Tang, an assistant professor of chemical and biomolecular engineering, will receive a prestigious National Science Foundation Early Career Development (CAREER) award of over $500,000 to perform revolutionary research in process systems engineering and provide novel educational experiences for NC State students.

This project aims to fundamentally transform how engineering systems achieve optimal performance and to develop a generic, data-driven framework to analyze and improve the dynamics of global optimization solvers. 

Process systems engineering is the area of chemical engineering that studies the modeling, design, operations, and control of processes. It uses computing technology and tools from applied mathematics, including optimization methods, control theory, and machine learning. This CAREER project adopts a dynamics and control perspective and uses the tools from machine learning and operator theory to analyze and improve algorithms. Such an operator-theoretic approach for the analysis of global optimization algorithms has not been systematically studied before.

“As a process control researcher, I believe that high-quality engineering decision making is dependent not only on the wide use of automated solvers and algorithms, but also on the engineers’ understanding of their mechanisms and formulations,” Tang said. 

Tang will use machine learning on large datasets to estimate the internal mechanisms of these solvers. This will allow human engineers to interpret the solvers’ behavior and enable the development of programs that can intelligently self-train for optimal performance. The project’s ultimate goal is to provide a comprehensive, end-to-end approach for the self-improvement of solvers, which could significantly accelerate new product discovery, enhance industrial automation, and boost the competitiveness of existing engineering systems.

In addition to the research, Tang’s educational goals are to improve the computational skills and data literacy of NC State engineering students. He plans to coordinate a set of advanced electives in computational and data sciences, develop courses on machine learning and optimization, and establish an engineering concentration in these fields. 

“I want to better prepare students for data-driven engineering careers, and connect classroom learning with cutting-edge research,” he said.  “Living in an era of artificial intelligence, engineers enjoy the unprecedented convenience of modern toolsets. Our challenge, as educators, is to ensure that they also can think critically and know what lies behind these toolsets.”