceleration Consortium’s new scale-up program for translational research
The Acceleration Consortium (AC) has awarded nearly $400,000 in funding through a new program at its scale-up self-driving lab at the University of British Columbia (UBC)
By encouraging collaboration across institutions and disciplines at the University of Toronto (U of T) and UBC, the program supports translational research that bridges fundamental science and practical applications. The awardees will help advance the science of self-driving labs (SDLs), and the AI and automation tools that support these SDLs.
“At the Acceleration Consortium, we are focused on advancing scientific discovery, but the real challenge lies in how we catalyze early-stage research and align it with industry standards” said AC scientific leadership team member Jason Hein. “This program is specifically designed to bridge that gap, helping translate research into scalable, real-world applications.”
Selected by the scale-up SDL’s Industrial Scientific Steering Committee, the funded projects span a wide range of research areas from new AI methods for smarter self-driving labs, to high-throughput electrochemical experimentation, scalable synthesis of electrocatalysts, and smart platforms to predict the solubility of molecules. Each initiative exemplifies the transformative potential of AI-guided experimentation and innovation at the intersection of chemistry, engineering, and data science.
The 2025 pre-competitive research grant recipients include:
Geoff Pleiss (Statistics, UBC), Alán Aspuru-Guzik
Research goal: The project aims to develop new machine learning and AI methods for self-driving labs that can be applied to materials discovery for carbon dioxide (CO2) conversion. By using AI methods to design experiments, researchers will determine the best conditions for producing materials known as electrocatalysts, which facilitate the conversion of CO2 into useful materials. As the search for better materials becomes more complex, the team aim to also create new methods to better connect machine-guided experiments with meaningful scientific insights, unlike conventional optimization which might execute a series of tasks in a lab but cannot explain the science behind it.
Eva Nichols (Chemistry, UBC), Jason Hattrick-Simpers, Yang Bai, Jason Hein
Research goal: To advance clean energy and sustainable chemical production, we need smart materials that can turn common and cheap ingredients like carbon dioxide (CO₂), nitrate, or plant-based molecules into useful fuels and products. One promising type of material for this job is called a metal–organic framework (MOF), a very high surface area sponge‐like material that can be customized for different chemical reactions. This project focuses on a special type of MOFs called SURMOFs, which are grown directly on electrode surfaces to make precise coatings with controllable thickness and conductivity. Using automated, self-driving lab techniques, robotic tools guided by AI will test and refine how these coatings form.
Corinna Schindler (Chemistry, Biochemistry and Molecular Biology, UBC; BC Cancer), Alán Aspuru-Guzik
Research goal: Electrochemical synthesis offers a sustainable alternative to traditional chemical synthesis by using electricity instead of harsh reagents, but its study is challenging due to the complex interplay between electrochemical parameters, reaction kinetics, and reactor design. This project will help to establish a generalizable framework that can accelerate electrochemical method development, expand the applicability of machine learning in data-scarce chemical environments, and enable more sustainable and accessible descriptor generation that does not rely on computational infrastructure.
Bhushan Gopaluni (Chemical and Biological Engineering, UBC), Yang Cao, Hao Han, Alán Aspuru-Guzik, Naoko Ellis
Research goal: Using machine learning, the project aims to develop a smart system that predicts how well substances dissolve in complex, multicomponent solvent mixtures–a critical step in creating new drugs, materials, and personal care products. Traditional solubility testing is slow, resource-intensive, and doesn’t scale well, especially complex mixtures or rare compounds. This system will combine molecular simulation, data-driven modelling, real-time experiments, and expert input to learn and improve as it goes, with the aim of making solubility prediction faster, more accurate and efficient to reduce development time and material waste.
Laurel Schafer (Chemistry, UBC), Sophie Rousseaux, Han Hao
Research goal: The project aims to develop the first SDL for catalyst discovery of early transition metals (ETMs). ETMs are safer and more abundant than expensive metals currently in use but have been underused due to challenges handling their air-sensitive chemistry and a lack of supporting data. By establishing this SDL, the team will be able to leverage broad opportunities across catalytic small-molecule and materials synthesis, for example, photoredox and other emerging transformations. To overcome this, the team will build an automated, air-free lab that combines high-throughput testing with machine learning to accelerate discovery. This approach could lead to cleaner, scalable methods for producing drugs and other advanced materials.
