(7gv) Advanced Control for Next-Generation Materials Synthesis and Smart Manufacturing

Strong trends in chemical and biomolecular engineering have led to a rapid increase in complexity of the design and manufacturing of high-tech products. Systems engineering (including computer modeling, simulation, optimization, and control) encompasses a powerful set of tools that can be used to help overcome the many challenges of these complex systems that include: (1) nonlinear (multi-scale) physical phenomena occurring on a large range of length and time scales, (2) mixed continuous and discrete dynamics due to the presence of time delays, interconnected units, and component faults/failures, (3) process uncertainties, disturbances, and noise, and (4) tight product specifications and safety-critical constraints.

During my Ph.D. research at the Massachusetts Institute of Technology (MIT) (with advisors: Prof. Richard Braatz and Prof. Michael Strano), I developed practical optimization-based methods that are able to address many of these challenges for a wide-range of complex systems using first principles models in combination with advanced statistical techniques. I have applied these methods to a number of emerging applications including continuous pharmaceutical manufacturing, advanced manufacturing of biological drugs, carbon nanotube-based solar cells, and the templated assembly of nanocrystals. Specifically, I have shown the impact of uncertainty on standard model-based control methods and established ways to systematically account for this uncertainty in the design of system-wide control structures.

Throughout my postdoctoral work at the University of California, Berkeley (with advisor: Prof. Ali Mesbah), I have been expanding the scope of my research toward autonomous “fault-tolerant” model predictive control schemes. The main application focus has been on plasma processes for both medical therapy and semiconductor fabrication. The large number of reactions taking place in the plasma and their non-equilibrium nature make the models far too complicated to be directly used for optimization. In this work, more tractable models can be derived by applying identification and machine learning methods to data generated from the complex first principles calculations.

In my future research program, I am interested in using my systems engineering expertise to bridge the gap between the synthesis/design of next-generation materials (e.g., pharmaceuticals, or catalysts) and the development of processes for their smart manufacturing. This involves cooperative research toward (1) materials design by explicit modeling of molecular- and atomic-scale interactions, (2) efficient material screening through machine learning and reduced order modeling, and (3) the development of tractable methods for simultaneous process optimization and advanced control (across multiple spatial and temporal scales) in the presence of uncertainty. These methods can be applied to a wide-class of complex chemical and biomolecular systems including battery systems, (bio)pharmaceuticals, power/smart-grid networks, advanced biofuel production, and catalytic chemical processes. Based on my background in applied mathematics, I also intend to develop application-oriented theory and numerical methods.

Teaching Interests:

I have extensive experience as a teaching assistant (TA) both as an undergraduate at UT Austin and as a graduate student at MIT. At MIT, I TA'ed the Numerical Methods in Chemical Engineering graduate course. For my efforts as a TA, I received the Outstanding Graduate Teaching Assistant Award from the Chemical Engineering Department and the campus-wide 2015 School of Engineering Award for Extraordinary Teaching and Mentoring from the Dean of Engineering. I am particularly interested in advancing the computing, numerical methods, and control skills of chemical engineering students. In addition, I am very interested in teaching transport phenomena, thermodynamics, unit operations, and applied mathematics.