Automated material synthesis using deep-reinforcement learning

Reinforcement learning (RL) is researched extensively in game playing and has found important uses in finance, autonomous driving and commercial robotics. This project looks at the potential for RL to automate aspects of physical chemistry through gamification to discover new pathways for creating materials with desired properties. This project is a collaboration between the National Research Council (NRC) and University of Waterloo with the NRC providing expertise in physical chemistry and University of Waterloo providing expertise in AI and machine learning.

Project team

Dr. Mark Crowley

Dr. Mark Crowley is an assistant professor in the Pattern Recognition and Machine Intelligence group in the Department of Electrical and Computer Engineering at the University of Waterloo. Dr. Crowley's research focusses on algorithms, tools and theory at the intersection of machine learning, optimization and probabilistic modelling. 

Learn more about Dr. Crowley.

Dr. Colin Bellinger

Dr. Colin Bellinger is a Research Officer at the National Research Council of Canada and an Adjunct Professor at Dalhousie University in the Faculty of Computer Science. His research leverages real-world problems in health, science, security and industry as a gateway to understand how machine learning and data mining algorithms are impacted by adverse and limited data domains. Dr. Bellinger’s current research focuses on developing supervised learning algorithms and reinforcement learning algorithms for applications, such as materials design, where samples are limited and costly.

Find out more about Dr. Bellinger.