The W. A. "Tex" Moncrief Grand Challenge Awards are given to researchers that aim to find solutions to society’s biggest, most important questions. Named in honor of one of the Oden Institute’s greatest benefactors, the late W.A. “Tex” Moncrief was certainly not afraid of tackling any challenge he faced throughout his long and successful career.
The awards are chosen based on their highly compelling research proposals related to challenges affecting the competitiveness and international standing of the nation.
Successful applicants receive a stipend of up to $75,000 to cover the costs necessary for focusing on research and academic programs related to Grand Challenges in computational engineering and sciences.
The 2022 Grand Challenge Awards
Ann Chen, assistant professor in the department of Aerospace Engineering and Engineering Mechanics, will use her award to monitor and characterize groundwater aquifers from satellite remote sensing measurements. She will integrate large volumes of this Interferometric Synthetic Aperture Radar (InSAR) data with a 3D geomechanical model to infer aquifer properties in places where direct head data are not available. Chen’s goal is to characterize aquifers using satellite data and inform Aquifer Storage and Recovery (ASR) managers of the optimal pumping/recharge schedule and placement of wells. Chen believes that successful implementation of ASR at the H2Oaks facility in Twin Oaks, Texas will boost drought resiliency and protect endangered local species, and she hopes to see the proposed computational framework adopted by ASR facilities in other cities.
Graeme Henkelman, professor of Chemistry, will model the function of materials at the atomic scale for energy applications and design new materials with desired properties. Funding from the Grand Challenge Awards will make it possible for Henkelman to spend the majority of the Spring 2023 semester at a three-month program at the Institute for Pure and Applied Mathematics (IPAM) of UCLA. He believes that this program, entitled: “New Mathematics for the Exascale: Applications to Materials Science” will be a perfect fit for a primary research direction of his group: redesigning computational methods and code to take advantage of massively parallel and heterogeneous computing architectures. Henkelman is confident that his research group will be able to make progress on this front and is excited about the possibility of making substantial improvements in computational methods for use on exascale computing resources.
Per-Gunnar Martinsson, professor of mathematics, will use his award to develop faster and more energy-efficient algorithms for one of the most fundamental tasks in computational science: solving large systems of coupled linear equations. The fundamental innovation behind his group’s approach is to use randomized sampling and randomized averaging, to reduce the effective dimensionality of the problems to be processed. Martinsson hopes that the opportunity to fully focus on this project for an extended period will also have benefits in terms of building collaborations, as this research is being carried out in a collaborative framework with Rachel Ward of the Oden Institute, Vladimir Rokhlin of Yale, Joel Tropp of Caltech, and Yuji Nakatsukasa of the University of Oxford.
Rachel Ward, also a professor of mathematics, aims to enhance state-of-the-art randomized algorithms by reducing hyperparameter sensitivity and accelerating convergence to be compatible with real-world learning problems in scientific machine learning. Her work identifies “pain points” in standard randomized algorithms and outlines an approach for removing the pain using a combination of tools from probability, optimization and linear algebra. Two specific tasks of this form will be the incorporation of automation of momentum in stochastic gradient descent and the construction of fast and adaptive random Fourier features.
Tom Yankeelov, professor of Biomedical Engineering, Diagnostic Medicine and Oncology, is developing advanced mathematical methods for predicting and optimizing the response of individual breast cancer patients to pre-surgery treatment: neoadjuvant therapy (NAT). Yankeelov wants to create accurate and rapid methods for predicting the response of breast tumors to NAT so that treatments can be customized on an individual patient-by-patient basis. Yankeelov’s group will simulate a range of interventions via patient specific digital twins, and then employ the methods of optimal control theory to identify personalized therapeutic regimens designed to dramatically outperform the standard-of-care treatment protocols. In this way, they will be performing clinical trials on individual patients using the methods of computational science. This project represents a transformative approach for predicting and optimizing the outcomes of cancer treatment on an individual patient basis before therapy commences.