Scientist Presents Out of this World Research at International Conference
The Society for Industrial and Applied Mathematics (SIAM) hosted this week its biennial activity group conference on computational science and engineering (SIAM CSE23) in Amsterdam.
There, nearly half of the faculty body from Georgia Tech’s School of Computational Science and Engineering (CSE) had research accepted for presentation to fellow members of the world’s largest organization devoted to applied mathematics, computational science, and data science.
One notable, literally out of this world, presentation comes from Elizabeth Qian, an assistant professor with joint appointments in the School of CSE and the Daniel Guggenheim School of Aerospace Engineering.
Qian presented at the conference a new computational framework for engineering analysis that was applied to the James Webb Space Telescope (JWST). The method proved to reduce the time required to perform a design analysis from more than two months to less than two days.
In addition to reducing the time required to perform this analysis, the framework also makes results more consistent and robust. These technical performance improvements can help keep complex space missions, like the JWST, on schedule and on budget, contributing to overall mission success.
“Working with domain experts at NASA on the analysis of a space system is very exciting,” Qian said. “There are unique challenges that are encountered in the design of real-world systems that we don’t encounter when prototyping methods.”
Qian’s framework speeds up and reduces error of uncertainty quantification calculations. Uncertainty quantification is the science of accounting for uncertainty and determining the likelihood of certain outcomes if some aspects of a system are unknown.
Since uncertainties are present in most complex, large-scale engineering systems, like the JWST, this framework provides design methods and tools that can be applied to many other projects here on Earth.
“This framework is already having an impact on applications from plasma fusion to water resource management,” said Qian. “The beauty of CSE research is that it can be applied to almost anything in science, engineering, and medicine.”
Some uncertainties engineers must account for when designing the JWST are how temperature changes affect the telescope’s optics when it slews from one observation target to another. These temperature changes can distort the telescope’s images.
To overcome this challenge, Qian teamed with NASA engineer Giuseppe Cataldo and Jeremy Auclair, a research engineer at Centre d’Etudes Spatiales de la Biosphère in Toulouse, France.
In their study, the new method analyzed the optical error caused by temperature changes and identified which telescope structures had the largest impact on this error.
Tested through thousands of rigorous simulations, the team’s framework uses a mix of low and high-fidelity models to reduce error and improve calculation speed.
While less accurate, low-fidelity models run faster and improve the framework’s overall computational speed. A more expensive high-fidelity model is used in the framework to ensure accuracy of results and retain the physics of the phenomena the model is trying to represent.
Grounded upon interdisciplinary collaboration and poised for real-world application, the framework made for worthy scholarship accepted at SIAM CSE23. There, Qian represented the research team and the School of CSE when she gave a presentation of the paper.
The School of CSE formed a strong contingent of 11 presenting faculty at SIAM CSE23, nearly half of the School’s faculty body at a single conference.
Along with Qian, School of CSE faculty with papers accepted at the conference include:
- Spencer Bryngelson
- Ümit Çatalyürek
- Nisha Chandramoorthy
- Peng Chen
- Elizabeth Cherry
- Edmond Chow
- Felix Herrmann, joint with the School of Earth and Atmospheric Sciences and School of Electrical and Computer Engineering
- Surya Kalidindi, joint with the George W. Woodruff School of Mechanical Engineering
- Florian Schäfer
- Rich Vuduc
SIAM CSE23 is the designated, biennial conference for the SIAM Activity Group on Computational Science and Engineering. Here, School of CSE Professor and Associate Chair Edmond Chow serves as the activity group’s program director, a position he recently attained in December 2022.
School of CSE Regents' Professor and Chair Haesun Park served on the selection committee for for the James H. Wilkinson Prize for Numerical Software, which was awarded at the conference.
Another notable School of CSE highlight at the conference was the presence of Assistant Professor Peng Chen. Chen presented one paper that he authored, he co-authored five additional papers, and he organized two panels.
On the final day of SIAM CSE23, the conference announced that School of CSE Associate Professor and Associate Chair Elizabeth Cherry would co-chair the organizing committee for SIAM CSE25. The 2025 conference will take place March 2 – 7 in Fort Worth, Texas.
Whether it is research collaboration at an international conference or application on a telescope orbiting Earth, the School of CSE is distinguishing itself as a leader in solving scientific and engineering challenges through computational methods.
“It’s great to have such a strong Georgia Tech representation at the conference. It really points to Tech being an excellent place for impactful, interdisciplinary work,” Qian said.
A wrongful arrest. A “racist robot.” A call for new laws.— Georgia Tech Computing (@gtcomputing) November 10, 2023
A @GeorgiaTech experiment trained a robot to seemingly act out racist behavior, to prove bias can exist in #AI. @MatthewGombolay opens up his lab to show where research can help address tough social issues. https://t.co/21F7IV0vbH pic.twitter.com/P3GD29lth1