An international team of researchers created picture recognition technology by utilizing AI learning to analyze solely the surface morphology of a battery, which can precisely ascertain the battery's elemental makeup and the number of charge and discharge cycles. The journal Npj Computational Materials published this research work.
![](https://dp9eps5gd5xd0.cloudfront.net/images/news/ImageForNews_15041_17200088379411589.jpg)
Example images of true cases and their grad-CAM overlays from the best trained network. Image Credit: Korea Advanced Institute of Science & Technology
President Kwang-Hyung Lee announced that Professor Seungbum Hong of the Department of Materials Science and Engineering, working with Drexel University and the Electronics and Telecommunications Research Institute (ETRI) in the United States, had developed a method using convolutional neural networks (CNN) that can predict the major elemental composition and charge-discharge state of NCM cathode materials with 99.6% accuracy.
The research team observed that although scanning electron microscopy (SEM) is commonly used in semiconductor manufacturing to inspect wafer defects, it is seldom used for battery inspections.
In battery research, SEM is primarily used to analyze particle size at research sites. Reliability is assessed based on the broken particles and the shape of breakage in deteriorated battery materials.
The research team determined that it would be revolutionary if an automated SEM could be used to examine the cathode material's surface during the battery production process, similar to how semiconductors are made, to ascertain whether the material was synthesized in accordance with the intended composition and whether its lifespan would be dependable, hence lowering the defect rate.
To forecast the key elemental composition and charge-discharge cycle phases of the cathode materials, the researchers trained a CNN-based artificial intelligence system that may be used in autonomous vehicles to learn the surface images of battery materials. They discovered that the approach could predict additive-containing material compositions with high accuracy, but it was less successful in predicting charge-discharge states.
Image Credit: asharkyu/Shutterstock.com
The team intends to utilize AI to eventually check for compositional homogeneity and forecast the lifespan of next-generation batteries by further training them with a variety of battery material morphologies produced through diverse procedures.
In the future, artificial intelligence is expected to be applied not only to battery materials but also to various dynamic processes in functional materials synthesis, clean energy generation in fusion, and understanding foundations of particles and the universe.
Joshua C. Agar, Professor and Study Collaborating Researchers, Department of Mechanical Engineering and Mechanics, Drexel University
Seungbum Hong, Professor and study lead, stated, “This research is significant as it is the first in the world to develop an AI-based methodology that can quickly and accurately predict the major elemental composition and the state of the battery from the structural data of micron-scale SEM images. The methodology developed in this study for identifying the composition and state of battery materials based on microscopic images is expected to play a crucial role in improving the performance and quality of battery materials in the future.”
This research was conducted by KAIST’s Materials Science and Engineering Department graduates Dr. Jimin Oh and Dr. Jiwon Yeom, the Co-First authors, include Professor Josh Agar and Dr. Kwang Man Kim from ETRI.
The research was funded by the National Research Foundation of Korea, the KAIST Global Singularity project, and international collaboration with the US research team.
Journal Reference:
Oh, J., et al. (2024) Composition and state prediction of lithium-ion cathode via convolutional neural network trained on scanning electron microscopy images. Npj Computational Materials. doi.org/10.1038/s41524-024-01279-6.