Artificial intelligence’s (AI) visual capabilities extend past those of the human eye.
This makes AI a powerful tool for detection.
Dr. Ridwan Sakidja, CNAS, and Dr. Sheryl Brahnam, COB, want to see AI used in many new ways.
About their research
Sakidja and Brahnam propose using machine learning (ML) to examine materials for defects. This includes composites or high-strength alloys.
Electron microscopes can capture high resolution images of materials. These images can help expose materials’ interior parts.
Yet certain flaws, like micro cracks, are still nearly impossible to see.
“Currently, researchers devote a great deal of time to tedious visual inspections to locate defects,” Sakidja said.
AI helps identify key features in images with less time and effort.
Sakidja stresses that AI has made significant progress in recent years.
When used with high resolution imaging, AI offers even greater detection benefits.
“With our project, we aim to merge these two technological advances,” Sakidja said.
The NASA-sponsored Missouri Space Grant Consortium (MOSGC) recently awarded their project $10,000.
Upgrading technology
Merging AI and high-resolution imaging requires advanced technology for acquiring images.
Last year, the provost’s office and CNAS provided Sakidja’s department with funding for new equipment.
“With their support, we upgraded our electron microscopy facility to a much more powerful detector,” Sakidja said. “This gives us the ability to improve the quality of our image acquisition process and image data.”
The new electron microscope equipment will not be the only upgrade to occur, Sakidja explains.
He and Brahnam will also use their project to start upgrading to a new computer system that can perform ML.
“Our timely collaboration will in part use the new and powerful AI-enabled visual technology,” Sakidja said. “It will couple this with the new imaging technology added to our electron microscope.”
Supporting students across disciplines
Sakidja and Brahnam’s project will engage undergraduate students from both colleges starting this fall. The students will learn to write codes in Python. They will also explore how to use AI for image analysis.
Students will then be able to present their work at the annual meeting of MOSGC in spring 2021.
Sakidja anticipates the project will challenge and excite students.
He is grateful for all Brahnam will contribute to their learning. She is a key leader in advancing visualization research.
“This project will be a stepping stone toward further exploring cross-disciplinary research with a common interest,” Sakidja said. “Students who engage in our research will learn the wide applications of machine learning and artificial intelligence.”
About MOSGC
MOSGC supports aerospace and space-related research of the science, engineering and technology fields.
The organization trains students at all academic levels for careers in professional fields of interest to NASA.
MOSGC offers an award annually to competing Missouri universities as affiliates.
The awarded university must match the organization’s funding through in-kind support.