Department of Materials Science and Engineering Fall Seminar Series: Autonomous Experimentation in Materials Science

Description

Francesca Tavazza, group leader of the Data and AI-Driven Materials Science Group at the National Institute of Standards and Technology, will give a talk titled "Autonomous Experimentation in Materials Science" for the Department of Materials Science and Engineering.

• Note: Link to Tavazza's bio was added during the fall 2025 government shutdown and may have been changed once funding resumed.*

Abstract:

While artificial intelligence (AI) applications in everyday life are now very common, AI for science is still in early stages because of its intrinsic difficulties. Among those, the fact that scientific data are scarce, expensive to generate, and the investigated phenomena are complex and multi-faceted. Nevertheless, the potential of AI to revolutionize research is enormous, including the ability to conduct Autonomous Experimentation. Autonomous Experimentation (AE) uses automation (e.g., robots) for physical task and decision algorithms (or AI) to perform data analysis and select the next best experiment. AE eliminates much of the slow, tedious, and error-prone human labor that characterizes traditional experimental work, removing human bias and yielding greater experimental repeatability, all the while generating more knowledge from fewer experiments. In this talk we discuss a few examples of successful application of autonomous experimentation as well as its currents drawbacks and possible solutions. Among those, the need for standards/protocols to reduce the cost of autonomous platforms.