Rutgers University New Brunswick

NSF Awards · FY 2026 · 2026-09

Free boundary problems are models where one of the unknowns is a shape or interface rather than a function, with part of the model describing the rate at which the interface moves or the shape changes (much like how a differential equation describes how a function changes). This kind of model arises naturally in the study of fluids (the waves on the surface of a body of water), petroleum engineering (the evolution of a fully saturated region in a porous material), phase transitions (the shape of a melting block of ice), and combustion (the motion of a flame front in a forest fire). Our current mathematical tools work best for steady-state solutions to such problems, and moreover to ones which minimize an energy. The purpose of this project is to develop approaches to study moving interfaces and non-minimizing steady states. Better mathematical understanding may lead to smarter and safer approaches to the applied problems through rigorous approximation schemes, analysis of stability under perturbations, and rigid qualitative properties of solutions. At the same time, the project trains graduate and undergraduate students in a mathematical subject with important industrial applications. The specific topics covered by the project are compactness theorems for critical points to Bernoulli-type free boundaries, and applications of these to gravity water waves and other min-max constructions; general free boundary problems, arising from semilinear elliptic equations not admitting a strong maximum principle, with the goal of treating them all together with minimal assumptions on the structure of the nonlinearity and the minimality or positivity of solutions; and parabolic free boundaries of several types, including Bernoulli and free transmission, for which minimality is not even an available concept, and the task is to understand the structure of the free boundary to a level not previously possible because of this. The approach here represents a complete shift in perspective on nonconvex free boundaries and provides a framework to study the much richer global structure of nonminimizing solutions. Postdoctoral researchers, graduate students, and undergraduates are involved in the work of the project. 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

This award provides partial support for “Several Complex Variables and Complex Geometry: A Satellite Conference of the 2026 ICM in Philadelphia” which will be held at Rutgers University in New Brunswick, New Jersey, from July 17–21, 2026. The event will serve as an excellent platform to highlight recent and exciting advances in the field, offering a broader scientific perspective on Several Complex Variables and Complex Geometry. This important area of mathematics has experienced remarkable progress in recent years, fueled by deep and productive interactions with many other branches of the discipline. The significance of this growth is reflected in the ICM 2026 program, which features one plenary speaker and several invited speakers representing the broadly interpreted domain of SCV and Complex Geometry. This satellite conference aims to capture that broader scope and foster further advances by providing a valuable opportunity for interaction among researchers. The scientific program will cover, but is not limited to, the following topics: Dbar-equations in Complex Analysis and Geometry, Kohn’s Multiplier Ideals and Their Applications in Complex Analysis and Geometry, Rigidity Problems in Complex Analysis and Geometry, Normal Form Theory and Complex Dynamics, Asymptotic Expansion of Bergman Kernel Functions and Related Topics, Holomorphic Deformations of Compact Complex Manifolds, and Nevanlinna Theory and Complex Hyperbolicity. It will be especially beneficial for scholars in the United States, including young researchers, by offering a unique and convenient venue to engage with numerous leading experts and peers from around the world. More information about the conference is available https://sites.rutgers.edu/scvcgc/ . 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

This award supports a five-day conference titled Lagrangian Floer Theory and Applications to be held September 28--October 2, 2026, at the Center for Mathematical Sciences and Applications in Cambridge, Massachusetts. The meeting is organized by Chris Woodward (Rutgers) with the assistance of Denis Auroux (Harvard) and Jonny Evans (Lancaster). Lagrangian Floer theory arose from research on Hamiltonian dynamics, which is the mathematical study of such systems as the motion of the planets. In recent years, Floer theory has become connected to quantum invariants in topology. The meeting will bring together experts at the frontier of research in these areas from around the world. The conference is expected to generate transfers of knowledge, new collaborations, and a cross-fertilization of ideas, and further inspire graduate students and early-career mathematicians. The web-page for the workshop is https://cmsa.fas.harvard.edu/event/lftworkshop/. The scientific themes of the event include applications to dynamics and classification problems; connections with singularity theory; mirror symmetry and tropical techniques; categorical aspects; and Floer homotopy theory. Each day of the meeting will focus on one of these themes. There will be approximately 25 invited speakers, with each speaker giving a one-hour talk. The list of speakers includes two graduate students, five postdoctoral researchers, and five tenure-track assistant professors. The speakers come from small liberal arts colleges, distinguished public and private research universities, and highly regarded international research institutes and universities. The award supports invited speakers, and participants who are U.S.-based graduate students or postdoctoral fellows. 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

Rutgers Symplectic Summer School 2026 will take place at Rutgers University in New Brunswick, New Jersey, from August 17–21, 2026. Building on the success of the first two editions of the summer school, the conference will bring together graduate students, postdoctoral researchers, and faculty working in symplectic geometry and related areas of mathematics for a week-long program consisting of minicourses, research lectures, and collaborative interactions. Symplectic geometry is a branch of mathematics that grew out of the mathematical study of motion and classical mechanics and now plays an important role in modern geometry, dynamical systems, and mathematical physics, including areas connected to quantum theory. The program is designed to provide advanced training for young researchers while also fostering communication across neighboring areas such as topology, geometry, and mathematical physics. By creating opportunities for mentoring, professional development, and broad scientific exchange, the conference aims to strengthen the research community and support the next generation of mathematicians. Website: https://sites.google.com/view/rsss2026/home The Rutgers Symplectic Summer School is intended to establish a recurring meeting focused on current developments in symplectic geometry and its interactions with topology, algebraic geometry, and gauge theory. The 2026 edition will feature several minicourse series delivered by leading researchers together with research talks highlighting recent advances in the field, especially the arithmetic feature of symplectic geometry. A central goal of the program is to provide graduate students and early-career researchers with accessible introductions to active research areas while promoting collaboration among participants from different institutions and backgrounds. The award will support participant travel and related activities for the 2026 edition of the summer school. 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

Scientific research is changing quickly because advanced automation and Artificial Intelligence (AI) can now run laboratory experiments on their own. This technology helps scientists make new discoveries much faster than before. Building these automated laboratories requires combining ideas from different fields, including robotics, and laboratory science. Currently, researchers in the United States and Japan do not have an easy way to work together and combine their unique strengths in these areas. This project addresses this problem by organizing an online conference series. The virtual event will bring together scientists and engineers from both nations to build an international network. These partnerships will push scientific progress forward and help the economy grow through faster technological innovation, serving the national interest. In addition, the conference supports training by involving early-career researchers. The goal of this project is to create a partnership between researchers from the United States and Japan focused on combining automated laboratory equipment and AI. Research topics include flexible soft robotics, human-robot interaction, and programmable cloud laboratories for automating scientific testing. To achieve these goals, the investigators will host a two-day online workshop in late May 2026. The event will feature presentations and panel discussions that pair international experts from both countries. To help participants build partnerships, attendees will share research profiles and capability summaries to match interests between researchers based in the United States and Japan. The team will coordinate follow-up breakout sessions throughout June to design concrete joint research plans. The core activities and findings will be written into a comprehensive report summarizing the new collaborative framework. 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

Memory required to perform a computational task is one of the most fundamental measures used by theoretical computer scientists to assess how difficult a task is. Nevertheless, in practice, memory optimization received limited attention until the emergence of big data applications. More recently, the rapid growth of large-scale machine learning (ML) systems, including large language models (LLMs), has pushed model parameter counts far beyond improvements in memory hardware, raising concerns that memory may soon become the primary bottleneck in serving them. This growing need for memory-efficient alternatives is further amplified by interest in on-device learning, driven by concerns over data security, transmission costs, and the demand for personalized applications. The field of learning theory has already deeply explored the data and time required for various ML tasks; however, our understanding of their memory requirements remains limited. The goal of this project is to systematically address this gap and develop a foundational theory of the capabilities and limits of memory-constrained learning. The new approaches explored in the project aims to unconditionally answer whether the data and time requirements to learn drastically increase when using low-memory algorithms – which include the commonly used methods in practice such as stochastic gradient descent. This project draws tools from various mathematical areas such as complexity theory, learning theory and information theory. The educational plan of this project involves training of undergraduate and graduate students through (1) foundational courses in theoretical computer science (TCS), as well as advanced courses at the intersection of these mathematical areas, and (2) supervised undergraduate and graduate research aligned with the themes of this project. The first two thrusts of this project will systematically characterize the memory requirements of fundamental supervised and unsupervised machine learning (ML) tasks in the streaming model – a setting that closely reflects the realities of large-scale data processing. Despite extensive work on streaming algorithms, quantifying the memory requirements of ML tasks in this model presents new challenges that this project aims to address. First, the streaming literature typically focuses on worst-case data streams, whereas learning theory commonly assumes that data are sampled independently and identically from an underlying distribution. Second, establishing memory requirements even when the learner has access to a super-polynomial amount of data requires going beyond the sample regimes considered in independently and identically streams. The third thrust of this project will establish memory requirements in learning paradigms beyond the sample-based model, such as machine unlearning and learning in the query model, with broader implications to LLM safety and computational complexity theory, respectively. 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

The project investigates how to make intelligent robots more reliable, understandable, and safe when performing complex tasks over long periods of time. Modern learning-based systems can achieve impressive performance, but they often lack guarantees about correctness and safety, especially when tasks involve sequencing, repetition, and conditional behavior. The project addresses these challenges by combining robot learning with rigorous reasoning techniques. The project's novelties are a formulation that embeds temporal logic task constraints directly into the reinforcement learning objective to translate temporal progress into informative reward signals, a mechanism for automatically inferring temporal task specifications, and compositional methods for verifying long-horizon agent behaviors. The project's impacts are the development of trustworthy robotic systems capable of operating in uncertain environments, improved accessibility of formal methods for non-expert users, and new foundations for reliable AI/ML systems. The project will also benefit AI/ML tools and toolchains by introducing verification components that can be integrated into existing learning pipelines. In addition, the resulting algorithms and software provide reusable capabilities for specification inference and compositional verification in broader AI systems. The project will introduce formal methods into K–12 and undergraduate classrooms through the development of an interactive robot simulation tool that connects formal task specifications with robot behaviors, supporting hands-on coursework and mentoring activities that prepare a workforce trained in formal methods for building trustworthy systems. The project develops a principled framework for synthesizing and verifying robot control policies for Linear Temporal Logic (LTL) objectives in continuous, stochastic environments. The framework leverages Limit-Deterministic Buchi Automata (LDBAs) as an abstract representation to guide both learning and verification. First, it introduces LDBA-guided policy learning that generates temporally aligned reward signals from the agent's own exploratory trajectories, enabling efficient learning under sparse supervision. Second, it proposes methods for inferring temporal objectives and event detectors directly from demonstrations, reducing reliance on manually specified LTL formulas and enabling intuitive task specification. Third, it develops a compositional verification methodology that decomposes infinite-horizon temporal properties into shorter-horizon subproblems, enabling scalable reasoning through inductive invariants with probabilistic guarantees in black-box settings. Together, these contributions will significantly improve the reliability, scalability, and safety of robot learning systems, advancing the scientific foundations of formal methods for AI and supporting trustworthy autonomous behavior in safety-critical 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 · 2026-05

This award supports participation in the conference "Analysis of Partial Differential Equations arising in Physics" held July 13-16, 2026 at Rutgers University in New Brunswick, NJ. The conference is a satellite event for the 2026 International Congress of Mathematicians taking place July 2026 in Philadelphia, PA. By leveraging the excellent research environment of the northeastern United States, the event gathers global experts and early-career researchers to study the equations governing fluid dynamics, general relativity, and quantum mechanics, thereby reinforcing U.S. leadership in mathematical physics and fostering international collaboration. This project promotes the progress of science by creating a forum for specialized communities to share insights, address complex challenges, and support the development of the next generation of researchers in fields critical to our understanding of nature. The goal of the conference is to bring together researchers in general relativity, fluid dynamics, and quantum mechanics who are unified by their use of hyperbolic and dispersive partial differential equations (PDE). While these fields often advance independently, this forum facilitates the cross-pollination of techniques necessary to address fundamental challenges in mathematical physics. The technical scope of the meeting focuses on five key pillars: Asymptotic behavior and the long-term stability of coherent structures like solitons and black holes; Singularity formation and stability in semi-linear and quasi-linear PDE; Rigorous theories of turbulence in fluid and wave equations; Inverse problems and the scattering of waves; and Propagation of randomness and invariant measures for hyperbolic and dispersive differential equations. Information about the conference is available at https://sites.math.rutgers.edu/~mv715/icm26sat. 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

Metal nanoparticles often serve as catalysts in chemical manufacturing. They use light to promote chemical reactions. However, the details of how they work remain unclear. The reactions may be driven by heat or by energetic electrons generated when light hits the particles. Using new nanoscale thermal measurement tools, this project will uncover the mechanisms underlying the performance of metal nanoparticles. The results will improve catalysts that can use sunlight to drive chemical reactions with greater efficiency and with better control over the final products. This, in turn, will enable the design of more efficient energy conversion technologies. The project will contribute to workforce development by training students in advanced nanotechnology and measurement techniques, by providing hands-on learning to engineering education, and by engaging undergraduate and high school students in research. This project will directly measure and distinguish between the electron and lattice (phonon) temperature fields within individual nanoparticles and their supporting structures under resonant plasmonic excitation. To achieve this, the project will apply a novel in situ measurement technique called Two-Temperature Scanning Thermal Microscopy that is capable of simultaneous, nanometer-resolution mapping of electron and phonon temperatures. The research will quantify the nanoscale heat source density generated within individual nanoparticles, extract the local electron–phonon coupling rate governing energy exchange between subsystems, and measure the interfacial thermal resistance between nanoparticles and their supporting substrates, which collectively control heat dissipation pathways. The study will further investigate the role of multi-particle interactions, near-field coupling, and geometric and quantum confinement effects on energy transport and local temperature rise in nanoparticles. In addition, the project will explore thermal engineering strategies by tailoring nanoparticle geometry and support interfaces to control energy flow and enhance photocatalytic performance. Experiments will combine controlled nanofabrication, optical excitation, and in situ thermal measurements with reaction kinetics to establish quantitative relationships between nanoscale non-equilibrium electro-thermal transport processes and photocatalytic reaction rates. These results will provide a predictive framework for the informed design of next-generation energy conversion materials and microelectronics, especially systems that leverage quantum confinement of energy carriers combined with highly non-equilibrium transport to achieve high performance. 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

Tensor contraction sequences are foundational computations in quantum chemistry, quantum physics, and materials science. As these workloads grow in scale and complexity, scientists increasingly rely on optimization techniques — mixed precision, reordering, and fusion — to reduce runtime and memory costs. While effective, these optimizations can introduce subtle numerical errors that are difficult to detect and that erode confidence in scientific results. The project's novelties are a scalable formal verification framework for optimized tensor contraction sequences and a principled method for connecting correctness guarantees directly with performance optimization. The project's impacts are more reliable and reproducible scientific computing, open-source tooling for the broader high-performance computing (HPC) and formal methods communities, and interdisciplinary training opportunities that bridge formal verification, numerical analysis, and high-performance computing. This project develops a scalable, sound, and precision-aware formal verification framework for reasoning about optimized tensor contraction sequences under interacting optimization strategies. The project creates automated modeling and property-checking methods for mixed-precision tensor contraction sequences using domain-specific satisfiability modulo theories (SMT) encodings in quantifier-free floating point (QF_FP), automated formula generation, and counterexample-guided abstraction refinement. It extends verification to contraction-order optimization and fusion through incremental SMT solving with solver-state reuse, portfolio-based solving, and equality-saturation methods that preserve tight numerical bounds. The project further integrates verification into multi-objective design space exploration, using interaction-aware analysis and graph-based performance–error optimization to identify Pareto-optimal implementations across FLOP count, memory usage, and verified numerical error. The expected result is a practical methodology for discovering tensor contraction implementations that are both high-performance and provably correct for mission-critical scientific workloads. 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

This award will partially support US researchers to attend a five-day conference on July 13-17, 2026, at Rutgers University, New Brunswick, NJ, an official satellite event for the International Congress of Mathematicians (ICM), July 23-30, 2026, in Philadelphia, PA. The ICM is the most prestigious conference in mathematics and occurs every four years. Because of the location of Rutgers University near Philadelphia, conference participants will have a unique opportunity to interact with international mathematical leaders and rising stars. The conference will bring together distinguished senior speakers and a wide range of junior mathematicians, and will thus help to support, train and encourage the next generation of researchers. The scientific themes of the Rutgers conference include i) complex geometry, ii) gauge theory in dimensions 3 to 8, iii) geometric analysis, geometric flows, and minimal surfaces, iv) low-dimensional topology, knot theory, and Floer homology, and v) symplectic geometry and Gromov-Witten invariants. Each day of the meeting will focus on one of these themes. The conference is expected to generate transfers of knowledge, new collaborations, and a cross-fertilizations of ideas, and further inspire graduate students and junior mathematicians. The conference website is located athttps://tinyurl.com/36hn42by and maintained by Rutgers University. 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

The International Symposium on Computational Geometry (SoCG) is the oldest and most influential international conference in computational geometry. The premier event in the field, SoCG brings together researchers from all over the world to report on the state of the art in the fields of computational geometry and many related disciplines. The annual SoCG conference is now part of Computational Geometry Week (CG Week), which incorporates workshops and the Young Researcher Forum, as an effort to further broaden participation and visibility of the community. Research in geometric computing has had broader implications in technological development in many areas of engineering and science, including computer vision, geographic information systems, entertainment, virtual reality, graphics, manufacturing, networks, and geometric data processing. The computational geometry research community is truly international, with significant representation on at least five continents. CG Week 2026 (including the 42nd annual SoCG) will be held on the Rutgers campus in New Brunswick, New Jersey. While Rutgers is relatively accessible from most of the USA, the cost of travel is prohibitive for many students and young researchers at institutions outside the greater New York area. This project subsidizes travel to CG Week in New Jersey (June 2-5, 2026) for US-based student participants to engage in the important technical, professional, and social exchanges that SoCG fosters. 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

This award supports the 23rd annual Graduate Student Topology and Geometry Conference, to be held April 24-26 at Rutgers University in New Brunswick, New Jersey. The conference is designed by and for graduate students; organization is primarily student-led, and the majority of the talks are given by student participants. Such a gathering provides junior researchers from many institutions and geographic regions with an opportunity to present their work, to engage with research in the frontiers of geometry and topology, to interact with experts in their field, and to meet and network with same level peers and develop the seeds of future collaborations. In addition to a full slate of student talks, the conference will also feature three plenary talks from established faculty Tara Holm (Cornell), Aaron Naber (Institute for Advanced Study), and Lisa Piccirillo (University of Texas Austin), along with six talks from early career faculty speakers Ricardo Caniato (California Institute of Technology), Nir Gadish (University of Pennsylvania), Robbie Lyman (Rutgers University Newark), Abhishek Mallick (Dartmouth), Allison Miller (Swarthmore), and Juan Muñoz-Echániz (Simons Center for Geometry andTopology). Careful attention has been given to the inclusion of a mathematically broad range of speakers capable of acquainting their audience with a wide array of cutting-edge topics in geometry and topology. The conference website may be found at https://www.gstgc2026.com/. 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

Volumes and varieties of data generated by ocean observing systems and long-term research have increased substantially in recent years. Yet, many early career researchers lack essential skills that are needed to find, access, use and reuse these data. The EMBARK collaborative project will address this gap by teaching skills to minimize the need to move large data by accessing and (re)using data with software workflows. A short-format, virtual training course will be developed and delivered for graduate students and postdoctoral researchers in the ocean sciences. The open-access course materials will provide practical, hands-on training to promote flexibility in adopting cyberinfrastructure (CI) tools and resources provided by major facilities and data repositories in the ocean sciences. By improving CI literacy and enabling the (re)use of large, complex, and/or long-term ocean datasets, this project will promote research to better understand the ocean and its role in global environmental and ecological processes. This project serves the national interest by promoting the progress of science, expanding utilization of NSF-funded research infrastructure, and strengthening the American STEM workforce. The EMBARK collaborative project promotes the adoption of cyberinfrastructure (CI) tools, methods, and resources by early career ocean professionals (ECOPs) as CI Users. Specifically, the goal is to train early-career researchers with skills to use and reuse data from major facilities and data repositories in the ocean sciences. This project will develop and deliver virtual workshops to three cohorts, with a total of ~90 participants, focused on developing key skills including data discovery, access via application programming interfaces (APIs), and software workflows for data harmonization and reuse. Training materials will be developed through an iterative design process with CI experts and ECOPs. The materials will then be packaged into modular learning units inclusive of executable Jupyter Notebooks and practical exercises. The training will be based around four ocean science use cases. The use cases will illustrate workflows for accessing and reusing oceanographic time series, satellite data, model data, and imagery classified by machine learning. The efficacy of the training will be evaluated to assess changes in learners’ CI awareness, confidence, and skill application. All training materials will be openly accessible and disseminated for broader use and future development, ensuring use beyond the participants in the project. 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 · 2025-12

Minerals are essential to understanding our planet's past and guiding its sustainable future. This project builds on Mindat.org, the world's largest public database of minerals and their global distributions, which receives nearly 10 million visitors per year. With previous support from the National Science Foundation, the OpenMindat project has made mineral data more accurate and accessible for scientists and the public alike. This new phase of work, called OneMineralogy, will expand these efforts into a broader open science ecosystem. OneMineralogy aims to make mineral data and tools available to a wider range of users, including students, educators, researchers, and decision-makers. Its data, tools, and services will help them ask and answer big questions about how Earth's systems have evolved over time. The activities of OneMineralogy will strengthen science education, promote open data sharing, support mineral exploration (including critical minerals), and accelerate discoveries related to planetary science, Earth-life co-evolution, and environmental change. By fostering new partnerships and offering training opportunities, OneMineralogy will empower the next generation of scientists and ensure that mineralogical data benefit society as a whole. Scientifically, OneMineralogy will advance the frontier of data-driven geoscience by developing new data curation strategies, computational tools, and community engagement programs. It builds on the success of the NSF-funded OpenMindat project, which has already improved the quality and accessibility of over 6,000 mineral species records and data from more than 400,000 global localities. OneMineralogy will consist of three major activity clusters: (1) extending and curating data to support a wider range of geoscientific research, (2) building a data science toolbox to enable large-scale analysis of mineralogical systems, and (3) conducting workshops and outreach programs to grow the user community and build capacity. The project will integrate Mindat with other open cyberinfrastructure resources related to paleoenvironments, geomicrobiology, tectonics, and biosignatures. The enriched data and tools will provide strong support to the investigation of Earth's dynamic history through deep time. It will also provide a foundation for interpreting the growing body of mineralogical data from planetary missions to the Moon and Mars. Through these activities, OneMineralogy will create a sustained, open, and collaborative ecosystem to support transdisciplinary research and education in the Earth and planetary sciences. This award by the Geoinformatics program is co-funded by the Innovative Technology Experiences for Students and Teachers (ITEST) program, which supports projects that build understandings of practices, program elements, contexts and processes contributing to increasing students' knowledge and interest in science, technology, engineering, and mathematics (STEM) and information and communication technology (ICT) careers. 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 intrinsic high strength, light weight, and heat, corrosion, and irradiation resistance of ceramics positions them as nearly ideal structural materials. However, their inability to resist the growth of cracks causes any tiny flaw to grow into a catastrophically large crack; this renders ceramics brittle and impractical as structural materials for automotive, energy, aerospace, and defense applications, to name a few examples. The overarching goal of this Designing Materials to Revolutionize and Engineer our Future (DMREF) project is to transform the fracture resistance of ceramics by introducing heterogeneous metallic features across multiple length scales into the ceramic material. The fundamental challenges to be overcome in the project are: (1) how can heterogeneous ceramics be computationally designed when the space of possible designs is massive? and (2) how can such engineered ceramics be manufactured? These challenges look to be overcome by leveraging recent advances in machine learning for material modeling in conjunction with advanced low-temperature ceramic processing techniques. The revolutionary new class of materials to look to be designed through the project can be directly implemented in commercial applications, such as satellite structures, low-wear medical devices, armor, and hypersonic vehicles. Insights gained on the design, processing, and fracture of heterogeneous ceramics seek to drive future innovations enabling next-generation structural materials. Despite many decades of development of advanced ceramics, fracture toughness values have remained consistently below about 15 MPa-m0.5, a factor of three less than typical structural metals. This DMREF project aims to transform the toughness of ceramics through the introduction of hierarchically heterogeneous metallic interphases that will drive toughening via crack multiplication and deflection ahead of an advancing crack tip. The structure of the interphases and their distribution in the microstructure looks to be designed using molecular dynamics and finite element simulations of fracture. To accelerate computational materials design, finite-element-based physics-informed neural networks seek to be employed after training on cohesive surface and phase field finite element fracture models. The computationally designed ceramic microstructures aim to be produced using non-conformal coating of ceramic powders in conjunction with low-temperature sintering enabled by demonstrated doping technologies to avoid melting of the metallic interphases. R-curves intend to be used to quantify fracture response via macro and micro-scale mechanical testing. After establishing all necessary modeling, processing, and mechanical testing capabilities, a closed-loop design cycle looks to be executed utilizing the machine-learning accelerated model for microstructure 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 2025 · 2025-10

Many critical scientific challenges, from understanding complex diseases to designing innovative materials, rely on sophisticated computer simulations. However, scientists often encounter a "silicon ceiling," where current computational power restricts their ability to model these intricate real-world phenomena accurately enough to achieve major breakthroughs. The SINAPSE project directly addresses this issue by developing a powerful, open-source software toolkit that combines Artificial Intelligence (AI) with High-Performance Computing (HPC). This integration promises to enhance simulation capabilities, effectively offering significant orders-of-magnitude performance gains. SINAPSE will provide foundational software that benefits the broader AI-HPC research community, advancing the field itself. The project is also dedicated to supporting education and training for students in these cutting-edge computational methods, fostering the next generation of STEM professionals. By making advanced simulations more powerful and accessible, SINAPSE serves the national interest by driving innovation and enabling solutions to pressing scientific challenges. The project aims to overcome the "silicon ceiling" limiting complex simulations by developing a Scalable Infrastructure for AI-driven Predictive Simulation Enhancements (SINAPSE), delivering an open, sustainable Software Development Kit (SDK) that seamlessly couples Artificial Intelligence (AI) with High-Performance Computing (HPC) workflows. The project will provide functional capabilities through new and enhanced core software elements for AI-coupled HPC and integrated problem-solving frameworks for common scientific discovery patterns. The methodology begins by convening the SDK with a community focus. The SDK will then be populated by creating several novel core software elements and significantly enhancing existing tools like Colmena and RHAPSODY to support diverse AI-HPC coupling needs, including dynamic and asynchronous execution. These components will be assembled into problem-solving frameworks such as "Muse" for online surrogate model training, "Music" for model-directed sampling, and "Melody" for multi-scale campaigns. Finally, the entire SINAPSE SDK and its frameworks will be validated and strengthened through applications in biophysics, focusing on viral glycoprotein dynamics, and materials engineering, specifically for catalyst 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 2025 · 2025-10

This project addresses the growing challenge of data movement in high-performance computing systems with diverse processors, memory, storage, and networks. These systems are critical for national-scale efforts in drug discovery, materials science, energy research, and large-scale artificial intelligence and machine learning. Applications such as molecular dynamics, graph neural networks, and particle-in-cell simulations generate large data volumes that must be moved efficiently. Data transfers often limit performance rather than computation. This project develops tools to reduce data movement time and energy, improving throughput, efficiency, and scientific productivity. It empowers researchers and developers to scale workflows on complex HPC systems while fostering collaboration among academia, industry, and national laboratories to transition ideas into practical solutions for exascale platforms. The project also advances national interests by enabling scalable AI and simulation workflows and engaging students in systems research for next-generation infrastructure. The project develops a unified framework to reduce data movement overheads in heterogeneous high-performance computing systems. It integrates three core components: a cross-layer monitoring and learning framework that characterizes data transfer patterns and predicts contention; a heterogeneity-aware data movement scheduler that coordinates bandwidth usage across computation, memory, storage, and interconnect resources; and a collaborative caching and prefetching architecture that anticipates future data needs across workflows. The framework treats data movement as a first-class task, parallel to computation, and uses analytical and machine learning techniques to reduce interference and improve overall throughput. The research is validated through representative workloads on petascale and exascale systems, including simulations and machine learning pipelines. Results will provide generalizable strategies for optimizing data movement in next-generation scientific computing environments. 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 grant supports a workshop that will identify the challenges and risks involved in building AI-powered innovations in US ports, ultimately making the US more economically competitive. World trade depends heavily on the marine transportation system: Over 80 percent of world trade volume is carried by sea. Seaports handle a vast majority of international cargo and are critical links in global supply chains. With increasing world-wide demand for all kinds of goods, ports need to find new ways to meet the increasing volume. Automation has allowed modern ports to move goods more rapidly and safely, and to develop procedures that are designed to make them more resilient in the face of disruptions. Artificial Intelligence offers the opportunity to take the benefits of automation to the next level, to increase the efficiency and capacity of ports even further. While this greatly widens the scope of automation, it raises new challenges and potentially serious risks. The workshop will examine the opportunities for and the potential risks from the adoption of advanced AI for automation in ports. Modern ports have achieved greater efficiency and increased capacity through automation, through integrated and coordinated use of automated vehicles and cranes, drones, smart sensors, robots and robotic devices, more rapid communication and information sharing, and new logistics systems. However, they need to become more efficient, expand their capacity, and at the same time handle a wide variety of challenges such as tsunamis, power outages, accidents, and labor strikes. Traditional automation involves systems programmed for specific tasks without the ability to adapt to unforeseen situations, while AI-powered automation can learn and adapt to new information, and so it performs a much wider variety of tasks and makes decisions in new and increasingly complex environments, without human intervention. However, there are risks to interfacing AI with automation, including potential for new kinds of disruptions, lack of experience in dealing with them, and the potential for machines creating damage if given increased autonomy to make decisions. The workshop will explore such questions as: What is different about automation in the age of AI? What are some obstacles involved in implementation of AI-powered automation? What are some of the key optimization questions that arise in making ports “smarter”? What are the special risks of increased use of AI in automation in ports and what countermeasures might mitigate impact of new kinds of disruptions? The workshop will bring together academics and practitioners involved with AI and automation of ports and interfacing transportation modes (ships, rail, trucks, barges) and interfacing nearby warehouses, will initiate a dialogue among them, and identify key research questions they can collaborate on in the future. 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 COSMOS^3 project is to enhance and expand the COSMOS platform, a city-scale testbed for advanced wireless, networking, and sensing technologies, deployed in New York City (NYC). COSMOS is one of the projects under the NSF Platforms for Advanced Wireless Research (PAWR) program. The project aims to enable studying smarter and safer urban environments by deploying new wireless nodes and sensors—such as cameras, lidars, and radars—across key intersections in New York City. These tools will enable inventing and testing new technologies that could improve everyday life. The COSMOS^3 platform is designed to be open and accessible for a wide range of users to learn, experiment, and create technology that shapes the future of cities. The project is organized into several thrusts: (i) design and deployment of advanced hardware nodes that support a wider range of wireless frequencies (FR1/FR2/FR3) and multi-modal sensing; (ii) development of open-source software tools for managing experiments and processing real-time data; and (iii) development of educational materials and hands-on tutorials to support learning and experimentation. The COSMOS^3 infrastructure enables research on a wide range of next-generation wireless network scenarios including joint communication and sensing (JCAS), dynamic spectrum sharing, edge artificial intelligence (Edge AI), and smart intersections. By integrating programmable radios, multi-modal sensors, and high-performance edge computing, COSMOS^3 provides a flexible platform for experimentation and development that goes beyond what existing urban testbeds can offer, driving the advancement of next-generation wireless technologies. In terms of broader impacts, the COSMOS^3 testbed will accelerate wireless technology innovation, benefiting both the industry and end-users of mobile services. The project will also help to foster a growing community of research users and to encourage the use of the platform at local, regional, national, and global scales. The COSMOS^3 team hosts workshops, internships, and training programs to expand access to the testbed’s resources. By sharing all data, code, and training materials, the project is designed to accelerate discovery, foster collaboration, and enable new research and educational opportunities across many fields. Project information, software, datasets, and experimenter resources for COSMOS^3 are made available at https://cosmos-lab.org/. The project team is committed to maintaining the repository and website for the duration of the award and for at least two years beyond its completion. Long-term sustainability will be pursued through partnerships, community engagement, and integration with other research and education programs, ensuring COSMOS^3 remains a valuable resource into the future. 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 award provides funding to support the participation of early-career researchers, including graduate students, postdoctoral fellows, and recently appointed tenure-track assistant professors from U.S. institutions, in the six-month program on Geometric Spectral Theory and Applications, hosted and funded by the Isaac Newton Institute for Mathematical Sciences (INI), Cambridge, UK, from January to June 2026. The NSF funds will be used exclusively to support U.S.-based researchers attending one of the week-long workshops within the program. A significant portion of the funds is designated for early-career researchers (ECRs) participating in the Emerging Horizons in Geometric Spectral Theory: An ECRs Workshop, to be held February 2-6, 2026. Applications to attend other workshops, Geometry of Eigenvalues (March 23-27, 2026), Interactions of Geometric Spectral Theory with Numerical Methods and Applications (April 13-17, 2026), and Random and Arithmetic Models in Spectral Theory (May 11-15, 2026) will also be considered. Distributing funding across all workshops maximizes collaborative opportunities and offers greater flexibility, thereby enhancing participation by U.S. researchers. Participation in these activities is open to all who meet the scientific criteria. Applications will be submitted through the INI program website, where U.S.-based applicants will be offered the option to apply for NSF funding. The Principal Investigators of this grant will oversee the evaluation of NSF funding requests applying rigorous academic merit criteria. Geometric spectral theory is a vibrant and rapidly evolving area of mathematics with deep historical roots, from Ernst Chladni’s pioneering experiments on vibrating plates to Lord Rayleigh’s foundational work on acoustics and Mark Kac’s famous question, “Can one hear the shape of a drum?” The field maintains strong connections with differential geometry, mathematical physics, partial differential equations, number theory, dynamical systems, and numerical analysis. Applications range widely, from acoustics and scattering theory to computer imaging and scientific computing. The highly competitive INI program will focus on four interconnected themes: eigenvalues and geometry; geometry of eigenfunctions and spectral asymptotics; probabilistic and number-theoretic methods in geometric spectral theory; and numerical aspects and applications. These themes form the core of the workshops for which the funds of NSF award are used. The goal is to foster productive exchanges among these communities, leading to advances at their intersections, resolution of open problems, and the generation of new questions and applications. Participation in these workshops, embedded within a broader and vibrant six-month program, offers early-career U.S. mathematicians professional and academic benefits far beyond those of a single event. Given the concentration and breadth of expertise at INI, the workshops provide a unique opportunity for young researchers to present their work, engage with leading experts, and build lasting international collaborations that are critical to their future careers. 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

The central goal of geometry is to understand the structure of mathematical spaces, which may or may not be directly related to physical space. Such spaces can be described by both local properties, like curvature, and global properties, such as connectivity. Symplectic geometry, a specialized branch of this field, focuses on symplectic manifolds: spaces that are locally identical but can exhibit a wide range of global structures. These objects have originated in the study of motion and classical mechanics and play an important role in mathematical physics and applications. The primary tools used in their study fall into two broad categories: algebraic and analytic. Algebraic methods provide frameworks for encoding global information and facilitating computations, while analytic techniques involve solving differential equations and constructing these algebraic frameworks. This research project seeks to refine existing methods and develop new analytic tools to tackle longstanding challenges in the field. It will also contribute to outreach activities aimed at K–12 students through afterschool programs and support the training and professional development of graduate students via summer schools and seminars. At the technical level, this project focuses on several ambitious goals within symplectic geometry and mathematical physics. These goals all require refining existing techniques and developing new methods. First, it builds upon the method developed jointly by the PI and Bai to establish novel applications of Floer theory, including the longstanding Arnold–Givental conjecture. Second, the same framework will be used to investigate the relationship between integer-valued Gromov–Witten invariants and the Gopakumar–Vafa invariants of Calabi–Yau threefolds, a problem of interest to both geometers and physicists. Third, the project will continue the study of the framework known as the gauged linear sigma model, with particular emphasis on its role in classical and homological mirror symmetry. The PI will also set up afterschool outreach activities aimed at advancing the mathematical learning of K–12 students, as well as continue organizing summer schools and seminars to foster the training of Ph.D. students. 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 is using artificial intelligence (AI) to better understand how metals are used by cyanobacteria, the most abundant group of organisms to have ever existed in Earth history. Metals are essential for life and are held tightly by proteins inside of cyanobacterial cells. AlphaFold is a powerful AI tool that can predict the shapes of these proteins and how they hold onto metals. So far, these predictions have not yet been tested with laboratory experiments. In this study, college students are growing cyanobacteria in the laboratory and using advanced X-ray techniques to see how metals are bound inside the cells. By comparing the AI predictions with real data, the team hopes to better understand how metals move through environments when these cells die and break apart. This knowledge will provide a better picture of how dissolved metals are recycled in aquatic ecosystems. The project also includes outreach to schools and communities, including science activities for children, story-writing contests, and support for college students to get involved in science. The laboratory studies focus on resolving the chemical speciation of Zn and Fe in cyanobacteria. Experiments are performed by the researchers to assess whether AI-predicted metal-ligand binding environments reflect the actual speciation of Zn and Fe in living cells. Marine and freshwater cyanobacteria are being cultured under metal-controlled conditions, and proteins expressed under different growth phases are being identified using LC-MS/MS proteomics. The three-dimensional structures of the proteins will then be modeled using AlphaFold, and the protein structures will be annotated to identify putative metal-binding sites, coordination numbers, and ligand identities. These predictions will be experimentally tested using High Energy Resolution Fluorescence Detection (HERFD) X-ray absorption spectroscopy at the Zn and Fe K-edges, conducted at the Advanced Photon Source. Spectral data will be analyzed using linear combination fitting and principal component analysis to quantify the distribution of metals among cysteine, histidine, and carboxyl ligands. It is anticipated that AI predictions will correlate with experimental data, particularly in conserved protein families. These findings will provide mechanistic insights into metal-ligand complexation in cyanobacteria and establish a framework for AI-enabled investigations of metal cycling and biogeochemistry in natural aquatic 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 · 2025-09

Radar technology for monitoring vital signs, including heart rate and respiration rate, shows significant potential in transforming health monitoring via remote, continuous observation of multiple individuals. While phased arrays have been widely used in radar-based vital sign monitoring due to their low cost and effectiveness in single-person applications, they lack the flexibility to generate beams with well-controlled sidelobes required for monitoring individuals in close proximity. This project introduces VSMART (Vital Sign Monitoring via Remote Tracking), a system that integrates hardware, signal processing, and machine learning to enable continuous, remote monitoring of vital signs across multiple individuals. VSMART is designed to reduce interference and deliver high-accuracy performance at low cost. The research also explores the system’s adaptability for remote blood pressure monitoring. VSMART is particularly suited for deployment in crowded critical care environments, assisted living facilities, and nurseries, with applications including apnea detection and identification of sudden infant death syndrome in newborns. Additionally, the project contributes to curriculum development and offers opportunities for both graduate and undergraduate students to engage in research. The research pursues the following objectives: (i) Develop a novel, cost-effective radar transmitter architecture capable of producing well-controlled beams targeted at specific regions of the human body. This innovation centers on enhancing phased arrays with double phase shifters, providing additional degrees of freedom for improved beamforming capability. Unlike traditional arrays, the proposed configuration can control both magnitude and phase of the transmitted waveform at each antenna element, allowing for greater flexibility in beam pattern design. These improvements are achieved with minimal added cost, primarily due to the use of inexpensive phase shifters. For multi-target scenarios, the main beam is successively directed at each target while nullifying interference from others, enabling effective monitoring of multiple individuals. Computationally efficient algorithms for calculating beamforming weights that consider non-ideal characteristics of the phase shifters are developed and a prototype of the enhanced phased array system is constructed. (ii) Develop advanced methods for heart rate and respiration rate estimation using learning-based techniques that reconstruct waveform morphology similar to that of contact-based sensors. This includes the design of a waveform reconstruction network based on Long Short-Term Memory to capture temporal dependencies and a self-attention mechanism to emphasize relevant segments of the radar signal. A signal augmentation framework that simulates realistic radar echoes reflecting human dynamics is implemented. A Conditional Generative Adversarial Network serves as the foundation for generating training data to enhance model robustness. Radar waveforms that preserve the integrity of vital sign information within the radar echoes are also investigated. (iii) Leverage the flexibility of the proposed transmitter system to generate multiple beams focused on different regions of the body, and utilize the reconstructed vital sign waveforms to extract blood pulse information from each region. By analyzing time delays between these signals, estimates of blood pressure are obtained using both traditional physiological models and machine learning techniques. 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 explores a unique type of fungus that parasitizes aspen trees. Known as the aspen bracket, this fungus has a remarkable trait: insects tend to avoid it. This behavior suggests that the fungus — either on its own or in partnership with the aspen tree — may produce substances that are toxic or unappealing to insects. By studying how the aspen bracket functions, researchers hope to discover new insect-repelling or insect-killing compounds that are both effective and environmentally friendly. These natural substances may work in ways not previously seen in fungi, offering new approaches to pest control. The findings could have valuable applications in agriculture, forestry, and public health. In addition to its scientific goals, the project has a strong educational mission: it will involve undergraduate and high school students in hands-on research. By engaging students and sharing results with both the public and the scientific community, the project aims to inspire broader participation in science, technology, engineering, and math (STEM) fields. The metabolic profile of the aspen bracket (Phellinus tremulae) will be characterized with respect to the content and origin of insecticidal or repellent compounds. First, the project will identify and characterize metabolites in the fungus. Both untargeted and targeted metabolomics will be conducted using liquid chromatography–mass spectrometry (LC-MS), and compounds of interest will be purified and tested on insects to confirm their repellent or toxic activity. Second, the origin of these compounds in P. tremulae will be investigated. Comparative analyses will assess the presence of these compounds in wild-collected fungal specimens, host aspen tissues, and cultured fungal samples to determine whether the compounds are biosynthesized by the fungus or accumulated from the host. Lastly, the project will explore the mode of action of these compounds using computational modeling. A suite of in silico approaches will be applied, including molecular docking, molecular dynamics simulations, and free binding energy calculations, to evaluate interactions between the compounds and known insecticidal protein targets. Additionally, chemometric models such as quantitative structure-activity relationship (QSAR), read-across, and q-RASAR, combined with machine learning (ML) and artificial intelligence (AI) algorithms, will be used to predict insecticidal efficacy, potency, and chemical stability. Environmental and human safety assessments will also be conducted computationally to evaluate toxicity and regulatory viability. Together, these approaches may uncover new natural insecticides, reveal previously unknown biosynthetic pathways, and provide a model for integrating metabolomics, pest management, and predictive computational toxicology. The results of this project can be translated into new biotechology, i.e., new pesticides. 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.