Engineers from the University of Colorado Boulder (CU Boulder) are advancing the process of integrating artificial intelligence (AI) with advanced computer simulations in an effort to predict the failure of electronics, such as transistors used in mobile phones.
The study was recently published in the journal npj Computational Materials, led by physicist and aerospace engineer Sanghamitra Neogi.
In their new study, Neogi and her colleagues designed the physics of small individual building blocks and then applied machine learning methods to predict the behavior of larger structures produced by these small building blocks. It's like looking at a Lego brick to predict the strength of a relatively larger castle.
"We are trying to understand the physics of devices with billions of people" says Sanghamitra Neogi, Assistant Professor at Ann and HJ Smead Department of Aerospace Engineering, University of Colorado Boulder.
This quest could benefit the electronics that support our daily lives, from electric cars and smartphones to emerging quantum computers. According to Neogi, engineers could one day use researchers' techniques to identify weaknesses in the design of electrical components in advance.
"Instead of waiting years to figure out why devices fail, our methods can give us an advance knowledge of how a device works before we build it," says Sanghamitra Neogi, Assistant Professor at Ann and HJ Smead Department of Science Aerospace Engineering, University of Colorado Boulder. See also: IBM UQ360 will help measure AI uncertainty
The study is part of Neogi's broader view of how the realm of very small things, such as atom wiggling, can help build new computers that are more efficient or even computers that are inspired by the human brain. Artem Pimachev, a research associate in aerospace engineering at CU Boulder, also participated in this study.
Source of information: azorobotics.com