![Ying Lin [left] and Qianmei (May) Feng from the Industrial Engineering Department at the Cullen College of Engineering have received a grant from the University of Houston's Advanced Manufacturing Institute (AMI) to use machine learning techniques to understand and improve the manufacturing process for superconductors.](/sites/ccoe.egr.uh.edu/files/images/news/2023/ying-lin-and-may-feng-01-crop.jpg "Ying Lin [left] and Qianmei (May) Feng from the Industrial Engineering Department at the Cullen College of Engineering have received a grant from the University of Houston's Advanced Manufacturing Institute (AMI) to use machine learning techniques to understand and improve the manufacturing process for superconductors.")
A pair of professors from the Industrial Engineering Department at the Cullen College of Engineering have received a grant from the University of Houston's Advanced Manufacturing Institute (AMI) to use machine learning techniques to understand and improve the manufacturing process for superconductors.
“Machine Learning-based Process-Structure-Property (PSP) Modeling and Monitoring for Superconductor Manufacturing” is funded for $35,000. The AMI supports the transition of lab-scale technology to fully-fledged manufactured products for the market, and addresses manufacturing challenges by creating solutions in manufacturing R&D.
Ying Lin, an associate professor in the Industrial Engineering Department, is the lead PI for the project. Lin is also the director of the Smart Health & Intelligent Engineering Systems (SHINES) Lab. Qianmei (May) Feng – a professor, a Brij and Sunita Agrawal Faculty Fellow and the Graduate Program Director in the Industrial Engineering Department – is a co-PI for the project.
According to an abstract for the project, their research is focused improving the manufacturing process for high-temperature superconductors (HTS).
“To achieve high yield and improved performance in superconductor manufacturing, this project aims to develop a machine learning-based process-structure-property (PSP) modeling and monitoring system with enhanced interpretability and reproducibility,” the authors wrote.
“The objective of this proposal is to fill these gaps by achieving two specific aims: 1) Construct a novel interpretable machine learning model (i.e., graphical models) to systematically investigate and visualize the underlying PSP relationships (i.e., A-MOCVD process parameters, sub-surface microstructure, and the uniformity) of superconductor tapes on each HTS tape; and 2) Develop a graph dictionary learning model to build a dictionary of reproducible PSP relationships across heterogeneous tapes, which can be leveraged to real-time monitor the key process parameters and sub-surface microstructure during the production of new HTS tapes.”
The research was started earlier this year, and will continue through the end of the year. The grant continues earlier research that the pair also received AMI funding for.