Deep-Learning System Uncovers Materials’ Interiors From the Outside

The Deep Learning System explores the interior of materials from the outside

Researchers at MIT have developed a new method that allows engineers to determine what’s inside a material by observing its surface properties. This could be anything from an airplane component to a medical device. The new method allows engineers to determine what is happening inside a material by simply observing its surface properties.

The team used deep learning, a form of machine learning that compares a large number of simulations of materials’ external forces fields with their corresponding internal structures. They then used this to create a system capable of making reliable predictions about the interior based on the surface data.

The results were published in Advanced Materials in an article by Zhenze Yang, a doctoral student and Markus Buehler, professor of civil engineering and environmental engineering.