Full automation of research systems for materials science may be on the horizon

Since the early 2000s, automation has been dramatically increasing data throughput in the field of materials science, which combines existing data with physics to create and test new hypotheses via experiments or simulations. However, the scientific process has typically been driven by the creativity and expertise of individual researchers, human beings who update theories and models as new data or knowledge emerges. Now, with the latest advances in machine learning and artificial intelligence, some in the field say the automation of the entire closed-loop process is foreseeable.

Providing a unique industry perspective, Montoya et al. presented an overview of autonomous research systems in materials science, including both challenges and opportunities.

“The goal is to accelerate the pace of research and development through automation -- not simply to automate research,” said co-author Brian Storey. “Our vision is that autonomous systems cannot replace the researcher, but they can automate and streamline some tedious tasks, therefore providing more time for human creativity.”

While the paper highlights some successes toward this goal, Storey said “there’s still a long road ahead to get to a point where materials developed in this manner will find their way into applications, such as the battery in your electric car.”

For the authors, there is great impetus driving these efforts.

“Our interest is in applications such as electric vehicle batteries, hydrogen fuel cell catalysts, and other new materials needed to bring electrified mobility to all,” said Storey. “The pace of materials development is traditionally very slow, and if we want to meet electrified vehicle targets, we need to move faster.”

Source: “Toward autonomous materials research: Recent progress and future challenges,” by Joseph H. Montoya, Muratahan Aykol, Abraham Anapolsky, Chirranjeevi B. Gopal, Patrick K. Herring, Jens S. Hummelshøj, Linda Hung, Ha-Kyung Kwon, Daniel Schweigert, Shijing Sun, Santosh K. Suram, Steven B. Torrisi, Amalie Trewartha, and Brian D. Storey, Applied Physics Reviews (2022). The article can be accessed at https://doi.org/10.1063/5.0076324 .

This paper is part of the Autonomous (AI-Driven) Materials Science Collection, learn more here .