Skip to main content

NCSA welcomes 2019-2020 Faculty Fellows

The National Center for Supercomputing Applications (NCSA) has named seven new Faculty Fellows for the 2019-2020 academic school year. The NCSA Faculty Fellowship is a competitive program for faculty and researchers at the University of Illinois at Urbana-Champaign which provides seed funding for new collaborations that include NCSA staff as integral contributors to the project.

The Big Picture: Media, Capital and Networks of Influence

Faculty Fellow: Rini Mehta (Dept. of Comparative and World Literature, College of Liberal Arts & Sciences)
NCSA Collaborators: Kalina Borkiewicz, Sandeep Puthanveetil Satheesan, Luigi Marini

Abstract: The Big Picture proposes to build a map of the global network of media, corporate power, and political influence that was engendered by globalization and which in turn continues to shape our world in the 21st century. The world that we inhabit today is caught in the interstices of political, economic, and cultural forces that operate in between fault lines of nations, regions, and ideologies. Visual technologies and arts are as much complicit in manufacturing and manipulating history as they are involved in disseminating history as it unfolds. Our project will connect realpolitik and representation to capture a media history of the current times, in a global context. Using methods gleaned from the ontology-based models such as FOAF (friend of a friend), The Big Picture will bring together media studies, history, statistics, data curation, and advanced visualization to produce a dynamic interface accessible through a website. This project will partner with the PI’s Global Film History from the Edges project that has been granted $150,000 by the University of Illinois Presidential Initiative for the Advancement of Humanities and the Arts.

Predicting International Food Security Crises: A Data-Driven Approach

Faculty Fellow: Hope Michelson (Dept. of Agricultural and Consumer Economics, College of Agricultural, Consumer, and Environmental Sciences)
NCSA Collaborators: Liudmila Mainzer, Aiman Soliman

Abstract: In food crises, faster and more accurate evaluation and response can save lives and resources. Methods currently in use to predict such crises have limitations that delay and impede humanitarian response: they are not model-driven, and they do not engage the full scope of available data. Because government policy-makers and Non-Governmental Organizations often fail to recognize specific food insecure populations, scarce resources to mitigate hunger can arrive too late and in the wrong places. In many parts of the world, crises of this sort are on the rise, requiring improved methods to identify their scale and scope. Developing and deploying an effective early warning system is urgent, given the expectation that climate shocks disrupting agricultural production and market functioning will increase in frequency and severity in coming decades.

We propose to develop and test a new model-driven method for predicting food crises across the world. Dr. Michelson’s previous work (Lentz, Michelson, Baylis and Zhou, 2018) demonstrates that we can improve prediction of food security crises by exploiting publicly-available high-frequency, spatially-resolved data. The proposed collaboration with NCSA’s Data Analytics Group and the NCSA Genomics Group will take this research to the next level: developing new data sources for predicting and applying state of the art machine learning techniques to the prediction problem.

Discovery of Advanced Nanodielectrics through AI-accelerated Multiscale Simulation from First Principles

Faculty Fellow: Yumeng Li (Dept. of Industrial and Enterprise Systems Engineering, The Grainger College of Engineering)
NCSA Collaborators: Erman Guleryuz

Abstract: The research objective of this proposal is to create a new generation artificial intelligence-enabled multiscale simulation framework and its enabling techniques, which would directly root in first principles theory, for a comprehensive understanding of the cross-scale multiphysics phenomena in dielectric polymer nanocomposites. In addition to features like easy processing and light weight, polymer nanocomposites demonstrate a great potential in realizing highly enhanced combined properties to meet the needs of advanced dielectrics in applications from energy storage to power delivery. However, the current lack of understanding in fundamental mechanisms leading to the property enhancement necessitates modeling of complex phenomena (ranging from nanoscale to macroscale) using a high-performance multiscale simulation framework. The new multiscale simulation framework employs artificial intelligence (AI) for an effective integration of first principle calculations, physics-based atomistic simulations and data-driven predictive analytics, thereby concurrently leveraging high accuracy of first principles calculations and high efficiency in AI enabled predictive data analytics. Built upon the PI’s research experience on both multiscale simulation and polymer nanocomposites, this proposal will focus on four research thrusts to address grand challenges in developing the new framework: 1) developing machine learning potentials for the interface based on high-throughput first principles calculations, 2) characterizing nanoscale local interfacial electro-mechanical-thermal properties using AI-accelerated atomistic simulations, 3) predicting macroscale electro-mechanical-thermal properties considering the interface effects, and 4) model validation of the proposed multiscale simulation framework.

Enabling Long-Term Reuse of Experimental and Computational Datasets on Protein Dynamics

Faculty Fellow: Diwakar Shukla (Dept. of Chemical and Biomolecular Engineering, College of Liberal Arts & Sciences)
NCSA Collaborator: Luigi Marini

Abstract: Modern molecular simulations of proteins on high-performance computing resources such as Blue Waters generate extensive atomistic-detailed information about protein dynamics, which could be leveraged for obtaining insights about molecular origin of human diseases, design of therapeutics, bioengineering of plants. However, the key challenge is to convert the terabytes of biomolecular dynamics data generated on supercomputers into a format accessible to an experimental researcher. In this proposal, we present an approach that not only generates suggestions for optimal experiments based on simulation data (e.g. for validation of simulations) but also integrates the existing experimental and simulation information to generate comprehensive models of protein dynamics that are missing from the current literature. We have developed algorithms that provide an approach that maximizes information gain for the design of experiments given simulation data. We propose to work with NCSA collaborators to implement a cloud-based platform and a user interface for this proposed service. NCSA will benefit from working on this project by gaining more expertise in applying cyberinfrastructure in the realm of biomolecular dynamics. The biggest impact of the proposed study is that it provides an accessible tool for experimental researchers to help harness the knowledge hidden in the big protein simulation datasets generated using Blue Waters and other high performance computing resources. This work will have a transformative impact on how protein science is conducted by experimental and computational research groups.

High-Performance, Multi-Objective, and Multi-Physics Design Optimization of Next-Generation, Patient-Specific Implant Scaffold at Scale

Faculty Fellow: X. Shelly Zhang (Dept. of Civil and Environmental Engineering, The Grainger College of Engineering)
NCSA Collaborator: Erman Guleryuz

Abstract: With recent advances in tissue engineering, the design and fabrication of implant scaffolds have become emerging areas of research, as the traditional implants fail to fulfill required functionalities for specific patients. While topology optimization offers a promising method for scaffold design, existing studies have limited capabilities of addressing multiple design scenarios and fine control of the porosities to achieve highest performances. To address these challenges, the proposed research aims to create a high-performance, multi-physics, and multi-objective topology optimization framework for the design of next-generation patient-specific scaffolds implant scaffold with enhanced multifunctionality. The proposed formulation addresses both mechanical and mass transport design requirements using multi-objective formulations and simultaneously controls the location, size, and shape of porosities through local constraints. To successfully realize the high complexity of the scaffold structures, the proposed research requires large problem size (hundreds of millions of degrees of freedom) and 3D multi-physics simulations, which must rely on massively parallel supercomputers. The PI will work closely with NCSA to develop highly scalable algorithms and high-performance computational frameworks for efficient optimization and to utilize the large-scale supercomputers in order to achieve ultra-high-resolution designs.

The proposed work will be built upon an open-source parallel code based on PETSc suite of libraries. Through a proof-of-concept benchmark on Blue Waters, the workflow was successfully tested and showed great scalability. The supercomputing infrastructure and domain experts at NCSA will provide essential support for the success of this project. The state-of-the-art methods created in this research will carry a great potential to contribute to the synergy between NCSA and the members of NCSA’s Industry Program from the life sciences sector. With the optimized structures developed through this project, patients implanted with the optimized scaffolds would be benefited from better functionality, better clinical results, and ultimately contribute to better health and living conditions.

The War on Professional Expertise: The Global Spread of Online Myths about Medicine and Health

Faculty Fellows: Kevin Leicht (Dept. of Sociology, College of Liberal Arts & Sciences), Brant Houston (Dept. of Journalism, College of Media)
NCSA Collaborator: Loretta Auvil

Abstract: The spread of dubious or downright false information (sometimes referred to as “fake news”) is a growing social, cultural and scientific dilemma, and the situation is especially troubling when it comes to information about medicine and public health. The most recent manifestation of the real world consequences of dubious medical information is the spread of measles and its link to anti-vaccination websites and memes. But that is only the most recent manifestation—others include the peddling of conspiracy theories and fake cancer cures, organized misinformation about stem cell research, and the spread of dubious claims about alternative medicines. There is further evidence that some of this dubious information is deliberately produced for financial gain or to fuel cultural discord.

The purpose of this project is to examine the routes through which medical misinformation spreads in the news and social media. The research will examine medical misinformation in four areas; (1) vaccinations, (2) cancer cures, (3) the spread of the Ebola virus, and (4) the safety of contraception. Misinformation is defined as information that is publicly available and disseminated that is not supported or actively contrary to established medical advice. For this fellowship, we will use the considerable news resources of the Cline Center archive and then use the resources and expertise of NCSA to (1) explore methods for searching and applying models to identify relevant new articles on our selected healthcare topics, (2) develop a model for identifying dubious and false information in these articles, (3) rendering the data suitable for quantitative analysis, and (4) aiding the principal investigators in conducting the analysis. The pilot research from this fellowship will form the basis for a much larger research grant to be submitted to the Knight Foundation or the National Institutes of Health.

Disclaimer: Due to changes in website systems, we've adjusted archived content to fit the present-day site and the articles will not appear in their original published format. Formatting, header information, photographs and other illustrations are not available in archived articles.

Back to top