Machine learning methods to accelerate design and synthesis of clean energy materials

The aim of this project is to establish a long term collaboration between researchers at NRC/McMaster University and KIT, in order to develop machine learning (ML) methods to support and boost research on clean energy materials. Initial contact between Canadian and German researchers was established within the German Canadian Materials Acceleration Center (GC-MAC), which aims at the development of materials acceleration platforms (MAPs) for clean energy materials, bringing together expertise in materials science, automation, and machine learning in Canada and Germany.
The main focus of this MITACS visit will be to bring together a postdoctoral researcher working on machine learning methods for materials science with experimental researchers working on the development of MAPs in Canada. The goal is to generate a common understanding of theory and experiment, study the current experimental procedures for synthesis and characterization of electrocatalysts, membranes, and other energy conversion or storage materials, and identify the most promising and impactful ways of applying ML methods developed at KIT in those experiments. For promising areas, we will then create databases to train ML models or design closed-loop workflows where ML methods directly interact with automated experiments.

Faculty Supervisor:

Robert Black

Student:

Partner:

Karlsruher Institut für Technologie

Discipline:

Engineering

Sector:

Education

University:

McMaster University

Program: