Large-Scale Multidisciplinary Design
Optimization of Aerospace and Robotic Systems
John Hwang
Associate Professor | Department of Mechanical and Aerospace Engineering | University of California San Diego
Thursday, November 20
11am - 12pm
Guggenheim 442
About the Seminar:
The design of complex engineering systems such as aircraft and robots is dominated by discipline-partitioned workflows that rely on expert judgment and use simulations mainly to evaluate a few candidate concepts rather than systematically explore the full design space. As these systems grow more complex and performance-critical, this paradigm leaves critical trade-offs underexplored and leads to suboptimal or overly conservative designs. Multidisciplinary design optimization (MDO) offers a formal alternative that uses numerical optimization to solve design problems involving multiple engineering disciplines. Large-scale MDO focuses on high-dimensional problems (dozens or more design variables) which are typical in system-level design, where time- and space-dependent fields are parametrized, subsystems from multiple disciplines are optimized together, and control variables across many operating conditions are included in a single MDO problem. Over the past decade, large-scale MDO has become tractable thanks to advances in automatic differentiation and modern MDO software frameworks. This seminar will highlight recent work aimed at making large-scale, system-level MDO practical for industry-relevant applications, focusing on three areas: (1) graph-based modeling, a new computational modeling paradigm that automates adjoint sensitivity analysis and accelerates model development and execution; (2) new optimization algorithms and theory tailored to the structure of large-scale MDO problems; and (3) cross-cutting modeling infrastructure that simplifies assembling and coupling models across disciplines, fidelity levels, and operating conditions. These developments will be demonstrated through applications to urban air mobility vehicles, blended-wing-body aircraft, soft aquatic robots, and surgical robots.
About the Speaker:
John Hwang is an Associate Professor in the Department of Mechanical and Aerospace Engineering at the University of California San Diego. He holds a B.A.Sc. (2010) from the University of Toronto and an M.S.E. in Aerospace Engineering (2012), an M.S. in Mathematics (2013), and a Ph.D. in Aerospace Engineering (2015) from the University of Michigan. Prior to joining UC San Diego in 2018, he was a research engineer at NASA Glenn Research Center. Hwang specializes in large-scale multidisciplinary design optimization (MDO), developing theory, algorithms, methods, and software for efficient system-level design of complex engineering systems in aerospace, robotics, and wind energy. He is a recipient of a 2022 DARPA Young Faculty Award and a 2025 Presidential Early Career Award for Scientists and Engineers. He leads a $4.8M research center funded by the Air Force Research Laboratory (AFRL) focused on large-scale MDO of military aircraft and recently led a $5.8M NASA University Leadership Initiative (2021–2025) project on large-scale MDO for urban air mobility vehicle design. His research has been supported by AFRL, NASA, ONR, DARPA, NSF, General Atomics, Samsung, and Hyundai Motor Company.