University of Minnesota-Twin Cities

NSF Awards · FY 2026 · 2026-10

Since the first direct observation of gravitational waves in 2015, a new field of astronomy has fundamentally changed how the universe can be explored. A gravitational wave is a ripple in spacetime produced when two extremely dense objects, such as black holes or neutron stars, collide. The detectors that observe these signals now record hundreds of such events per year, and the rate is expected to grow to several events per day within the next few years. Sifting through this flood of data to find the rare astrophysical signals buried in detector noise, and then alerting telescopes around the world quickly enough to capture the light produced by a collision, is one of the most demanding computational problems in modern science. This project develops an open, community machine learning framework that makes gravitational-wave discovery faster, more accurate, and more accessible. The framework will reduce the time between the crossing of a gravitational wave through the detectors and a public astronomical alert to less than a second, make it possible to detect events that traditional analysis methods miss, and lower the computing cost of these analyses by orders of magnitude. The project trains the next generation of scientists, including high-school students, undergraduates, graduate students, and researchers at smaller institutions, by sharing open code, open data, open trained models, and open lessons. It also strengthens shared national computing infrastructure that benefits not just gravitational-wave science but also particle physics, neutrino astronomy, and time-domain astronomy more broadly, advancing the national interest by accelerating discovery and broadening participation in science. The project develops ml4gw, an open-source PyTorch-based machine learning (ML) framework for gravitational wave (GW) data analysis. ml4gw provides Graphics Processing Unit (GPU)-accelerated implementations of the data ingestion, signal-processing, waveform-generation, and inference operations that historically ran on central-processing-unit clusters of the Laser Interferometer Gravitational-wave Observatory (LIGO) Data Grid, and integrates the resulting models into the international low-latency alert pipeline. The award covers three coordinated work packages. The first extends model coverage to long-duration binary-neutron-star and neutron-star-black-hole signals using multi-rate and multi-band processing, integrates auxiliary detector channels through multimodal architectures, and develops state-space and ensemble models together with a shared foundation backbone from which task-specific models can be fine-tuned. The second work package builds production-grade cyberinfrastructure that targets the National Artificial Intelligence Research Resource, the National Research Platform, and the Open Science Data Federation, including an Inference-as-a-Service deployment built on the NVIDIA Triton inference server and on the SuperSONIC service that already supports particle and neutrino physics experiments. The third work package delivers community resources: standardized benchmark datasets with persistent digital object identifiers on Zenodo, versioned reference models on Hugging Face, comprehensive documentation and tutorials hosted on Read the Docs, containerized release artifacts, an upgraded continuous integration system that reduces test runtime by an order of magnitude, and an agent-driven development scaffold for community-led code contribution. Training and outreach activities include hands-on tutorials at international collaboration meetings, an annual hands-on lesson at the University of Minnesota Time-Domain Astrophysics Summer School, and a public machine learning challenge focused on binary-neutron-star detection that builds on a prior challenge that engaged hundreds of teams and roughly a thousand individual participants. This award by the Office of Advanced Cyberinfrastructure is jointly supported by the Physics at the Information Frontier Program within the Physics Section of the Directorate for Mathematical and Physical Sciences. 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 2026 · 2026-07

Discovering new materials, safer chemicals, and better manufacturing processes often depends on running costly experiments in real-world laboratories. In many areas of science and engineering, researchers must choose from a large number of possibilities, where testing each experimental design is expensive in terms of resources consumed. This project seeks to transform the practice of selecting experiments to accelerate engineering design and scientific discoveries. The project will develop novel artificial intelligence (AI) methods to help engineers and scientists to adaptively decide which experiments to run next so that promising discoveries can be found with far fewer trials than traditional trial-and-error methods. The project will also strengthen the future AI workforce by training undergraduate and graduate students through new courses and research opportunities, create open-source tools, and benchmark problems to advance AI and scientific discovery. The overarching goal of this project is to develop a general framework for AI-driven adaptive experimental design over large structured design spaces (e.g., sequences and graphs), where each experiment is expensive and only a small fraction of candidate designs can be tested. The research has four inter-connected thrusts. First, developing new probabilistic surrogate models that map high-dimensional discrete designs to experimental outcomes, so that reliable predictions can be made even from small experimental datasets. Second, designing uncertainty-aware deep learning surrogate models that can use large historical datasets while still producing reliable prediction intervals. Third, developing a unified information-theoretic framework for selecting experiments toward a broad set of scientific goals, including optimizing multiple properties, identifying feasible regions, and finding diverse sets of high-quality candidates. Fourth, developing look-ahead planning methods for physical laboratories that account for setup time, limited equipment, and the need to prepare materials before experiments can be run. Together, these advances will provide broadly useful AI tools for selecting valuable experiments more efficiently, and they will be evaluated in three real-world application domains in materials discovery, chemical design, and additive manufacturing. 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 2026 · 2026-07

The partial differential equations (PDE) of fluids describe the dynamics of vortices, which appear across a wide range of fluid flows and play an important role in turbulence. Trailing vortices behind aircraft and vortex rings provide familiar examples. One goal of the project is to better understand the behavior of vortices from a mathematical perspective. Another is to study the “typical behavior” of solutions of the basic equations. For example, why are certain scenarios, such as singularity formation, not observed, while we are unable to rule them out mathematically? Are we missing important information about the equations that could be helpful in designing models and in computer simulations of fluid flows? These questions will also be studied in simpler settings that can serve as stepping stones to the full equations. In a somewhat different direction, problems related to PDE of elasticity will also be explored. Together with traditional PDE methods, the project will explore possibilities offered by recent developments in AI for the study of turbulence. The project provides research training opportunities for graduate students. The project focuses on several related directions: vortex dynamics and models for vortex filaments, the formation and evolution of singularities, mechanisms by which singularities may be avoided, and geometric and variational approaches to incompressible flows. Key themes include the emergence of vortex dynamics from the fundamental equations, instabilities and singularity avoidance, the role of random perturbations, and geometric approaches to open problems for the incompressible Euler equations. The project also examines properties of classical energy functionals in variational integrals, including Morrey’s quasiconvexity and its relation to other convexity conditions. Overall, the research aims to identify essential parameters in infinite-dimensional systems. Such an identification can be very helpful for computer simulations of the equations. 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 2026 · 2026-06

Artificial intelligence (AI) is becoming a powerful tool for helping people make decisions, but most AI systems still deliver advice on desktop screens. This is a poor fit for people working in physical environments, such as surgeons, facility teams, first responders, and coaches. Augmented reality (AR) can place AI guidance directly into these settings, but simply moving information off a screen is not enough. In these settings, people must divide attention among the environment, movement, and the task itself, which can lead them to accept or reject AI advice too readily. This award develops adaptive AR interfaces that help people engage with AI more deliberately in real time while avoiding unnecessary disruption to the physical task. By improving how people and AI work together in real-world context, the project can support safer, more accurate, and more accountable decision-making in settings where errors are costly. The project will also train students in spatial computing, create open tools and learning materials, and broaden participation through courses, tutorials, and workshops. This award studies how adaptive situated visualizations in AR shape appropriate reliance on AI during decision-making tasks. The research will first conduct controlled experiments to identify which visual design choices, such as placement, visual prominence, lighting consistency, proximity, and scale, encourage overly fast judgments or more deliberate reasoning. It will then use these findings to design and evaluate adaptive visual prompts that change in response to user behavior, task demands, and AI confidence, for example by highlighting overlooked alternatives or revealing uncertainty at important moments. Finally, the project will develop learning-based methods that decide when and how to present these prompts using real-time signals from the user, the environment, and the AI system. The methods will be tested in sports analytics and facility management scenarios, producing general design principles and open-source tools for AR systems that help people and AI make better decisions together in physical spaces. 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 2026 · 2026-06

This project will advance understanding of how beneficial soil fungi and bacteria interact to improve plant nutrition and support sustainable agriculture. Arbuscular mycorrhizal fungi form symbiotic partnerships with most land plants, helping roots acquire nutrients, while bacteria associated with fungal hyphae can influence fungal growth, colonization, and nutrient exchange. However, the mechanisms governing these cross-kingdom interactions remain poorly understood. This work will address a fundamental question in soil ecology and plant biology: how mycorrhizal fungi recruit helpful bacteria and how those bacteria, in turn, enhance fungal symbiosis with plant roots. By uncovering the rules that shape these partnerships, the project will contribute to basic knowledge of microbial ecology and may inform future strategies and biotechnology priorities to reduce fertilizer dependence, improve crop productivity, and support environmentally sustainable agriculture. The project will also provide interdisciplinary training for students in microbiology, plant science, and molecular biology, strengthening workforce development in agricultural research. The project will use maize, the arbuscular mycorrhizal fungus Rhizophagus irregularis, and the bacterium Azotobacter vinelandii as a model system to test how fungal hyphal exudates influence bacterial persistence and how bacterial partners affect mycorrhization. Using compartmented growth chambers that separate plant roots from fungal hyphae, the project will analyze fungal exudates by metabolomics, measure gene expression changes in plant roots, fungi, and bacteria by transcriptomics, and quantify fungal colonization and bacterial population dynamics across varying nitrogen and phosphorus conditions. The project will also test whether bacterial strains differ in their ability to promote fungal growth and whether these effects depend on bacterial nitrogen metabolism or other helper functions. Together, these experiments will identify metabolites, genes, and ecological conditions that regulate mutualistic interactions between fungi and bacteria. The results are expected to reveal how microbial partnerships are established and maintained in soil and how they can be harnessed to improve nutrient acquisition and plant health. 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 2026 · 2026-06

Two-dimensional (2D) materials with the thickness of a single atom exhibit unconventional electronic, optical, and magnetic properties. This project builds novel material platforms by stacking multiple layers of 2D materials on top of each other, whose material properties can be designed with versatile choices of material combination and orientation between each layer. Material design and assembly is guided by multiscale modeling, followed by experimental characterization of novel material properties, and material optimization via a machine-learning feedback loop. Material platforms that are identified with novel material properties are fabricated into advanced nano-scale devices, to further understand the microscopic mechanism of unconventional quantum material properties and to demonstrate new device functionality. By opening new frontiers in materials science with AI-facilitated designer material properties, this project is strengthening the nation’s leadership in materials science and its application. This project also provides K-12, undergraduate, and graduate students with interdisciplinary training in computational and experimental techniques for materials science, physics, quantum science, chemistry, computer science and machine learning to develop our next generation quantum workforce, improve public awareness of next generation materials, and stimulate public engagement with quantum science and technology. This project is drastically expanding the scope of 2D material platforms by developing and investigating novel twisted-multi-layer systems consisting of graphene, 2D semiconductors, 2D superconductors, and 2D ferromagnets, whose material properties are designed and tuned by versatile material choices, twist-angle combinations, chemical intercalation, and modification. Fast turn-around cycles of theorical prediction, experimental characterization, and model optimization are being developed with a machine-learning approach to accelerate materials discovery. Advanced quantum electronic, optical, and magnetic devices are being fabricated to further study the microscopic mechanism of unconventional quantum phenomena, and to benchmark its potential application in next-generation quantum electronics and computing platforms with unprecedented device functionality. The project provides a fundamental understanding of novel properties and emergent phenomena in quantum materials and demonstrates device applications in low energy electronics, quantum sensing, and quantum computing. 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 2026 · 2026-06

This project aims to develop a new energy-efficient memory device to address the rapidly growing energy demands of artificial intelligence (AI) and data centers. As AI models become larger and more powerful, the electricity required to train and operate these computational models is growing exponentially and is projected to consume up to 12% of total electricity in the U.S. by 2028. Within this current data-centric computing scheme, non-volatile memory devices that store and retrieve information are a major contributor to energy dissipation. To address this challenge, the project will demonstrate a new cryogenic memory technology - superconducting magnetic random-access memory (SC-MRAM), which can dramatically reduce energy consumption by several orders of magnitude, even when accounting for the refrigeration cost. The proposed SC-MRAM device leverages the zero-resistance, dissipation-less nature of superconducting currents to perform memory read and write operations. The research integrates materials development, investigation of underlying physical mechanisms, and device engineering to establish a scalable pathway toward high-performance cryogenic memory. Such devices could lead to a paradigm shift toward energy-efficient cryogenic data centers with superior computational performance. On the education front, the project also entails a Personalized Education with AI for Quantum Engineering (PEAQ) program to modernize cross-disciplinary workforce training at the intersection of electrical engineering and quantum technologies. The PEAQ effort includes AI-assisted personalized curriculum design, developing intelligent textbooks with interactive learning experiences, and community outreach to increase quantum literacy among the public. Technically, the project seeks to understand and harness how spin-polarized supercurrents can control magnetization in superconductor/ferromagnet heterostructures. The research is organized into three coordinated thrusts. First, it will investigate spin transport and magnetization dynamics in various superconductor/ferromagnet device geometries through carefully designed magneto-transport measurements. Second, it will optimize material stacks and interfaces, with particular emphasis on strongly spin–orbit-coupled superconducting nitrides such as tantalum nitride (TaN), to simultaneously maximize both the spin Hall angle and the superconducting critical current density required for supercurrent-driven switching. Third, peripheral superconducting devices and circuitry, including superconducting diodes and superconducting transistors, will be developed to achieve all-superconducting read and write operations. By combining material optimization, transport characterization, device modeling, and circuit-level validation, the project aims to establish the scientific foundation and technological feasibility of all-superconducting memory, advancing both fundamental knowledge in superconducting spintronics and practical pathways toward cryogenic, energy-efficient computing architectures. 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 2026 · 2026-05

Software quality is critical to the success of software systems, and effective testing practices are key to achieving high quality. Yet, many organizations continue to struggle to release high-quality software because they often rely heavily on large end-to-end tests that validate complete user workflows, making it difficult and costly to verify frequent code changes effectively. The project’s novelties are approaches that generate small, focused tests from larger existing tests, recommend the most relevant of these generated tests in response to code changes, and keep them aligned as software evolves. The project’s broader significance and importance are that it helps organizations use existing testing resources more effectively, supports the development of higher-quality software, and advances education in software testing. The project ultimately prepares students and practitioners to build and maintain stronger test suites, with the long-term benefit of more dependable software. This project is centered on three research goals. First, it develops automated techniques for generating valid, actionable, and understandable carved tests, which are focused tests extracted from larger tests. Second, it develops automated and AI-based techniques for suggesting carved tests that are relevant to specific code changes, enabling focused and effective regression testing. Third, it develops automated techniques for maintaining alignment between carved tests and the evolving original tests so that extracted tests remain useful over time. These goals are pursued through a combination of dynamic analysis, static analysis, code-change analysis, and traceability analysis, and the resulting techniques are integrated into an open-source framework. The project also has an educational goal to incorporate test repurposing concepts into software engineering curricula through instructional modules and projects for students, while also supporting the continuing education of software professionals. The project advances the foundations and practice of software testing and helps prepare a workforce skilled in producing high-quality software. 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 2026 · 2026-04

Semiconductor devices are key to all modern technology, and continued innovations are needed to achieve high-performance computing, robust high-speed communications, machine learning, image and video processing, as well as energy-efficient and sustainable power generation and control. For continued U.S. leadership in these important fields, it is critical that the next generation of STEM researchers be trained to address emerging needs for electronic, optoelectronic and quantum devices. Federal investments in the semiconductor manufacturing industry make the training of device engineers even more imperative so we can provide the skilled workforce that will support expansion of domestic chip production. The annual Device Research Conference, which is currently in its 84th year, is the premier forum for innovative and emerging semiconductor devices. The 2026 Device Research Conference (DRC) will be held at the University of Michigan in Ann Arbor. One of the key features of DRC is the balanced participation of students and leading world experts in the field, which provides a unique learning and training opportunity for the student participants. In addition to contributed oral talks and posters, the 2026 DRC will offer a half-day technical short course taught by prominent experts, three plenary sessions and over 40 invited talks by academic and industry leaders, and an evening rump session at which panelists interact with attendees to debate a technical topic of particular current importance. This year, the main session topic will be “Everything Switches: Sorting Fact from Fiction in Future Devices.” The short course will focus on “Best Practices for Reporting on Electronic Devices.” ECCS support for the 2026 DRC will provide student participants with the opportunity to gain exposure to new materials and devices, their basic physics, as well as engineering applications. 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 2026 · 2026-04

Social media is often blamed for making mental well-being worse for adults. We have a general understanding of how platform design and use patterns can cause these problems. However, there is still much to learn about the mechanisms by which social media use affects well-being and what people and platforms can do to make things better. This project addresses these challenges by developing better models of how people use and respond to social media content, along with new AI-driven technologies that improve platform design and the algorithms that choose what, when, and how to show content to users. These new models and technologies will be guided by a wide range of studies of social media use that will enable early detection of possible harm to well-being. These detection models will be integrated into tools that help people think about the benefits and risks of using social media, as well as design. Finally, this project will create new designs for social media that reduce harms to well-being. To increase real-world impact, the research team will also engage in a multi-part public engagement effort to share the scientific results of the work with industry, policymakers, and the general public. This project develops new scientific knowledge, methods, and design interventions to understand and create more beneficial social media "engagement infrastructure" -- the interface and algorithmic features of social media platforms that help shape people's behavior and well-being. The research team will first conduct several studies to empirically characterize and precisely measure the engagement infrastructure that induces harmful use, developing a validated typology of the most problematic elements and mechanisms of action. These insights will be used to create and evaluate multimodal AI models for inferring when harmful use is occurring. Finally, the research team will design new interventions that interrupt and displace harmful use, either through platform-level changes or, for individual use, through overlay tools that change the engagement infrastructure. This approach will enable the decomposition of the precise mechanisms that mediate negative well-being outcomes, advancing human-computer interaction, social computing, and applied AI research. The project's outreach and education plan includes bi-directional interdisciplinary workshops at a number of venues to bridge research gaps, academic-industry collaborations for networking and knowledge transfer, public engagement through Minnesota State Fair exhibits and science communication partnerships, and curriculum development to train software engineers on responsible and ethical design. 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 2026 · 2026-04

This award supports the participation of graduate students and postdoctoral researchers in the "2026 Riviere-Fabes Symposium on Analysis and PDE" which is scheduled to take place May 1-3, 2026 at the University of Minnesota. The award gives early-career researchers and researchers without other sources of funding a chance to participate in the conference. In this way, the award supports the communication of state-of-the-art mathematical techniques and promotes the development of future generations of scientists working in important, cross-disciplinary fields. The symposium focuses on recent developments in mathematical analysis, this year especially in areas involving harmonic analysis, singularities in minimal surfaces, geometric flows, microlocal analysis and mathematical general relativity. More information can be found on the symposium web page https://cse.umn.edu/math/riviere-fabes. 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 2026 · 2026-03

The human brain exhibits complex mechanical behavior. Its deformation under external forces depends on the extent and speed of loading. Rapid deformation of the brain during events such as blasts and automotive crashes can cause traumatic brain injury. Understanding the mechanical behavior of the human brain under such extreme conditions is critical to developing computer models for predicting brain injury. This knowledge is also needed to design safer personal protective equipment and brain injury management and prevention strategies. Unfortunately, the current understanding of the mechanical behavior of living humans' brains is restricted to small deformations and a narrow range of loading rates that do not represent the full spectrum of injury-causing conditions. This award supports fundamental research combining high-rate mechanical testing, analytical and computational modeling, and machine learning to generate insights into how the living human brain responds to large and rapid loading. Results from this research will positively impact U.S. national health and welfare and will contribute to the fields of tissue mechanics, traumatic brain injury, and machine learning. This project will lead to new courses and involve contributions from underrepresented minorities. The overarching goal of this research is to understand the high strain rate mechanics of the brain in its native biophysical environment. The first stage will focus on tissue responses under small deformations and dynamic strain rates. Wide-band Magnetic Resonance Elastography experiments will be conducted on brain tissue specimens from multiple brain regions to develop linear viscoelastic constitutive models. Multi-fidelity models will be developed to fuse the observed responses with available narrow-band in vivo brain tissue responses for predicting linear viscoelastic properties of the in vivo brain tissue in a wide range of loading frequencies. The second stage will focus on tissue responses under large deformations and extreme strain rates. Quasi-static and dynamic mechanical testing will be conducted to develop visco-hyperelastic constitutive models. Physics-informed multi-fidelity models will be developed to fuse the ex vivo visco-hyperelastic responses with the in vivo linear viscoelastic responses characterized in the previous stage. This study will significantly advance our understanding of brain biomechanics by generating insights into the relationship between in vivo and ex vivo tissue mechanics and the first-ever full-field maps of the living brain’s mechanical properties applicable under extreme loading conditions. This project is jointly funded by the Biomechanics and Mechanobiology (BMMB) program and the Established Program to Stimulate Competitive Research (EPSCoR). 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 2026 · 2026-02

This proposal seeks to fund US-based students to attend the 2026 International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks (WiOpt), held in Columbus, Ohio on June 3-6, 2026. WiOpt is a premier international forum that attracts high-quality, forward-looking research contributions and provides a vibrant forum for technical and professional exchanges. WiOpt 2026 will expose selected students to cutting-edge developments in the field and enable interactions with world-leading researchers. Students will gain feedback on their ongoing work, broaden their academic perspectives, and build lasting professional connections. WiOpt 2026 will feature topics on next-generation 6G networks, including AI/ML techniques for semantic and goal-oriented communications; security, privacy, and trust; distributed or multi-agent learning; and integrated sensing and communications (ISAC). This project supports students from US universities to attend the 2026 WiOpt conference in person. Students will have the opportunity to present their work and be exposed to state-of-the-art developments in the field. They will also have the opportunity to interact with peers from institutions worldwide, meet with senior researchers, and participate in discussions that are likely to shape the future of the field. 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 2026 · 2026-02

Rare diseases are individually rare, yet collectively affect millions of individuals worldwide. They pose persistent diagnostic challenges due to variability in clinical presentations, limited clinical expertise outside specialized centers, and fragmented patient data. These challenges are further compounded by small and geographically dispersed patient cohorts and strict privacy and regulatory constraints that limit data sharing across institutions and national borders. This makes it difficult to develop robust and widely applicable diagnostic models at any given site. This project addresses these fundamental barriers by developing a privacy-preserving artificial intelligence framework for identification of rare genetic disorders from observation alone, using a genetic disease of cell component, cilia, as a representative and scientifically challenging disease class. The central goal is to enable large-scale learning from distributed clinical data without direct exchange of patient-level information, reducing diagnostic delays and supporting faster and more accurate clinical decision-making. The proposed research develops a federated learning architecture that enables collaborative model training across multiple institutions while ensuring that sensitive patient data remain local and secure. A core technical contribution is deep clinical phenotyping through accurate extraction, normalization, and representation of observational information from both structured and unstructured electronic health records, leveraging multilingual natural language processing and large language models. These data representations are integrated with biomedical ontologies and rare disease knowledge bases and combined with patient similarity modeling to support rare disease recognition across a wide range of healthcare environments. The project includes algorithm development, cross-site federated evaluation, and real-world validation in collaboration with health care experts. The resulting methods and open resources are designed to generalize beyond the initial single group of diseases to a wide range of rare diseases. It will advance fundamental research in privacy-preserving machine learning and trustworthy AI while contributing to improved population health, international scientific collaboration, and responsible deployment of artificial intelligence in healthcare. 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 2026 · 2026-01

Rock deformation is a sub-discipline of Earth science that employs experimental techniques from geology and engineering to measure the strength of rocks. Observations and data from rock deformation are essential to a wide range of research in geoengineering, geologic hazards, geophysics, and planetary geology. However, most institutions do not have active research programs in rock deformation, due to the scale, cost, and technical needs of an experimental rock deformation lab. The Research Opportunities in Rock Deformation (RORD) REU site provides training in experimental rock deformation and generates a robust pipeline of students and industry professionals from all backgrounds. The RORD REU site provides research and mentorship opportunities for undergraduate students in the field of experimental rock deformation. The long-term objective is to expand the pipeline of students pursuing research or industry careers in rock deformation or related fields. Student participants receive training in research methods and professional development topics that provide a stable foundation for graduate school or related career paths. A large team of PIs and senior participants, composed of academic researchers in rock deformation, ensures that students who participate in the program have a deep professional network to support their future endeavors. Students are drawn from the full spectrum of higher education institutions. Strong emphases are placed on recruiting students from smaller colleges and universities that do not have research programs in rock deformation. The REU site includes three integrated sessions: a field session to introduce students to the geological study of deformed rocks, a laboratory session where students conduct experiments on specimens collected during the field session, and a conference session where students have the opportunity to present the results of their research projects. The REU site uses an innovated distributed model, leveraging the combined lab capacity of the PIs and other senior participants to support 10 students per year. 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 2026 · 2026-01

Streams and associated riparian areas supply water for a multitude of downstream uses, reduce wildfire risk, and provide habitat for many species. Streams are increasingly threatened by drought, flood, and changing patterns of land and water use. Accordingly, billions of dollars have been spent to restore these ecosystems so that they can continue to provide vital services for people and support valuable fish and wildlife resources. Stream restoration increasingly focuses on beavers, which were once common throughout North America but were harvested nearly to extinction in the 18th and 19th centuries. Beaver-based restoration, which includes reintroducing beaver to historically occupied habitat and building structures that mimic beaver dams, has the potential for far-reaching beneficial impacts on stream and riparian ecosystems. However, it is not clear how these benefits vary across different environments and with different restoration practices. This proposal will provide critical data on how beaver-based restoration improves stream and riparian health. Such information is critical for promoting responsible use of restoration funding and effective stewardship of natural resources in a changing environment. This proposal will investigate the effects of beaver-based restoration on stream and riparian habitats and associated fish and wildlife species across divergent precipitation regimes using a large-scale, five-year field experiment and a complementary set of artificial stream experiments. Specifically, investigators will evaluate: 1) the effects of beaver reintroduction and the construction of beaver dam analogs on fish, amphibians, birds, bats, and mammals in wet vs. dry climates, 2) the specific mechanisms underlying the effects of beaver-based restoration on key habitat features under different streamflow conditions, and 3) the efficacy of different beaver-based restoration practices for achieving specific habitat and biodiversity outcomes. The field experiment will feature four types of sites: 1) beaver reintroduction (the translocation of beavers into currently unoccupied sites), 2) beaver-mimicry (the construction of beaver dam analogs), 3) unrestored controls, and 4) beaver-occupied reference sites. These site types will be replicated on the wet west side and the drier east side of the Cascades Range in the Pacific Northwest. The artificial stream experiments will provide detailed mechanistic insight into how different types of beaver-related structures, and different spatial layouts of such structures, influence stream and riparian habitat variables under varying streamflow conditions. This project is jointly funded by the Divisions of Environmental Biology and Integrative Organismal Systems through the Partnership to Advance Conservation Science and Practice Program. 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 · 2025-11

Title: Electronic Microparticles That Help Microbes Clean Waste and Generate Energy Microorganisms can produce both electricity and useful chemicals from wastes, but there are multiple technical problems that prevent these capabilities from being fully realized in real world applications. These problems slow the reactions of bacteria and make energy harvesting inefficient. This project will create tiny particles that combine solar-powered electronics with bacterial activity to accelerate these important biological reactions. Each particle is designed to provide a surface at a voltage where certain microbes can best do what they are expert at: breaking down organic compounds. The bacteria enriched on these surfaces have a unique ability; their internal metabolism is connected to their outer surface. A metal electrode on the particle can collect electrical current from these bacteria and send it to a photovoltaic cell. The additional power drives hydrogen gas production from a second electrode, which can be collected or further converted into methane. These particles can flow freely through water, allowing them to access fresh nutrients and avoid many limitations of traditional systems. The silicon and glass-based particles are non-toxic, largely using materials common in sand and soil or using other materials, such as tiny aluminum or gold electrodes, that do not harm the surrounding environment. This technology opens new possibilities for wastewater treatment, energy production, and environmental monitoring. Our research supports a collaboration between students and scientists in microbiology and electrical engineering, offering unique training and educational opportunities. The discovery of electroactive microorganisms enabled a new generation of biological systems integrated with electronic devices. When microorganisms can transfer electrons directly to or from electrodes, new kinds of wastewater treatment, soil bioremediation, biofuel synthesis, biomaterial production, and biosensing are possible. However, all microbial electrochemical technologies are constrained by diffusional limitations and inefficiencies, especially in water with low buffering capacity. This project addresses these challenges through a novel approach: self-powered, mobile microparticles that integrate anodes, photodiodes, and cathodes. These devices, having electrodes smaller than a typical diffusion length, harness spherical diffusion dynamics to enhance reaction rates while the photodiodes bias electrodes to voltages favorable to microbial metabolism, simplifying operation and reducing resistance losses. In addition, the particles themselves are small enough to be carried by fluid flow, which allows them to constantly travel through new regions of solutions where nutrients are available. Our interdisciplinary effort bridges microbiology and electrical engineering to fabricate and systematically test these devices. This completely new platform for growing electroactive microorganisms free of diffusional limitations and external power sources has the potential for transformative advances in catalytic rates and increased throughput of microbial experiments. 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 · 2025-10

Pathogens originating from wildlife pose an increasing threat to agriculture and public health. Yet, predicting when and how these pathogens spill over into domestic animals or humans remains a significant challenge. This is mainly due to limited knowledge about which reservoir hosts interact with potential recipient species, how long those interactions last, and which types of contact are most likely to result in pathogen transmission. Recent advancements in animal biologging and pathogen genetic sequencing are opening new opportunities to map contact networks and trace pathogen transmission across species. This research integrates animal tracking data, pathogen genomics, and mathematical modeling to better understand and predict cross-species transmission dynamics. The project focuses on blood-feeding vampire bats, key wildlife reservoirs for rabies virus, and the livestock upon which they feed. By analyzing interactions between these species, the project aims to identify patterns of contact and transmission that can inform more effective surveillance and control strategies. Given rising concerns over emerging zoonotic pathogens, the research is timely. It not only advances methods to integrate diverse data streams to study spillover but also provides practical insights into managing vampire-bat livestock conflict, an issue of growing concern. The research will also broaden participation by mentoring trainees, strengthening scientific capacity in our study areas, and engaging communities through school programs and public outreach on infectious disease prevention. This project aims to (1) characterize dynamic, multi-species contact networks between vampire bats and livestock using animal-borne proximity sensors; (2) map pathogen-sharing networks by analyzing genetic similarity among common viral, bacterial, and protozoan pathogens (e.g., coronaviruses and rabies virus, hemoplasmas, and trypanosomes); and (3) develop statistical and simulation models to identify likely transmission routes, understand epidemiological dynamics, and inform rabies virus control strategies. The outcomes will advance our fundamental understanding of pathogen spread in complex host communities. Additionally, our integrative approach, combining animal tracking with pathogen diagnostics, will offer a valuable framework for studying cross-species transmission in other 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 · 2025-10

Modern artificial intelligence (AI) and data-intensive applications, from language models to scientific simulations, demand ever-growing computational capability. Yet, conventional electronic-based computing systems face a critical bottleneck: the energy and time required to move data between processors and memory now exceed the costs of performing the calculations themselves. This imbalance has slowed progress in AI and deep learning, while contributing to rising energy demands. This project explores a transformative computing paradigm grounded in photonic-based, in contrast to the conventional electronic-based paradigm, enabling information processing through light. Using diffractive optical neural networks (DONNs), our team will develop specialized, ultrathin computing systems composed of engineered nanostructures that manipulate light waves. These DONN-based systems are capable of performing inference tasks at the speed of light while consuming significantly less energy than conventional electronic platforms. These breakthroughs will support future embedded and edge devices, from smart sensors to autonomous systems, enabling AI to scale sustainably. The project will expand the American workforce in AI and optical computing by integrating education and outreach. The PIs will connect students from local community colleges to research opportunities and build upon successful undergraduate and K-12 outreach programs at both NC State and the University of Minnesota. The team will introduce new AI- and photonics-focused courses into the undergraduate curriculum and engage students through REU programs, design projects, and public-facing events such as youth AI labs, science competitions, and museum activities. Together, these efforts will prepare students with expertise in optics, nanofabrication, and machine learning, ensuring they are equipped to lead future innovations in photonic computing. This project seeks to design, fabricate, and experimentally validate a new generation of metasurface-based diffractive optical neural networks (DONNs) that overcome key limitations of existing optical accelerators. The research targets three major advances: (1) Multi-modality: developing a unified DONN framework capable of processing diverse data types, including images, text, and graphs, and integrating programmable metasurface layers to significantly broaden its application scope. (2) Scalability: enabling deep, large-scale network inference using iterative train-prune-retrain, weight clustering, and tile-wise sparsification to minimize diffraction errors and optical crosstalk, combined with nonlinear activation reduction techniques such as low-degree polynomial network approximations for efficient, stable inference; and (3) Robustness to physical non-idealities: integrating fabrication-aware optimization and phase smoothing to mitigate meta-atom geometry variations, inter-element crosstalk, and environmental instabilities. The DONNs will employ multi-channel metasurface layers that leverage wavelength and polarization multiplexing for parallel, multi-task processing, and will incorporate architectural strategies such as shared diffractive layers, spatially distinct task routing, and optical skip connections to extend functionality to transformer and graph-style architectures. Fabrication will follow a staged cleanroom-to-foundry pipeline, with TiO2 nanofin metasurfaces prototyped at NC State and scaled through commercial foundry tape-outs to achieve device areas exceeding 250 μm2. The resulting systems will be experimentally benchmarked for accuracy, energy efficiency, and throughput across different workloads. By tightly integrating algorithmic innovation, photonic device engineering, and experimental validation, this work will establish the foundations for compact, energy-efficient, and adaptive optical AI processors, offering a pathway toward practical deployment in embedded and edge-computing applications. 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 · 2025-10

Traditional machine learning often involves collecting data from multiple sources, which can raise significant privacy concerns. One approach has emerged as a promising solution to solve this challenge by enabling models to be trained across many different sources without directly sharing private data. This approach has become particularly valuable in sensitive sectors such as medical diagnostics, where individual data privacy is legally protected. Despite these advancements, existing systems for training models across multiple sources lack standardized assessment tools, posing challenges to research reproducibility, validation, and trust. Without proper testing tools, organizations cannot verify that their privacy protections work as intended, creating barriers to adoption in critical areas like healthcare, finance, and national security. This project addresses this challenge by developing comprehensive testing tools that ensure privacy-preserving artificial intelligence systems work reliably, serving the national interest by enabling secure collaboration on AI development while protecting individual privacy, supporting American competitiveness in artificial intelligence technologies, and strengthening data security across critical infrastructure. This project designs, develops, and sustains FLTest, an interdisciplinary testbed that automates privacy and robustness evaluations in federated learning systems, addressing gaps often overlooked by traditional tools. The research activities include developing automated test orchestration frameworks, implementing privacy attack simulation models, creating configuration vulnerability detection systems, and building recommendation engines for optimization. The testbed's key innovation streamlines evaluations through automated orchestration assisted by a pitfall checker that detects configuration issues and vulnerabilities in privacy evaluations. FLTest empowers both novice and expert users with actionable insights tailored to real-world applications. The team will validate FLTest across multiple domains and datasets, develop standardized benchmarks for assessment, and create detailed reporting mechanisms for security analysis. By utilizing distinct datasets and offering a standardized solution, FLTest verifies model privacy and robustness across heterogeneous data distributions, supporting the development of reliable privacy-preserving federated learning systems. The project includes collaboration with three industry partners to ensure practical adoption and long-term sustainability. 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 · 2025-10

This project aims to serve the national interest by improving curricula for strengthening students' data analysis skills in solid Earth geoscience courses. The ability to process, analyze, and visualize data is increasingly vital for conducting independent research and preparing students for the STEM workforce. The project plans to advance geoscience education by developing upper-level teaching modules that will help students use advanced data analysis to explore the mechanical behavior of the solid Earth. This collaborative IUSE Level 1 Engaged Student Learning track project will also lead and facilitate faculty professional development to enhance data-centric teaching approaches for solid Earth courses. To support the development and application of data analysis and strengthen the quantitative reasoning and self-efficacy of geoscience students, the goals of the project are to: 1) create undergraduate teaching materials for investigating solid Earth geoscience problems with a focus on data analysis skills; 2) lead and facilitate professional development for increasing faculty confidence in using data-rich teaching activities in solid Earth courses; and 3) evaluate the impact of data-rich teaching modules on changes in students' confidence in data-analysis skills in solid Earth courses. The teaching modules will provide freely available instructional faculty resources to teach data analysis in solid Earth contexts to help students develop critical thinking and other skills necessary to succeed in STEM disciplines. During the project, approximately 30-40 instructors and over 400 students will benefit from these efforts. The Science Education Resource Center (SERC) will provide research support, assessment, project evaluation, and website hosting. The National Association of Geoscience Teachers (NAGT) will support dissemination. The project plans to contribute knowledge of how professional development activities may increase instructors’ confidence in including more data analysis in their courses. The project also aims to enhance the quantitative reasoning and data analysis skills of geoscience students, preparing them for graduate research, careers in the geoscience workforce, and engaging with challenges such as natural hazards and energy resources. The NSF IUSE: EDU Program supports research and development projects to improve the effectiveness of STEM education for all students. Through the Engaged Student Learning track, the program supports the creation, exploration, and implementation of promising practices and 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 · 2025-10

People’s body movements can reveal a lot about students’ mathematical reasoning. Examples such as using one’s whole body to explore properties of geometric shapes and raising eyebrows during mathematical insight illustrate ways people’s movements are closely linked to how they think. Despite the value of these nonverbal indicators of students’ learning experiences, this body-mind connection remains understudied. In this project various data sources are gathered from learners playing a video game designed to improve mathematical reasoning through movement and speech. While learning, students occasionally use their bodies to express mathematical insights and trouble spots. These events are important enough that they can influence students’ learning and attitudes toward mathematics, but subtle enough that teachers may miss them in the buzzing dynamics of the classroom. AI will help the research team select when to interview students about these rare but significant moments in their learning, combining the speed and pattern recognition of computers with the depth and insight of humans' natural conversation. This approach creates a rich dataset for analysis and to develop design principles that support mathematics learning through movement. The findings will advance a deeper understanding of how people learn using nonverbal and verbal thought processes and ways to better support these thought processes. The broader impacts include improving mathematics learning for everyone, especially for those who rely on nonverbal thought processes that might be overlooked using current learning designs. In this project, investigators target two common and critical moments in mathematics learning: 1) forming mathematical insights, and 2) encountering trouble spots that indicate a student is struggling to adapt their understanding to new information. Findings from the five phases of research provide the investigators with detailed findings that can be leveraged to develop new theories and improve classroom learning. In Phase 1 of the project, initial data are collected (including movement, speech and interaction) as students play a mathematics learning game that encourages them to use their bodies. Phase 2 involves creating automated AI detectors, driven by scientific hypotheses, that can recognize students' insights and trouble spots in real-time. In Phase 3, additional data collected with the detectors from phase 2 alert trained interviewers to critical events during student learning, prompting data-driven interviews that gather evidence on cognitive, metacognitive, and affective processes. Phase 4 uses learning analytics and data from these AI detection-driven interviews to build a theoretical model explaining the interplay of cognition, metacognition, affect, and embodiment. Finally, Phase 5 generates practical design principles for classroom instruction and embodied learning technologies, informed by the resultant model and empirical findings. The broader impacts include improving mathematics learning for everyone, especially for those who rely on nonverbal thought processes that might be overlooked using current learning designs. 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 · 2025-10

While generative artificial intelligence (GenAI) tools and methods could help researchers discover new insights about STEM teaching and learning, researchers face three major problems: First, they haven't been trained how to use these technologies in their research. Second, the training that exists is often too technical or not specific to STEM education research. Third, researchers at resource-limited colleges and universities have even fewer opportunities and resources to learn about GenAI than their colleagues at other institutions. These challenges could mean that talented researchers will be left behind as GenAI transforms how research is designed, conducted, and disseminated, ultimately slowing advancements in STEM education research. This project addresses these challenges by creating a training program that gives researchers at resource-limited institutions the skills they need to evaluate and use GenAI effectively, which will lead to better STEM education for all students and support the National Science Foundation's goal of advancing scientific progress. The project aims to teach early- and mid-career education researchers how to use GenAI to improve studies of how students learn science, technology, engineering, and mathematics. This BCSER Institutes for Methods and Practices project is designed to build the GenAI capacities of early- and mid-career STEM education researchers by training 50 participants over the course of the institute. The project aims to (1) increase the number of researchers who can effectively, ethically, and responsibly use generative AI (GenAI) in their work; (2) strengthen participants’ professional networks; and (3) expand the broader research community applying GenAI methods in STEM education. In Year 1, the team will design and launch the learning experience and curriculum. In Years 2 and 3, two cohorts of 25 researchers will engage in a year-long program that begins with a 5-day intensive, in-person training on GenAI tools, techniques, and applications across the research lifecycle—including practical and ethical considerations. This will be followed by an extended online community of practice that offers resources and support as participants integrate GenAI into their own projects. By the end of the institute, participants will have developed the skills, judgment, and confidence to use GenAI in research and continue learning in this rapidly evolving field. The project is supported by NSF's EDU Core Research Building Capacity in STEM Education Research (ECR: BCSER) program, which is designed to build investigators' capacity to carry out high-quality STEM education research. The project is also supported through a collaborative NSF activity with the Bill & Melinda Gates Foundation, Schmidt Futures, and the Walton Family Foundation. 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 · 2025-10

Twenty-first century research is enabled by the availability of vast amounts of data, collected across a wide range of temporal and spatial scales, often in real time. Entire fleets of satellites, drones and other devices monitor the Earth at ever-increasing resolution, adding to the enormous corpus of maps and metadata that enable sciences from geology to biodiversity. While citizen science - or crowdsourcing science - has been successfully leveraged over the past several decades to close the analysis gap arising from large amounts of data, the sheer scale and complexity of these new data sets presents new challenges. This project addresses these challenges through novel Citizen Science Cyberinfrastructure (CSCI) where new tools and techniques, including teaming humans with Artificial Intelligence (AI), are being developed to enable researchers to efficiently extract the best results from large and complex data sets. This effort incorporates mapping, machine learning, and data sharing innovations in biodiversity, geoscience, and astronomy research and expands the capacity of research communities across a wide range of disciplines to use citizen science as a suitable, open and sustainable research methodology. This project leverages NSF-supported cyberinfrastructure, including the Zooniverse citizen science platform with its nearly three million volunteers, to provide a novel cyberinfrastructure by: (1) integrating mapping infrastructure into Zooniverse to accelerate accurate processing and curation of often complex data sets for geoscience and biodiversity projects; (2) providing an "incubator" hub for researchers and developers to design and deploy innovative citizen science projects with a fast production turn-around particularly for AI training; and (3) providing cyber-pathways to existing cyberinfrastructure, creating documentation for Zooniverse projects to responsibly deposit their data into appropriate repositories, and facilitating workshops for Zooniverse project teams to develop best practices for data and model sharing. To ensure wide dissemination of the new CSCI, a strong Community of Practice is engaged and supported throughout the project effort. The project is led by the University of Minnesota in collaboration with the Adler Planetarium and the University of Oxford, as core Zooniverse institutions and astrophysics expertise, and joined by the University of Florida and the Florida Museum of Natural History with expertise in biodiversity, museum specimen collections, and data repositories, as well as Northern Arizona University with expertise in Earth Sciences, mapping cyberinfrastructure, and the study of the Antarctica’s Dry Valleys. As the largest citizen science platform, and with key tools for human-in-the-loop data analysis tasks, the Zooniverse is uniquely positioned to develop skills and best practices to lower the barrier for research teams to responsibly share data and models. This is carried out through fact-finding and demonstration with several key groups who are exemplars in the data and model sharing ecosystem, including GitHub, the National Science Data Fabric (NSDF), the Global Biodiversity Information Facility (GBIF), and Open Science Chain. 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 · 2025-10

The conference "Equivariant, Motivic, and Physical Topology in the Midwest’' will take place October 24-26, 2025, at the University of Minnesota, Twin Cities. This conference will bring together leading researchers working in a wide array of different aspects of algebraic topology. Topology — the study of the intrinsic properties of geometric objects — has found broad application in recent years in other branches of mathematics such as geometry and the study of symmetries, as well as in high energy physics. This conference will promote the continued interactions of these disciplines through a series of lectures by experts as well as providing a venue for collaboration. The conference will focus on equivariant homotopy theory, motivic homotopy theory, physics, and symplectic topology pertaining to physics. There has been ongoing progress at the intersection of these fields, and these interactions have been fruitful for both the development of theory and the advancement of computations. The invited speakers are on the forefront of the recent developments in these fields, and their work highlights the intersections. Further details, including the program and speaker information, will be available on the conference webpage at https://cse.umn.edu/math/events/equivariant-motivic-and-physical-topology-midwest. After the conference, this website will be used to post the recorded talks, to allow the broader community to benefit from our scientific program. 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.