University Of Illinois At Urbana-Champaign

NSF Awards · FY 2025 · 2025-09

Geoscientific computational models are used to predict a wide range of natural processes such as weather, water supply, air pollution, eruptions, floods, and tsunamis. Progress in geo-modeling is held back by the need for geoscientists to also acquire expertise in computer science. The project will introduce a novel modeling workflow to overcome this hurdle. This project supports the development and community adoption of a symbolic equation-based modeling system. This system separates automates the numerical processing optimization so geoscientists can focus on the equations that describe natural processes. This system also advances the use of artificial intelligence in geosciences for simplifying model reduction and parameterization. The proposed work consists of three thrusts involving 1) community organization, 2) model development, and 3) de-centralized model management and education. A series of workshops at major international geoscientific meetings domestically and abroad will help define project priorities. Models will be documented on a dedicated website, run by a decentralized governance system and supported with interactive educational experiences to transition to a user-supported network for long-term growth. Geoscientific computational models simulate natural processes such as weather, water supply, and air pollution; for analyzing risks of volcanoes, floods, and tsunamis. The proposed project will introduce a new process of geoscientific model development, where model components and their interrelationships are specified as a system of equations that a compiler automatically transforms into a computer model. By separating the model design (the equations) from model implementation (the code compiler), geoscientists can focus on building equation systems that represent their areas of expertise, greatly increasing the participation in geoscientific modeling. This system will also provide an ideal base for integrating AI model reduction and parameterization into the geosciences. Project activities are divided into three thrusts. Thrust 1 will convene a series of workshops to create a shared roadmap for model development at major international geoscientific meetings domestically and abroad. Thrust 2 will expand on the types of systems that can be studied with equation-based models by implementing model components and capabilities as prioritized by project members and workshop participants. Model capabilities will be documented on a dedicated website. Thrust 3 will implement a decentralized governance system and interactive educational experiences to transition to a user-supported network of equation-based modelers. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Understanding the forest understory is essential for effective forest management, biodiversity conservation, wildfire prevention, and environmental monitoring. However, traditional satellite and aerial remote sensing technologies are fundamentally limited in their ability to observe vegetation beneath the forest canopy due to occlusion and signal attenuation. As a result, critical indicators of ecosystem health, such as aboveground biomass carbon pools, combustible understory fuel loads, and other measures of biodiversity in the understory, remain largely unmeasured at scale. This project addresses this critical gap by envisioning a low-cost, scalable sensing system that uses radar and wireless communication technologies to detect and characterize the forest understory, complementing existing orbital and suborbital remote sensing systems. The project advances radar sensing and physics-aware modeling by demonstrating an IoT-powered forest observatory in real-world scenarios, and creates new educational and outreach materials, including open-source software and student training modules, to broaden the impact and accessibility of the research. The project outlines a new approach that bridges radar sensing and backscatter communication for environmental sensing via a novel channel modeling approach through vegetation and generalizable physics-aware models that characterize the effect of biomass on radar signals. The key intuition is that vegetative dielectric and moisture content will alter the RF signatures in both frequency and time domains as they penetrate through the vegetation medium. These signatures can be learned using complex physics-aware models as long as the RF reflections that carry this signature can be reliably separated, modeled, and interpreted. The project team will realize this vision through four inter-connected tasks: (i) formalizing the radar backscatter through multi-layer forests, which results in a radar forest synthesizer (ii) proposing physics-aware radar backscatter models that offer spatial characterization and mapping of forest understory biomass (iii) introducing an RF-coded tag design that can pair up with off-the-shelf radar platforms and serve as ground references for accurate understory characterization; (iv) Fully evaluating the end to end system in wide-area testbeds and demonstrating the accuracy of characterizing forest understory. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

This project aims to develop and test a novel human-cognizant decision-making framework for bridge maintenance that integrates predictive modeling and institutional trust, behavioral adaptation, and participatory insight to improve infrastructure resilience and expand long-term economic opportunity. Bridge maintenance decisions have wide-reaching societal consequences – affecting public safety, job accessibility, freight mobility, and economic productivity – yet current models often overlook the human and institutional dynamics that shape those outcomes. By embedding decision-maker trust in machine learning models for bridge deterioration prediction, modeling how individuals and businesses adapt to disruptions, and incorporating stakeholder input into maintenance prioritization, this research transforms how infrastructure decisions are made. The project supports the progress of science and engineering and serves the national interest by informing infrastructure strategies that enhance safety and deliver greater economic impact for the public. Technically, the project advances four key innovations: (1) it reframes trust as a measurable design element of machine learning models by testing how model explainability, data quality, and uncertainty influence adoption by public-sector agencies; (2) it introduces modeling of behavioral adaptation among commuters and businesses in response to bridge maintenance disruptions, drawing on theories of risk, habit, and loss aversion; (3) it develops formal participatory models that assess how public input can enhance maintenance prioritization under budget constraints; and (4) it integrates these insights into an opportunity-sensitive planning framework that quantifies access and economic impacts alongside engineering performance. The research uses multimodal bridge deterioration models, stakeholder interviews and experiments, behavioral simulations, and scenario-based evaluation to test its framework. By combining civil engineering, behavioral science, and decision theory, the project advances the analytical foundations of infrastructure planning and delivers tools to support more adaptive and economically effective decisions for transportation agencies and communities nationwide. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Non-technical Abstract Molecules can arrange to form larger structures, a process that is key to both complex living tissues and new, advanced materials. For example, scientists have long studied how specific sequences of amino acids fold to create proteins that act as tiny machines. Similarly, surfactants (e.g., the molecules in soap) can assemble into spheres, layers, and tubes. In both cases, the assembled structure is important for their practical use. For example, long, tube-shaped surfactant structures help to thicken shampoos while also cleaning hair. However, the ability to form this tube-like structure is usually related to shape of the surfactant molecule itself. This project seeks to learn from the ways in which long, charged molecules with protein-like sequences attract oppositely-charged surfactants, and form materials with desired structures. This effort uses both experiments and computation and will benefit society and the U.S. by establishing a versatile class of biology-inspired materials for use across chemical, agricultural, and industrial applications. The research will also involve the interdisciplinary training of researchers with broad expertise in chemistry, engineering, and physics, via both student mentorship and engagement with K-12 students. Technical Abstract This project will establish how sequence-controlled polymers can be used for the rational design of surfactant-containing materials. This effort will leverage sequence-defined polypeptides to modulate the assembly of surfactants into a variety of different nano-scale structures. The resulting materials will be evaluated by optical and electron microscopy, as well as scattering and rheological methods, to determine the relationship between polypeptide sequence and assembled structure. These experimental aspects will be integrated with a modeling effort that connects molecular simulation, colloid science, and polymer field theory to obtain predictions of assembly in polyelectrolyte-surfactant complexes. The overarching goal is to establish a fundamental understanding of how sequenced polypeptides can be used to manipulate the nano-scale structure of bioinspired surfactant-based assemblies. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-09

ABSTRACT Chirality, the property of being non-superimposable on its mirror image, plays a pivotal role in biological systems. Homochirality, the exclusive use of one chiral form, shapes the structure and function of vital biomolecules like proteins and DNA. This uniform chirality raises profound questions about life’s nature and origins. However, current research faces challenges in bridging molecular and cellular understanding. Existing methods, while high in resolution, often lack the capacity to provide a comprehensive view at larger cellular scales. Conversely, tools designed for cellular-to-tissue analysis typically overlook finer molecular details. This gap hinders our full understanding and manipulation of chirality in complex biological contexts, highlighting the need for versatile tools capable of comprehensively exploring chirality across these varying scales. Central to this effort is the development of cutting-edge nanophotonic technologies tailored for probing and manipulating chirality. These advancements will enable high-resolution, label-free analysis of molecular and cellular chirality, potentially revolutionizing our comprehension of chirality in biological systems and impacting drug development and disease treatment. This research aims to provide groundbreaking insights into the interplay between molecular and cellular chirality, specifically how they affect cellular functions and disease mechanisms. It will tackle key questions about the fundamental connections between cellular and molecular chirality, define and measure molecular chirality at extremely low concentrations, and explore how altering cellular chirality could impact molecular- level mechanisms and cellular functions. The anticipated outcomes include the creation of novel diagnostic tools, safer chiral drugs, and non-invasive treatment methods, significantly advancing chirality research and bridging the divide between molecular intricacies and cellular complexity.

NSF Awards · FY 2025 · 2025-09

With the support of the Chemical Measurement and Imaging Program in the Division of Chemistry, Professor Mary Kraft of the University of Illinois Urbana-Champaign is developing new computational visualization and analysis tools for depth profiling secondary ion mass spectrometry (SIMS) data. The distributions of various molecules within a sample influence its properties and function. Thus, knowledge of the molecular distribution within a sample facilitates understanding the material’s functions, and ultimately, designing new functional materials with advanced capabilities. Three-dimensional (3D) images of the molecular distributions within a sample may be acquired with depth profiling secondary ion mass spectrometry (SIMS). However, the 3D SIMS images acquired from contoured samples are distorted along the z-axis, and this distortion hinders their interpretation. Professor Kraft and her team plan to tackle this problem by developing software tools that use only SIMS depth profiling data to correct this distortion and produce accurate 3D images of component distribution within contoured samples. This research will entail the development of computational strategies for extracting height information from secondary ion intensities collected from organic and inorganic materials, accuracy assessment algorithms, and user-friendly interfaces that facilitate usage of these tools by others in academia and industry. The proposed research will produce new software tools that enable the user to create accurate component-specific 3D SIMS images of contoured organic and inorganic samples using any SIMS depth profiling dataset. These new computational tools would improve both the accuracy of SIMS 3D depth profiling images and their interpretation. Ultimately, this research could increase understanding of structure-function relationships in biological and synthetic materials, which could enable the design of new materials with advanced functions. The project will train graduate and undergraduate students in materials characterization using SIMS depth profiling, MATLAB coding, computational data visualization, image processing, image interpretation, semiconductor fabrication and characterization, cell culture, and polymer synthesis. This training will prepare students for careers in our semiconductor, electronics and pharmaceutical sectors. The computational tools and SIMS depth profiling datasets generated by this research will serve as STEM education tools. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

With the support of the Division of Chemistry, the Office of Strategic Initiatives, and the Division of Materials Research, Prof. Ying Diao and collaborators at the University of Illinois at Urbana-Champaign and the University of Washington will develop novel approaches to imparting materials with the ability to control the electron spin. This project will focus on unveiling fundamental design rules for chiral emergence from achiral semiconducting polymers. By leveraging the ability of chiral structures to control the electronic spin, the research team aims to modulate chemical reaction pathways to increase the efficiency of electron transfer and energy transduction. The fundamental knowledge to be gained from this research will inform future design of electronics and redox active materials critical for advancing semiconductor technologies. The multidisciplinary research training of students and the educational activities aim to increase the STEM workforce. The educational activities through the high school summer camp and the Electrochemical Bootcamp will be on the topics of chirality, self-driven labs, directed assembly, polymer sciences and electronics, which are aligned with the research program. This research project seeks to revolutionize electronic and energy materials by introducing supramolecular chirality and thus imparting materials with the ability to control the electronic spin. This will be achieved through merging distinct areas of research: semiconducting polymers, chirality-induced spin selectivity, and electrocatalysis, which represent three separate topics with minimal overlaps to date. This project will pursue a two-pronged approach complementing hypothesis-driven discoveries (Aim 1) with data-driven autonomous experimentation (Aim 2) to discover chiral emergence from achiral high performance redox active conducting polymers. This approach will circumvent the complex synthetic challenge for inducing chirality while also meeting the demanding requirements on electronic and redox properties. This research will test the hypothesis that spin momentum control conferred by chiral structures can serve to precisely and dynamically modulate chemical reaction pathways. The research team will demonstrate this concept through an example of green hydrogen peroxide production through electrocatalysis (Aim 3). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Supported by the Chemical Theory, Models and Computational Methods program in the Division of Chemistry, Nancy Makri of the University of Illinois at Urbana-Champaign is to develop accurate and efficient real-time path integral methods for simulating quantum mechanical processes in large environments. Quantum mechanics governs the interactions among electrons and nuclei, giving rise to complex and subtle dynamics that dictate the outcome of processes of immense importance to biology, energy harvest, and quantum computers. Understanding and controlling such phenomena hinges on the availability of accurate and robust simulation tools. Traditional approaches are hindered by the infamous scaling of the quantum mechanical equations, whose solution requires computational effort that increases exponentially with the number of particles. Makri’s approach is based on Feynman’s path integral formulation of quantum mechanics, which circumvents this issue by avoiding the explicit calculation of wave functions. However, the path integral encounters other severe computational obstacles, such as numerical instabilities and astronomical numbers of terms that must be included. Makri has developed a small matrix path integral (SMatPI) methodology that overcomes these problems in many important situations. She will further develop this approach, to increase its efficiency and make it suitable for simulating processes of increased complexity. A broader impact of this work will be the development of powerful simulation code that will enable fully quantum mechanical calculations on complex processes, such as the energy transfer involved in the early steps of photosynthesis. For nearly three decades, the quasi-adiabatic propagator path integral (QuAPI) methodology developed by Makri and coworkers has offered an efficient, fully quantum mechanical tool for calculating the evolution of systems interacting with harmonic bath degrees of freedom. The recent development of the SMatPI algorithm eliminates the tensor storage requirements of QuAPI, extending the capabilities of quantum simulation to systems of unprecedented size interacting with long-memory environments. The proposed work will augment the SMatPI methodology with important components that will extend its applicability to systems and environments of increased complexity. These developments will advance the frontiers of theoretical chemistry, leading to new insights and establishing a closer connection between theory and experiment. This work will lead to new, powerful computer code for quantum dynamics, which will be added to the package PATHSUM recently developed by Makri and coworkers, raising the predictive power of simulation to a new level and aiding in the interpretation of experimental results. The proposed developments will thus have a broad impact spanning chemistry, physics, biology and materials research. Powerful visualization software will aid researchers and enhance the learning experience. Makri’s research group benefits from cross-fertilization by students in chemistry, physics, and chemical engineering. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

With the support of the Macromolecular, Supramolecular and Nanochemistry Program in the Division of Chemistry, Professors Smith and Schleife of the University of Illinois Urbana-Champaign are studying a new approach to manufacturing nanocrystals. Nanocrystals are an important part of the next generation of electronics, visual displays, solar cells, and medical tests. However, they are expensive because they require new manufacturing procedures that are not yet scalable, and their production generates new types of waste products that are difficult to process. The new approach is based on “alkoxy reactions” which replace oil-based liquids used in traditional nanocrystal reactions with liquids that are more similar to water. As a result, methods can be used that have already been industrially scaled and do not generate complex waste products. The new approach can also be performed safely and at low cost in classrooms and educational labs. The project will focus on understanding alkoxy reactions, including how the nanocrystal attaches to molecules within the reactions that cause the products to have long-term stability, how the liquids influence the quality of the products at high reaction temperatures, and how the nanocrystal products can be joined to biological molecules for use in medical applications. Both experimental and theoretical approaches will occur in tandem in order to reach both a fundamental and applied understanding of alkoxy reactions. With success of the project, it is expected that low-cost nanocrystals will become more readily available for advanced devices and medical tests. With the support of the Macromolecular, Supramolecular and Nanochemistry Program in the Division of Chemistry, Professors Smith and Schleife of the University of Illinois Urbana-Champaign are studying a new synthetic method for diverse classes of nanocrystals. Current compound nanocrystals are normally synthesized in nonpolar alkane-based solvents with high boiling points but they require extensive processing for applications in polar solvents such as water. In a new approach, nonpolar solvents will be replaced with ones that are polar to enable immediate dispersion in both nonpolar and polar solvents including water. The primary focus is on the use of solvents and reagents with alkoxy functional groups such as ethers, esters, and alcohols. The project aims to understand (1) how ligands in these reactions bind to the nanocrystal surface to elicit long-term stability, (2) how solvent properties (dielectric constant and reactivity) determine the colloidal stability of the nanocrystals at high temperatures needed for the production of high-quality products, and (3) how chemical reactions occur on terminal functional groups of the ligands to allow covalent conjugation to biomolecules. The reactions are expected to allow scalable manufacturing using minimal unit operations to generate colloidally stable products in polar solvents. The reactions are also expected to generate benign waste products due to processing and purification in aqueous solvents. With success of the project, low-cost and scalable manufacturing of diverse nanocrystals should be more accessible to non-experts and in educational settings. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Data protection regulations specify how personal data must be protected when used by others. Tracking and accounting for protections of data privacy has emerged as a pivotal requirement in contemporary data protection regulations. As a result, those who are responsible for using personal data, so-called data controllers, must actively enhance the privacy safeguards they provide. This project addresses the intricate challenges surrounding privacy accountability within the mobile software ecosystem, characterized by the opacity of third-party code modules, particularly third-party libraries. Existing methods for achieving privacy accountability primarily emphasize data transparency, often overlooking essential principles like data minimization and purpose limitation and facing integration challenges within mobile software development lifecycles. This research project seeks to address these limitations by presenting innovative approaches to enforce privacy accountability throughout the mobile software development process. The goal is to establish a more privacy-conscious and accountable mobile ecosystem, benefiting both users and data controllers. The outcomes of the research will contribute to educational curriculum and training to help developers achieve privacy goals plus additional outreach through workshop and bootcamp venues. The project's technical objectives are divided into three research thrusts: (1) understanding privacy accountability challenges in the mobile third-party code modules; (2) designing a privacy-accountable disclosure framework; (3) continuously enforcing privacy accountability properties in mobile software development lifecycle. The technical contribution of this research lies in advancing the socio-technical understanding of privacy non-compliance risks and accountability challenges within the mobile software supply chain. Additionally, it involves designing novel technical foundations that seamlessly integrate various methodologies and disciplines. This includes program analysis, formal methods, natural language processing, and human subject research, culminating in a privacy-accountable disclosure framework and continuous privacy accountability enforcement mechanism. These innovations are designed to be easily adoptable within the mobile software supply chain. The research will foster a holistic approach to enhancing privacy protection and accountability in mobile software development lifecycle and contribute to the creation of a safer and more privacy-conscious mobile ecosystem. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

ABSTRACT Electron and fluorescence microscopy now provide exquisite structural (~ nm) information using specific, labeled molecular species. However, only a small fraction of cellular species can be illuminated in any experiment and labeling techniques work primarily for large molecules and assemblies. Small molecules, such as metabolites, remain largely hidden without significant effort (e.g., single cell mass spectrometry). Molecular vibrational spectroscopies (infrared and Raman) offer a unique opportunity to study all molecules of life directly but are typically limited in the far field by optical resolution, suffer from low sensitivity when sampling volumes are reduced to nm scale or do not yet offer consistent, reproducible data for near-field imaging. Atomic Force Microscopy (AFM) combined with Infrared (IR) spectroscopy provides a powerful tool for label-free chemical imaging with nanoscale resolution. There were three main challenges that limited AFM-IR performance: (a) mechanical artifacts coupled the morphology and chemical domains, preventing chemical analyses, (b) high noise reduced analytical accuracy and precision and (c) low sensitivity precluded quantification of multiple species. Consequently, recorded data of biological materials were restricted to studies using point spectra to avoid image artifacts, used thick samples (~micron scale) for sufficient sensitivity or examined well-described and simple domains like amyloid deposits. Recent scientific advances have paved a path to overcome these limitations; however, legacy instruments cannot take advantage of these advances. Hence, in this proposal we will develop a new AFM-IR technology, dubbed resonance enhance null-deflection IR (RENDIR) spectroscopic imaging. RENDIR will include custom electronics for controlling the instrument based on our recent closed loop null deflection technique, optimized data recording using theoretical understanding of the image formation process, an optimized optical train with expanded range quantum cascade lasers, and novel optics integrated into a custom designed AFM. These advances will enable high-quality AFM-IR signals for all types of samples and by users who are not necessarily doctoral level experts. This innovative technology will be tested using gold standard samples and its limits characterized using fabricated polymeric samples of known composition and geometry as well as biomedical samples. Lithographically patterned polymer samples will first be used to validate both spectral and spatial performance. An isogenic cancer progression cell line will be used as a biological model, wherein all subcellular domains and local metabolites in cytoplasmic regions will be identified. The end of this project period will yield an AFM-IR measurement technology (RENDIR) that significantly surpasses state of the art by being free of mechanical artifacts, allowing ~5 nm (~64-fold smaller pixels) resolution, images with ~25x lower noise (due to null deflection), ~19x higher sensitivity (due to resonance enhancement), and ~50x higher data acquisition speed (real-time controls) as well as a quantitative concordance with gold-standard Fourier transform IR (FT-IR) spectra.

NSF Awards · FY 2025 · 2025-08

This project team will collect high-resolution, post-wildfire soil data using drone-mounted microwave sensors in response to the ongoing Monroe Canyon Wildfire in Utah's Fishlake National Forest. This wildfire intersects a region identified by the Fire and Smoke Model Evaluation Experiment (FASMEE) project as a critical data collection zone. As a result of this team's prior involvement in FASMEE's field campaigns and established relationships with local authorities and agencies, they are uniquely positioned to deploy and collect urgently needed post-fire soil hydrologic data. They will deploy their existing drone-mounted passive L-band radiometer system for two weeks following a precipitation event to measure fire-induced changes in soil moisture, infiltration dynamics, and hydrophobicity. These parameters are critical for understanding post-fire flood risks, erosion susceptibility, and ecosystem recovery. The resulting dataset will fill a crucial gap: no high-resolution, post-fire soil moisture datasets currently exist at this scale and resolution. Integration with NASA's UAVSAR pre-, during-, and post-fire L-band SAR observations will enable multi-scale, multi-modal analysis that is unprecedented in the wildfire recovery domain. The project will offer field based training for students and share resulting data in open repositories. The researchers will collect the first high-resolution, drone-based microwave remote sensing dataset focused on soil moisture dynamics and fire-induced hydrologic change immediately following a high-intensity wildfire. By capturing daily soil moisture dry-down behavior over both burned and adjacent unburned areas using passive L-band microwave radiometry, the project will generate spatially explicit time series suitable for estimating key soil hydraulic parameters such as field capacity, critical point, and wilting point. These parameters are essential for understanding post-fire infiltration, storage, evaporation, and flood risk, yet are currently unavailable at landscape scales due to the lack of timely, high-resolution measurements. This data collection effort fills a critical gap in post-fire ecohydrology, providing spatial continuity that is lacking in point-based studies and finer granularity than satellite products. Through alignment of their UAS data collection trajectories with the NASA UAVSAR airborne SAR missions, which have collected pre- and during-fire data in the area and are anticipated to conduct post-fire flights, this team's measurements will form part of a unique multiscale dataset that links surface, near-surface, and airborne observations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

The human nervous system enables effortless focus on a single conversation in crowded and noisy environments, through mental filtering-out of the surrounding noise. This phenomenon, known as the Cocktail Party Problem (CPP) highlights the brain’s remarkable ability to focus attention, integrate sensory input, and enhance a specific auditory signal while suppressing all others in a way that no current machine can replicate. Emerging evidence suggests that the noise-filtering ability relies on ‘brain rhythms’ shaped by the structure of the brain. This award will support research inspired by auditory perception in the brain. The project is organized around three core thrusts: (1) developing mathematical models to describe the neural computations underlying this capability, (2) engineering neural systems with interfaces to stimulate and record neural behavior, and (3) creating a framework to examine the societal implications, ethical considerations, and public understanding of this research, aiming to foster trust and responsible innovation. This project could lay the foundations for future design of neural computing systems. This research seeks to develop both the mathematical foundations and technological platforms for computing using neural oscillations. The theoretical framework builds on recent advances in Bayesian inference and mean-field game theory to model core bio-computational processes – such as encoding, learning and inference – within structured neural substrates. On the technological front, the project will engineer in vitro neural substrates that replicate cortical architectures guided by mathematical models derived from in vivo data. These 3D neural constructs will be integrated with electric, fluidic, and optical interfaces, enabling high-resolution access to neural dynamics allowing unprecedented insight into how oscillatory activity shapes learning and inference across network topologies. Taken together, the theoretical and technological components will provide a foundational understanding of design principles for next-generation 3D neural computing systems. Complementing this scientific work is an embedded ethical component that explores the broader societal implications of AI and biohybrid computing. This project will engage the public to explore the social, cultural, and ethical factors that influence the responsible development of these technologies. Ultimately the project aspires to deliver a comprehensive framework for ethical innovation – one that fosters public trust while supporting sustainable progress at the intersection of neuroscience, artificial intelligence, and society. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Short-video systems deliver videos only a few seconds or at most a few minutes long, in a personalized way, to billions of users around the globe. In the last few years, they have become increasingly popular for delivering societally beneficial content, such as microlearning, news, citizen reporting, advertising, user-generated content, testimonials, sports feeds, and more. Unlike long-form multimedia which is well-studied, short-video systems are unique in terms of the associated user behaviors, video and audio content, recommender algorithms, and the delivery pipelines that transport videos from content providers via content distribution networks (CDNs) and eventually to user devices. This transformative project will help improve both the efficiency of, and our understanding of, the class of short-video systems. The project plan draws on a unique and innovative combination of systems and networking design philosophies, along with machine learning (ML) techniques, complemented by real human user studies; this combination is essential for short-video systems because of their user-facing nature. The expected project outcomes include new video delivery techniques that use less compute and network resources and reduce consumption of user devices’ energy; new analytical understandings of the behavior of short video systems; and browser plugins and open software. Educational content will include course modules, including ones for online courses, focused on short-video streaming systems. The project will broaden participation in computing for Americans from any and all backgrounds, including high-school students, undergraduate researchers, and graduate students. Technically, this project will build transformative new ways of building learning-based and adaptive techniques for short-video systems. Our project, called "LANDS - Learning-based Adaptive Networked systems for Delivery of Short videos," will build three systems: (A) MidLand, a system that uses novel video reordering to reduce content distribution network (CDN) costs of midgress and cache size investment, while maintaining high user engagement and QoE (Quality of Engagement); (B) HighLand, a system that leverages ML pipelines, such as large vision language models, to predict user behavior and capture recommender algorithm performance as well as improve system-level metrics like cache effectiveness and adaptive bit rate adaptation; and (C) LowLand, a system that executes the ML pipelines of HighLand in fast, resource-efficient, and scalable ways, with support for expressing many types of useful analytic pipelines. Overall, LANDS will consist of both "learning-independent" layers and rich ML-driven layers atop them to further improve performance. The project plan contains a carefully crafted mix of system design and implementation, along with ML techniques (e.g., Large Language Models) as well as human user studies (with IRB approval). The team is interdisciplinary, with expertise across distributed systems, networking, ML systems, and human-computer interaction. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This research project explores the statistical aspects of diffusion models, an emerging class of generative modeling techniques that are transforming current practices in image and video synthesis, scientific simulation, inverse problems, and offline reinforcement learning. The project will apply information theory methods to explain when and why diffusion models succeed under certain statistical assumptions and when they do not. The results of the research are expected to advance the understanding of diffusion-based generative models and inform how they can be improved in terms of generation quality, computational efficiency, and user privacy preservation. The project provides research topics for training undergraduate and graduate students in modern statistical and machine learning techniques. Specifically, the project aims to address three technical questions: (1) What are the statistical limits of diffusion models in the minimax sense, especially the effect of low probability regions that may explain hallucination behaviors of generative models? (2) What is the optimal query complexity for sampling in diffusion models? The investigator will provide a systematic approach for the optimal query complexity by establishing connections with information-theoretic techniques previously used for analyzing the channel capacity. Accelerated diffusion methods will be constructed that nearly achieve this optimal complexity. (3) What is the fundamental trade-off between accuracy and differential privacy of diffusion models, and how can they be achieved? This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Sea ice is a key feature of the Southern Ocean that shapes the physical structure of the water column and regulates phytoplankton community dynamics and primary production. Phytoplankton are the base of the food chain, and the type of phytoplankton present, along with their overall productivity, impact the abundance of zooplankton and larger animals. Phytoplankton communities and production are also an important link for carbon export to the deep sea, a critical service provided by the Southern Ocean. However, sea ice extent and duration are decreasing in the Antarctic Peninsula region of the Southern Ocean, potentially affecting carbon export. This project aims to evaluate physical and chemical characteristics of sea ice in the Weddell Sea near Seymour Island and quantify effects of melting sea ice on phytoplankton and zooplankton growth and carbon export. This work will promote the progress of polar science and allow for better predictions of the ecosystem effects of changing sea ice conditions, such as shifts in krill abundance and its ability to support macrofauna and fisheries, and changes in carbon export. This project will further support the training of new undergraduate and graduate polar scientists and confer key transferable skills, such as data analysis and visualization and science communication. This project brings together an interdisciplinary team of physical, chemical, and biological oceanographers to comprehensively evaluate characteristics of sea ice in the Weddell Sea near Seymour Island and parse the effects of sea ice melt on biological systems and carbon export through a dual approach involving environmental observations and replicated factorial incubation experiments. Snow pits and ice cores will be used to characterize physical (snow density, hardness, temperature, and grain size and shape) and chemical parameters (δ¹⁸O, nitrate + nitrite, and Fe concentrations) of sea ice. Physical characteristics will be used to better calibrate satellite observations and improve future remote sensing. Chemical parameters will help determine how much meltwater is derived from sea ice and will further constrain the impact of sea ice melt on macro and micronutrient availability in the surface mixed layer. The spatial evolution of sea ice melt across the Weddell Sea continental shelf will be quantified through conductivity, temperature, and depth transects extending from Seymour Island to the shelf break. These transects will also include seawater sampling, zooplankton tows, and deployment of an in situ particle imager to characterize microbial, phytoplankton, and zooplankton communities and calculate particle flux. Experimental incubations will constrain the impacts of changing salinity, light, and nutrient regimes associated with ice melt on biological productivity, particle formation, and export potential. Broader Impacts include training an undergraduate and a graduate student and a synthesis workshop to share results and build a network of scientists interested in the impacts of ice-melt on polar systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

The future of the engineering and technology workforce will rely on robotics engineers to aid in unleashing ground breaking discoveries and translational solutions to the nation’s greatest challenges. Robots developed through a human-centered design approach have a focus on the end user and designs to best meet their needs. Soft robots describe those electro-mechanical devices made from low-modulus materials that can safely interact with people for a variety of applications. Soft robotics innovation sits at the nexus of advances in materials, automation, and biotechnology. Current strategies to attract K12 students to robotics-aligned engineering majors and careers are not equipped to meet the needs of the future workforce. This project will aim to deliver new curriculum for students to develop their science, technology, engineering, and math (STEM) talent by embedding a curricular strand of human-centered soft robotics throughout science courses in high school. This project will study the implementation of a design-based curriculum and the opportunities it creates for students to design solutions to challenges relevant to the United States and its people to increase engagement in engineering. By centering end users, we can reveal for students that robotics requires a multidisciplinary approach and can align with personal goals to advance the national health, prosperity and welfare. Given the intentional partnership with high schools in rural, suburban and urban schools, our findings can be translated across educational context to support professional formation of robotics engineers. This will ultimately lead to a stronger technical workforce and global science and engineering competitiveness. This project will build on pilot work to conduct the first mixed methods, longitudinal study of the impact of human-centered soft robotics on student career attitudes. This project will reveal aspects of a multi-year robotics curriculum that appeals to students’ engineering agency beliefs through validating a new measure, human-centered design self-efficacy for high school students. This project will ask, how can pre-college curricula be reimagined to change students’ perceptions of engineering majors through emotional connection to curricular applications? To answer this question, this CAREER project will conduct an integrated research and education project to disrupt traditional robotics education with a science-integrated, human-centered soft robotics curriculum and career resources. Robots developed for wearable technologies, for example, may align with engineering agency beliefs and aid in development of self-efficacy, identities and career interest. We argue that supporting engineering identity and career interest in robotics will require a multifaceted approach that equips teachers, schools, families, and students with resources, hence our planned educational activities. Understanding more deeply how human-centered design can create an enriching educational environment will fundamentally change approaches to engineering education to recruit roboticists and engineers to meet society’s most significant technical challenges. New knowledge generated in methods to measure human-centered design engineering self-efficacy can be used to improve existing robotics and engineering curricula. While studied in the context of soft robotics, findings from this project will be broadly applicable to other robotics and STEM fields. This integrated research and education project will produce a framework by which educators can teach engineering concepts in science to (1) change perceptions of robotics, and (2) empower students to develop robotics engineering career interest as early as high school. Directly through this project, thousands of students will benefit from a new multi-year science curriculum. Findings from this study will pave the way for increasing the number of students in the engineering workforce, so that new innovative technologies that benefit all of society can be realized. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

Summary In assisted reproductive technology (ART), morphological evaluation by bright- or dark-field optical microscopy is the most critical assessment on the implantation potential (viability) of an in vitro fertilized embryo. Ideally the obtained morphological information should be complemented with spatially resolved metabolic or functional information. This would provide a more complete picture of the embryo itself and may thus offer a more objective viability assay to increase the successful rate of an embryo transfer toward (singleton) live-birth. Currently, fertility clinic providers are unable to visualize embryonic metabolism, whereas the spectroscopy- based metabolic profiling under development has lacked spatial resolution. Therefore, a critical need exists to superimpose spatially resolved metabolic information on morphological information for full assessment of the embryo before implantation. An interdisciplinary partnership composed of collaborating investigators in optical instrumentation, reproductive biology, and computer science will address this critical need by multiplexed optical metabolic imaging (MOMI). In contrast to the intrinsically label-free bright- or dark-field optical microscopy unable to obtain molecular information and resolve a three–dimensional structure, MOMI enables these abilities via well-known endogenous metabolites and nonlinear optical excitation, respectively, while critically retains the label-free aspect required for potential clinical application. The preliminary results have demonstrated the utility of MOMI as a general imaging assay for unlabeled live cells and cultured spheroids or organoids. This partnership will thus aim to validate the MOMI in an established bovine model of ART. Future plans are to extend these methodologies to improve human ART. The goal of this partnership will be achieved through a systematic approach. First, the inclusion of fluorescence lifetime imaging will upgrade MOMI to an omics-like assay, and a dedicated environment of incubation will be built to interface the imaging with preimplantation embryo culturing. Also, an established bovine model for human ART will be employed to image 2500 preimplantation embryos and track subsequent live births of ~1000 embryo transfers. Finally, a novel machine learning algorithm based on multi-cell graph neural network from MOMI images will be developed to predict the viability of a preimplantation embryo (prospectively). The successful completion of this project will validate MOMI as a safe live-cell imaging assay for non-ART biomedical applications such as cell therapy and drug screening, considering the embryo development as a well-known sensitive bioassay for plausible optical imaging-induced phototoxicity. More importantly, this large- scale preclinical study will justify a randomized clinical trial for human ART. Ultimately, the clinical application of MOMI may significantly improve the success rate of live births versus embryo transfers that has remained stagnant in the last decade (35-37%) and may therefore reduce the associated pain and cost.

NSF Awards · FY 2025 · 2025-08

Gravitational wave astrophysics has entered a new era of discovery and exploration. In 2015, advanced LIGO detected gravitational waves for the first time, earning the pioneers of gravitational-wave physics a Nobel prize, and the collaboration several other prestigious prizes. Since then, advanced LIGO and its European partner, Virgo, and its Japanese partner, KAGRA, have detected over 300 additional events. All of these gravitational waves were emitted by black holes and/or neutron stars that spiraled into each other and collided at velocities close to half the speed of light. The neutron star events revealed important astrophysical information about matter in extreme environments, such as the fact that most of the gold in the universe is produced in these cosmic collisions. The black hole events validated the predictions of Einstein's theory of general relativity in a regime that had never been explored before: where the gravitational force is enormous and violently changing. The extraction and the interpretation of all this physical information required the accurate construction of gravitational wave models and data analysis techniques with which to pull the signal out of the detector noise. The main objective of this award is to develop ready-to-use models and artificial-intelligence-enhanced data analysis algorithms to trigger strong progress in science through the more detailed and robust extraction of information about matter and gravity in extreme environments. The award integrates this scientific research with educational activities, aimed at educating high-school students and the general public about physics and science, through activities at the Illinois Center for Advanced Studies of the University at the University of Illinois Urbana-Champaign, a fertile training ground for the next generation of gravitational scientists. This award is aimed at developing and implementing ready-to-use models and data analysis algorithms in two focus areas. Focus Area I is centered on the extraction of new, nuclear physics information from the gravitational waves emitted in the late quasi-circular inspiral of neutron star binaries. In particular, the models and algorithms this award develops enable inferences (robust to systematic uncertainties) on (i) non-trivial features in the speed of sound of the neutron-star fluid, (ii) (the principal-directions of the) nuclear physics parameter space of realistic equation-of-state families, and (iii) out-of-equilibrium processes that induce an effective viscosity in the neutron-star fluid. Focus Area II concentrates on carrying out new, model-agnostic tests of general relativity using current and future gravitational-wave data. In particular, this award (i) develops and implements a neural-network-enhanced, parameterized-post-Einsteinian framework to search for and to detect or constrain (post-Newtonian and non-post-Newtonian) deviations from general relativity in gravitational-wave data produced by inspiraling black hole binaries, and (ii) carries out model-specific tests of general relativity with ringdown-only gravitational-wave data. This award also implements broader impact activities aimed at educating and increasing the interest of high-school students and the general public in gravitational physics. In The Art of Physics/The Physics of Art, undergraduate and graduate students are educated in cutting-edge physics to create art that conveys gravitational physics concepts. The textbook "A Theorist’s Guide to the Physics of Neutron Stars" introduces junior graduate students to the gravitational and nuclear physics relevant to the study of neutron stars, which is then used in graduate classes and group meetings to train the next generation of multi-messenger physicists. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

As our society and technology continue to evolve rapidly, there is a growing interests by adults in taking an active role to continue to learn on their own as hey age. One of the most common lifelong learning scenarios is gaining knowledege from an algorithm-mediated information environment, such as using the internet to explore unfamiliar topics or interact with conversational agents to adopt new health behaviors. Contemporary algorithm-mediated information retrieval (IR) systems, including search engines and conversational agents, have limitations when it comes to adequately supporting users with complex information needs, particularly those related to learning-oriented search tasks. Finding information does not necessarily lead to effective learning or deep comprehension. Moreover, IR algorithms often neglect users needs or intentions for the search - one search does not fit all. Measures of intent and algorithms need to be tuned to individual users or risk misaligning with the learning goals of users. Therefore, developing IR systems for learning requires understanding how individuals monitor, assess, and regulate their learning progress and what factors shape their judgments to persist in or disengage from learning. To facilitate the translation of research into instructional and outreach practices, the project will collaborate with the Osher Lifelong Learning Institution and the National Multiple Sclerosis Society to co-design educational games and webinars that foster information literacy and its practical applications in everyday search and learning contexts. The project seeks to advance the understanding of self-regulated learning as mediated by information search behavior and IR systems, while also pioneering the development of an intelligent system that scaffolds self-regulated learning through adaptive information search. The project is grounded in an interdisciplinary theoretical framework that integrates cognitive science theories, specifically information foraging, metacognition, self-regulated learning, and comprehension, with the capacities of large language models. First, the project examines the cognitive mechanisms underlying learning through information search, aiming to develop generalizable user models that characterize the cues individuals use to assess learning and how users regulate their resources between exploitation and exploration to optimize learning outcomes. Second, the project expands user models by considering contextual factors from diverse IR systems, including search engines and conversational agents, across the lifespan. Third, the project employs user models, symbolic knowledge representation, and large language models to develop intelligent systems that support self-regulated learning through information search. Overall, the overarching goal of the project is to empower adults across the lifespan to adapt to evolving information environments for lifelong learning. The research will guide the development of personalized technologies that promote lifelong learning in various educational settings, including online adult vocational education and individualized tutoring systems for all learners. The research outcomes and approach can be translated to other domains, enabling proactive prediction of user behavior and the delivery of tailored information experiences to enhance cognitive performance or health behaviors for adults across the lifespan. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

Abstract: In this proposed research project, we seek to develop an advanced brain SPECT system that offers a unique hyperspectral imaging capability substantiated by an excellent energy resolution (e.g., <2.5 keV at 140 keV and <3.5 keV at 250 keV) across a wide energy range (25-600 keV), and at the same time deliver a 1- mm spatial resolution and a very high sensitivity to allow detailed visualization of multi-tracer uptakes in various brain regions. This device could potentially have a transformative impact on brain research by allowing for microscopic, multi-functional assessment of brain functions under various experimental conditions. This proposed research project will integrate the disruptive high-performance 3D CZT imaging-spectrometer technologies with a novel synthetic compound eye (SCE) camera design, as well as an innovative iterative image reconstruction method using deep-learning based priors, to develop a next-generation clinical brain SPECT imaging system with transformative spatial resolution and imaging sensitivity unattainable previously. The long-term objective is to apply this innovative imaging system to human brain SPECT studies using a collection of various SPECT radiotracers, and develop and advance physiological parametric imaging methodologies, in order to investigate the long-standing issues in neurobiology and improve our understanding of the interplay and relationship among cerebral blood flow and perfusion, brain tissue oxygenation, neuronal cell metabolism, and brain cell tracking under different cognitive challenges and biophysical conditions in healthy and in disease. We would envision the proposed system to serve as a unique imaging platform to significantly advance our understanding of neural cell biology and regional brain functions in response to various cognitive, behavioral, and physiological challenges by employing these unprecedentedly innovative SPECT imaging methodologies in order to assess all relevant quantitative physiological measurements that will be interpreted in an integrated fashion and synergistically so that new perspectives in brain research unattainable previously can be formulated.

NSF Awards · FY 2025 · 2025-08

This award funds planning activities for a new Industry University Cooperative Research Center (IUCRC), the Center for Manufacturing Ultrasound Systems for Intelligent Care (cMUSIC). The cMUSIC center is led by the North Carolina State University (NCSU) in partnership with the University of Illinois Urbana-Champaign (UIUC). The global medical ultrasound devices market is rapidly growing due to accelerating rates of chronic and lifestyle-related illnesses. However, the U.S. ultrasound industry faces unprecedented challenges to exploiting growth opportunities, largely related to insufficient device development, system design, manufacturing, and commercial translation capabilities following decades of manufacturing outsourcing. The center will integrate electronics and intelligence into ultrasound systems to transform the manufacturing sector of ultrasound products for sensing, imaging, therapy and drug delivery. The center aims to enhance the global strength and competitiveness of the U.S. manufacturing, electronics, and medical sectors of the economy. The cMUSIC center is also dedicated to workforce development through specialized training programs that will equip future workers with the skillset essential for 21st-century manufacturing, electronics, biomedical ultrasound and healthcare industry. Through a collaborative approach involving universities, industry leaders, and government agencies, the mission of cMUSIC center closely aligns with national priorities to grow economy sustainability and advance technological innovation. The mission of cMUSIC is to provide international leadership and train next-generation U.S. scientists and engineers in the fundamental ultrasound technology and its translation to practice. The cMUSIC center’s research focuses on the following four research thrust areas: advanced manufacturing of ultrasound systems with a large number of array elements by using semiconductor manufacturing processes; artificial Intelligence (AI) integrated ultrasound imaging for disease diagnosis, surgery guidance and post-treatment assessment; implantable and wearable ultrasonic sensing, communication, imaging, therapy, drug delivery and ultrasound power transfer; advanced thermal and non-thermal ultrasound therapy including histotripsy and neural modulation; and advanced characterization of ultrasound devices for intelligent ultrasound devices and systems. The UIUC site will specifically focus on advanced ultrasound electronics, integrated systems, AI in diagnostics, and wireless power transfer for implantable devices. UIUC has a history of designing larger scale ultrasound systems for volumetric imaging and designing and building smaller scale systems for wearable and implanted devices. UIUC is a world recognized leader in electronics and integrated circuits. With world class integrated electronics, systems, and power research labs in UIUC and cross-collaborations between NCSU/UNC and UIUC, cMUSIC center will significantly contribute to the next generation of biomedical ultrasound product manufacturing for intelligent care. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Non-Technical description Engineering living materials is a promising strategy to develop smart, resilient materials that can grow, adapt, and repair themselves, expanding possibilities in next-generation materials science and engineer-ing. A striking example of nature’s ingenuity in creating living materials is the way it regulates bio-mineralization in coral reefs. These dynamic structures not only provide crucial support for marine life, coastal protection, and local economies, but also serve as a source of bioactive compounds with poten-tial as drugs for treating cardiovascular diseases and cancer. If a coral fragment is detached from the main colony and attached to a substrate, the wound heals, allowing the fragment to grow into a new cor-al. While this process, known as coral fragmentation (or fragging), presents a promising strategy for cor-al conservation and restoration, fragging also offers a pathway for developing coral-inspired living mate-rials. However, the fundamental understanding of the processes involved remains quite limited from a materials science perspective. The goal of this project is thus to advance from infancy to maturity the understanding and control of coral fragment biomineralization. The project outcomes advance the sci-ence of biomineralization and how it can be used to design materials that live, grow, and heal— inspired by corals. Broader impacts of this project include training two graduate students at the intersection of materials science and biomineralization and hosting undergraduate students in the principal investiga-tors´ lab to further develop their interests in materials science, thereby enhancing the STEM pipeline. Discoveries in coral-inspired living materials have the potential to drive economic prosperity through product innovation and new markets, while simultaneously improving societal welfare by enhancing health and safety. Technical description The goal of this NSF project is thus to advance the knowledge of the biomineralization processes in cor-al fragment healing and growth and how these processes relate to the resulting material properties. Moreover, the project provides insights into how to influence early-stage coral biomineralization and material properties by tuning the substrate composition and/or water chemistry, i.e., the local chemical environment around the fragment. To achieve this goal, the project tasks are to (i) study the self-attachment of (micro)fragments and the mineralization at the body basal wall, (ii) characterize fragment growth and the evolution of skeletal morphology, (iii) elucidate the evolution of fragment microstructure during early growth (2–3 years) after attachment and the biomineralization pathway, and (iv) understand the effects of extrinsic parameters (substrate composition, water chemistry) and fragment size on growth, properties of (micro)fragments, and the biomineralization pathway. Focusing on fragments of two spe-cies of reef-building scleractinian corals (Echinopora lamellosa and Favia fragum), the experimental plan includes an in-depth study of early-stage calcification during self-attachment using advanced liquid-cell transmission electron microscopy. Additional knowledge is obtained from a comprehensive dataset encompassing coral morphology, growth rates, and skeletal microstructure and composition using in-house techniques such as scanning electron microscopy, energy-dispersive X-ray spectroscopy, and nano-computed tomography with submicron resolution. These studies are complemented by experiments with collaborators at Argonne National Laboratory to determine the skeleton composition. Overall, the insights gained from this research contribute to establishing design principles for coral-inspired living materials and advance the knowledge about biomineralization processes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Computing and energy have always been interwoven. Large-scale energy systems were some of the earliest adopters of advanced computing systems. The earliest large-scale computers were intensive energy consumers. However, computing and energy are now colliding in ways unseen. Historically, networking, computing hardware, and software advancements limited the growth of data centers, AI, and high-performance supercomputing. Today, increasingly the energy and electricity grid infrastructure are limiting their growth. Leadership in AI, in particular, is necessary for national strategic imperatives. Energy challenges must be addressed for continued growth of AI. The Compute/Energy Nexus Workshop is proposed by a team at the University of Illinois Urbana-Champaign and collaborators as an invitation-based two-day event to assemble a wide range of stakeholders, articulate key challenges, and start an action plan. The workshop aims to organize a group to prepare a national strategic roadmap for integrated solutions to the challenges. The workshop will propose immediate actions and near-term tactical programs to address key issues. The roadmap group is to be organized among cross-cutting stakeholder groups, charging them to initiate activities. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This project aims to advance simulations of gas-liquid mixtures, especially when they break apart into small droplets or bubbles and when small droplets or bubbles come together. These fluid behaviors are commonplace in nature, for example in the breaking of waves at the ocean surface or formation of raindrops in clouds. They are also crucial for many engineering applications, such as the injection of fuel in combustion engines, the spraying of crop protection products, or the production of powders in the food and pharmaceutical industries. The most common simulation methods struggle to capture small but important details, such as very thin liquid sheets or tiny droplets, which limits accuracy and utility of the results. This project will develop new ways to represent and model these fine details of fluid behavior, resulting in more accurate simulations without requiring expensive computer resources. The approach will allow scientists and engineers to better predict how gas-liquid mixtures behave in complex situations, making engineering design more affordable and more accurate. The proposed research will also contribute to modernizing course content for training undergraduate and graduate students, while fostering collaboration with industry to promote the widespread adoption of open-source software tools. Multi-scale two-phase flows play a central role in many natural phenomena but also in several key industrial sectors, such as energy production, transportation, manufacturing, and the food and pharmaceutical industries. Traditional Eulerian interface capturing methods fail to accurately predict topology changes of the gas-liquid interface due to mesh resolution limits and numerical errors. To address these limitations, this project proposes two key innovations: (1) a new piecewise-quadratic interface representation that enables the capture of sub-grid scale structures such as thin films/sheets and ligaments, and (2) a new volume-filtered framework in which sub-grid scale surface tension-driven physics are accounted for through closure models. The outcome will be a framework capable of accurately predicting break-up and coalescence events as well as droplet size distributions in multi-scale two-phase flows, a feat that has so far remained elusive to even the most refined simulation frameworks. This project marks a shift from expensive and often insufficient direct numerical simulation to efficient, physics-informed modeling, through the introduction of novel sub-grid scale interface representation. This promises both more affordable and more predictive simulations of multi-scale two-phase flows. Industrial impact will be maximized by the open-source release of the developed numerical tools, their integration in commercial codes, and the organization of user workshops. The project will also modernize course content on multiphase flows to benefit engineering education. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.