University Of Florida

NIH Research Projects · FY 2025 · 2026-01

PROJECT SUMMARY/ABSTRACT Opioid use disorder (OUD) is a severely debilitating condition affecting nearly 10 million Americans. The sublingual drug product, Suboxone, which contains buprenorphine and naloxone, is the most frequently prescribed medication to treat OUD. In 2022, the FDA released a warning after several reports of patients prescribed sublingual forms of buprenorphine that exhibited severe dental caries and tooth loss. Furthermore, these issues arose in patients without prior dental concerns. Case reports and epidemiological evidence have supported these findings; yet the reason for this association is not understood. Dental caries (ie. tooth decay) is a chronic, polymicrobial, biofilm-based disease resulting from microbial dysbiosis and environmental changes in which microbial fermentation byproducts like lactic acid accumulate, demineralize the tooth structure, and can lead to severe odontogenic infections. Certain conditions favor the growth and persistence of aciduric and acidogenic microorganisms, including the cariogenic Streptococcus mutans and the yeast pathobiont Candida albicans at the expense of health-associated microorganisms, hence forming a cariogenic, dysbiotic microbiome. Sublingual tablets containing buprenorphine also contain several pharmaceutical excipients including binders, taste enhancers, and solubilizers, including lactose, mannitol, and citric acid. Our preliminary data suggests that these excipients combined promote the growth and biofilm formation of cariogenic microorganisms S. mutans and C. albicans. Importantly, they can do so at the estimated concentrations in the oral cavity following administration of the drug product. Thus, we hypothesize that the extended use of the current marketed drug products containing fermentable sugars and acidic excipients result in a shift in the oral microbiome by favoring cariogenic species and thus promoting the development of dental caries. In this proposal, we seek to address this concern by determining whether the current sublingual tablets containing buprenorphine promote a cariogenic microbiome and developing a new oral health promoting formulation that prevents the devasting dental effects associated with the currently marketed formulations. In Aim 1, the PI will first determine the growth and acidogenicity of S. mutans and C. albicans cultures exposed to each excipient and whole tablet. Then biofilms formed by these two microorganisms in response to the tablet will be examined by performing biofilm biomass assays and confocal microscopy. Metagenomic analysis will then be performed using an ex vivo dental plaque microcosm model after exposure to the tablets. In Aim 2, the PI will formulate a novel tablet formulation that replaces identified cariogenic excipients with validated non-cariogenic and dental safe excipients. Successful completion of the proposed studies will result in an improved formulation aiding in the prevention of dental caries, thus significantly improving the health and quality of life of OUD patients, and patient's willingness to accept buprenorphine therapy. Furthermore, the proposed training plan will cultivate the PI's development into a clinician-scientist through coursework, inter-disciplinary mentorship, dental practice, and rigorous research.

NSF Awards · FY 2026 · 2026-01

The ability to detect pathogens, toxins, and biological/chemical threat agents is critical to supporting public health, and national security. However, such detection is limited by instrument sensitivity and portability, slow analysis methods, and the inability to detect unknown hazardous substances. For example, to protect public health after extreme weather events, toxins and harmful bacteria in waterways must be monitored over vast distances, yet water quality samples are typically collected in the field and brought to a laboratory for analysis using time-consuming methods and non-portable instrumentation. One potential way to overcome these limitations is to use living cells as biosensors in versatile, portable, and sensitive devices. Decades of research have provided the ability to manipulate the genetic circuitry of living cells, programming them like computers. Unfortunately, biosensing devices made from such cells are not currently as reliable as everyday digital electronics. Taking inspiration from digital technology now relied upon worldwide, this project will develop biosensors that harness the fundamental physical principles underlying the functionality of modern electronic devices. The project will broaden participation in STEM fields by developing a program to guide team research, mentoring, and training activities that tap into the full spectrum of available talent. The research team will hold biannual workshops to learn new ways to expand the range of thoughts, ideas, impacts, and approaches for defining and solving important questions in research. Partnering with a local museum, an exhibit for continuous use in public education and outreach will be created, assembling high-school and undergraduate research teams to build the exhibit. This TRAILBLAZER project aims to develop biotechnology inspired by solid-state physics while leveraging the powerful complexity of genetics. Coherent, spatiotemporally patterned, and highly controlled collective biological states will emerge upon integrating programmed cell communication with cell positioning on lattices. This structure-function hypothesis is drawn from solid-state materials; semiconductors, magnets, and superconductors are made by arranging specific atoms into specific crystal structures, but the atoms alone do not have the properties of their assemblies. Thus, as in atomic crystals, emergent collective states can be designed by matching cell programming to cell assembly. Creating new biological states by integrating a top-down structural strategy with bottom-up cell-engineering has never been attempted; crystallinity has not yet been leveraged to control collective cell behavior. Thus, the main research thrusts are: (1) develop new types of engineered cells and synthetic biology programming specifically for incorporation into crystal lattices; (2) bring together theoretical modeling, advanced bio-fabrication, and experimentation to create and study the novel biological phases that emerge in analogy to the collective phases that emerge in solid state materials. The project will introduce new ways to think of biotechnology development, which will be used to create totally new classes of biosensors and other cell-based devices. Anticipated Transformative Impact: Disruptive options for 3-D biomanufacturing beyond bioprinting. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2026-01

PROJECT SUMMARY/ABSTRACT This application seeks to understand the epidemiological patterns and complex interplay of risk factors driving skin cancer incidence and severity among people with HIV (PWH) and to develop predictive tools that address disparities in diagnosis and prognosis. PWH face a higher likelihood of developing non-AIDS-defining skin cancers (NADSCs), including but not limited to basal cell carcinoma, squamous cell carcinoma, and melanoma, due to multifactorial biopsychosocial determinants of health. Despite this risk, the mechanisms underlying NADSC incidence and severity in PWH remain under-researched, and no tailored predictive models exist to guide targeted interventions for this vulnerable population. This proposal describes the application of epidemiological and machine learning approaches, including causal artificial intelligence (AI) methods, specifically probabilistic graphical models (PGMs), with two objectives: 1) characterizing patterns of NADSC burden among PWH and 2) identifying direct effectors among risk factors contributing to NADSC in PWH and developing robust predictive models accordingly. By leveraging multimodal healthcare data containing sociodemographic and clinical data, I aim to address critical gaps in the identification of high-risk subgroups and the development of targeted detection and prevention strategies. The central hypothesis of this proposal is that NADSC incidence and severity in PWH are influenced by a complex interplay of causal factors that can be modeled to inform personalized clinical interventions. In Aim 1, I will characterize the epidemiological patterns of NADSC burden (prevalence, disease sub-types, severity) among PWH compared to people without HIV. In Aim 2, I will identify factors that directly affect NADSC incidence and severity in PWH. Specifically, I will integrate high-dimensional data sources to uncover interactions between risk factors such as immunosuppression, comorbidities, medication history, and sociodemographic identifiers. I will then build and validate predictive models for NADSC incidence and severity among PWH, focusing on the integration of multimodal data to develop robust and clinically actionable models. These models will be validated using an independent dataset to ensure generalizability. This work is significant because it addresses a critical gap in understanding and predicting NADSC incidence and severity in PWH, a population with unique sociomedical vulnerabilities. The project is innovative in its use of causal AI and multimodal data to elucidate novel relationships and create predictive tools tailored to the needs of PWH. These findings will advance the field of onco-dermatology and contribute to reducing disparities in skin cancer outcomes through data-driven clinical and public health strategies.

NIH Research Projects · FY 2025 · 2026-01

Project Summary and Abstract Inborn errors of metabolism (IEMs) are a diverse group of over 1,000 congenital disorders caused by enzymatic mutation, the vast majority of which prominently affect the nervous system. Phosphoglycerate dehydrogenase (PHGDH) deficiency is a rare IEM caused by loss-of-function mutations in PHGDH, the rate- limiting enzyme in the de novo serine biosynthesis pathway. PHGDH deficiency presents with a clinically heterogeneous spectrum ranging from lethal neonatal Neu-Laxova Syndrome to nonlethal infantile-onset manifestations. While it is understood that PHGDH loss leads to serine and glycine depletion, the mechanisms linking enzyme dysfunction to nervous system phenotypes remain poorly defined. This project investigates how PHGDH loss disrupts metabolic homeostasis during nervous system development by disentangling its dual effects on serine-driven one-carbon, and central carbon metabolism. Preliminary data using [U-13C]glucose tracing in patient-derived PHGDH-deficient cells reveal not only reduced serine synthesis, but also increased pyruvate synthesis and TCA cycle turnover, potentially exacerbating oxidative stress through reactive oxygen species (ROS) overproduction. We hypothesize PHGDH deficiency not only inhibits serine/glycine synthesis, and also increases TCA cycle turnover, predisposing cells to oxidative damage while simultaneously driving global dysregulation of energy metabolism. In Aim 1, we will use inducible PHGDH expression systems in patient-derived fibroblast cell lines and [13C] tracers to define how varying PHGDH levels alter metabolic flux and ROS generation, and test interventions that buffer oxidative stress. In Aim 2, we will characterize the developmental consequences of PHGDH loss in newly developed humanized PHGDH deficiency mouse models. We will leverage spatial metabolic imaging and advanced microscopy to characterize the metabolic functions of PHGDH during nervous system development. Lastly, we will test exogenous, gestational serine supplementation strategies to assess tissue-specific metabolic vulnerability and embryonic lethality rescue in development. This work will provide fundamental insights into how PHGDH loss causes metabolic pathway disruptions leading to nervous system developmental defects. Our work will establish a framework for applying in vitro and in vivo targeted metabolic tracer and gene regulation approaches to study and potentially treat PHGDH deficiency and other IEMs. This proposal comprises of a rigorous and reproducible research strategy and comprehensive training plan that will prepare me to be an independent principal investigator at a tier 1 research institute. This research project will incorporate training in mass spectrometry, isotope tracer analysis, gene regulation, and molecular genetics. The University of Florida, in combination with the exceptional joint mentorship of Dr. Matthew Merritt and Dr. Eric Wang, provides an outstanding research environment that will enable my professional growth and completion of this proposed research.

NIH Research Projects · FY 2025 · 2026-01

Title: Cryptococcal Nanotherapy: Engineering a Novel Fungus-Based Platform for Improved Drug Delivery to the Central Nervous System PROJECT SUMMARY/ABSTRACT Neurodegenerative diseases affect more than 1.2 billion people worldwide and this figure is steadily increasing. The exact causes of neurodegenerative diseases are still not entirely understood and the FDA-approved options for such diseases, especially amyotrophic lateral sclerosis (ALS), have very minimal effects on disease progression. This limited efficacy can also be attributed to the physiological barriers of the central nervous system (CNS), the physicochemical properties of the drugs, and the lack of understanding of how the drugs impact ALS. Riluzole and Edaravone, two FDA-approved drugs for ALS, possess the ideal properties for penetrating the CNS. However, their mechanisms of action in ALS remain unclear, they are likely to be absorbed by off-target tissues and can be ejected from the CNS barriers after ALS onset. Resveratrol has been cited as a potential alternative to existing therapeutics for ALS after demonstrating neuroprotective properties in preclinical animal models of ALS and is currently being investigated in clinical trials. Resveratrol is a polyphenolic molecule that scavenges reactive oxygen species and reduces oxidative stress in neuronal mitochondria, two hallmarks of ALS. Like Riluzole and Edaravone, resveratrol experiences off-target tissue absorption. The drug delivery field has sought to address this challenge by encapsulating resveratrol with nanomaterials. Despite some success, the nanoscale formulations of resveratrol are still limited by physiological barriers and the body's clearance mechanisms. Therefore, there is a critical need for a drug delivery platform that can withstand the body's clearance mechanisms and bypass the physiological barriers of the central nervous system to enhance ALS treatment. This work proposes using the neurotropic fungus Cryptococcus neoformans (Cn) to facilitate more efficient delivery of resveratrol into the central nervous system to treat ALS. Cn is well-equipped to traverse the physiological barriers of the central nervous system by “professionally” executing three separate mechanisms to penetrate the tissue: (1) paracellular transport (2) transcellular transport (3) immune cell hitchhiking and subsequent vomocytosis. Vomocytosis is a process by which Cn liberates itself from the intracellular environment of phagocytes to promote its survival. Given these unique characteristics, Cn is an ideal candidate for engineering the desired drug delivery platform as it can bypass physiological barriers to penetrate the CNS and escape premature degradation by the immune system. The overall hypothesis is that by tethering nanoparticle- loaded resveratrol onto the surface of Cn we can improve resveratrol delivery to the CNS and modulate ALS disease progression more effectively. To address this hypothesis, the first aim will optimize the synthesis of fungal drug carriers (FDCs) and assess their functionality in vitro. The second aim will evaluate the biodistribution of FDCs, quantify the bioavailability of resveratrol in the CNS, and investigate the therapeutic efficacy of the FDCs in a preclinical ALS model. A more efficient drug delivery platform that more effectively traffics resveratrol to the CNS will not only enhance the ALS treatment but also enhance treatment for other neurodegenerative diseases.

NSF Awards · FY 2026 · 2026-01

Butterflies are one of the most familiar and charismatic groups of insects. They are important in natural ecosystems as herbivores, pollinators, and components of food webs. In temperate regions, like North America, butterflies are well-studied and often used as indicators for how biodiversity changes across space and time, and as flagship species for conservation. In the tropics, however, where butterfly diversity is far greater, there is an urgent need for basic information about species diversity, distribution, natural history, and trends in abundance over time. This project will focus on the butterflies of protected areas in Ecuador, which is one of the world's three most diverse countries for butterflies. The project will document butterflies using observations from nature along with DNA, to help us better understand butterfly diversity and insect communities in the tropics. The project will provide educational opportunities for undergraduate and graduate students through independent research projects, courses in insect biodiversity, and training experiences in Ecuador. Live-stream events, a museum exhibit and school visits through the Scientist in Every Florida School program will help communicate the project to school children and the public. The project will initiate a multi-year survey of butterflies within national and local protected areas in Ecuador. This will build unique spatial, temporal, DNA sequence and functional trait datasets, and fill major gaps in knowledge of the distribution and diversity of Ecuador’s c. 4500 butterfly species, almost a quarter of the world's fauna. The project will conduct regular butterfly monitoring events throughout the year at a dozen sites in different biogeographic regions, environments, and elevations, where park rangers will receive equipment and training in field methods and science communication. The project team will also partner with park rangers to undertake expeditions to collect and study butterflies in remote, poorly explored mountain ranges and forests. Specimen collections and DNA 'barcodes' will facilitate systematic research and descriptions of new butterfly species. Functional trait data will be used to examine how environmental filtering and competition influence seasonal variation in butterflies, and how evolutionary and ecological processes influence spatial patterns of diversity. The project will train university students in the USA, reach school children and the public through museum programs, collaborate with scientists and naturalists in Ecuador, and build interest in biodiversity. 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

Membrane technologies have the potential to play a transformative role in addressing energy scarcity, which impacts the lives of millions of people. Atomically-thin two-dimensional (2D) materials represent a new kind of membrane material. 2D materials allow subatomic particles (e.g., protons) to selectively pass through the membrane while blocking even small gas atoms such as helium. The ability to separate protons from other atoms and molecules will enable disruptive innovations in energy generation and conversion, chemical processing and separations, electronics, and environmental protection. The project aims to develop fundamental understanding of proton transport through 2D materials. These scientific insights will be leveraged to develop novel catalytic and separation processes that serve to advance the U.S. economy and national security. A comprehensive education and outreach plan will complement and aid research efforts by a) reinforcing positive public perception towards science, engineering and mathematics and b) training the next-generation of scientists. Atomically-thin 2D materials such as graphene and hexagonal boron nitride offer fundamentally new opportunities to probe and control mass-transport. Pristine monolayer graphene and hexagonal boron nitride are impermeable to helium atoms but allow for proton transport. Selective proton transport through 2D materials offers transformative opportunities for fuel cells, isotope separations, hydrogen purification, photo-detectors, and artificial photosynthesis. However, a comprehensive understanding of proton transport mechanisms through 2D materials remains elusive. The overall objective of project is to develop fundamental understanding of the mechanisms governing proton transport through 2D materials. State-of-the-art advances in in-situ metrology will be used to study proton permeation through 2D materials. These fundamental insights on proton transport will be used to develop novel catalytic and separation processes that are of interest to the U.S. economy and national security. The research is integrated with a comprehensive education and outreach plan that focuses on i) providing under-represented and under-served groups with research internships for undergraduate and high-school students and engaging with their high-school teachers; ii) collaboration with professionals to develop content for outreach and dissemination of research findings via social media platforms; and iii) community engagement with hands-on science experiments via outreach activities at Vanderbilt University and the Nashville area. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2026-01

Project Summary Osteoarthritis (OA) is a chronic, progressive condition that affects over 32 million adults in the United States. Despite its widespread impact, preclinical research has struggled to develop effective treatments. A major obstacle is the knowledge and technology gap in OA research, where preclinical studies focus on understanding disease mechanisms, while clinical treatments primarily target pain relief. Although recent preclinical research has started to include pain assessments, these methods are still underdeveloped compared to those used in human patients. Historically, OA was thought to be caused by cartilage degradation due to wear and tear. However, it is now understood to be a more complex disease involving inflammatory and metabolic factors. Recent studies show that OA pain does not always correlate with joint damage, and pain severity is not necessarily linked to the degree of structural damage. This disconnect highlights the need for a unified approach that addresses both pain and disease progression. While clinicians focus on pain management, preclinical studies tend to emphasize cellular mechanisms and structural changes. Both perspectives are essential for developing comprehensive treatment strategies. To bridge this gap, researchers have begun incorporating pain as a key metric in preclinical studies, especially in animal models; however, measuring pain in animals is challenging. Human pain is commonly assessed through standardized questionnaires such as KOOS and WOMAC, while pain in animal models is inferred from behavioral changes, such as alterations in movement or weight-bearing. OA manifests in a variety of phenotypes in humans, making it difficult to replicate in animal models. The pain associated with OA is multifaceted, influenced by biological, psychological, and social factors. Current preclinical models utilize both stimulus-evoked and non-stimulus-evoked methods to assess pain in rodents. Non-stimulus- evoked methods, such as gait analysis and weight-bearing assessments, are gaining popularity due to their closer alignment with clinical pain experiences. Despite significant physiological differences between rodents and humans, such as differences in gait and anatomy, rodent models remain crucial for studying OA pain. However, to enhance the relevance of these models to human conditions, pain assessment techniques in rodents must be further developed. Therefore, the objective of this proposal is to identify and validate translatable pain metrics that can be used in both preclinical models and human patients. We propose to assess pain-like behaviors in rodents and humans using various methodologies, including gait analysis, quantitative sensory testing, and activity monitoring. By identifying reliable, translatable pain measures, we aim to improve the translational relevance of preclinical research and ultimately enhance clinical outcomes for OA patients.

NSF Awards · FY 2026 · 2026-01

Configuration troubleshooting in large-scale cyberinfrastructure (CI) software systems is a complex and costly task due to numerous configurable parameters. Existing methods like log mining and machine learning analysis face challenges in such environments. To address this, we present BGT4AutoCI (Automating CI Configuration Troubleshooting with Bayesian Group Testing), a groundbreaking solution that leverages Bayesian Group Testing, ensuring accurate results even with minimal prior knowledge and testing errors. Experienced CI operators can expedite the process with approximated prior knowledge. This research aims to revolutionize CI configuration troubleshooting, introducing a highly precise and efficient approach that will optimize the utilization of current and future large-scale CI systems. The primary focus of this research is to address critical challenges in automated configuration troubleshooting within large-scale CI systems. The proposed three-fold approach encompasses: (1) Formulating Bayesian Group Testing for CI configuration troubleshooting, which employs lattice models to accurately identify risks at the individual configuration parameter level, taking uncertainty into account during troubleshooting. (2) A multinomial paradigm for Bayesian Group Testing, which introduces multinomial responses to simultaneously observe multiple aspects of CI systems, enabling efficient test selection algorithms for jointly testing configuration parameters that impact various aspects of CIs. (3) Automated configuration troubleshooting, which involves the designs of several key components to establish BGT4AutoCI as an automated configuration troubleshooting framework that minimizes the need for human intervention. The outcomes of this project hold the potential to significantly enhance the efficiency and accuracy of CI configuration troubleshooting, benefiting current and future large-scale CI 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 2026 · 2026-01

High-Performance Computing (HPC) has revolutionized various scientific fields, including climate research, wildlife health, agricultural sciences, and scientific simulations and modeling. With the emergence of HPC-accelerated deep learning (HPC-DL) systems and applications, there is a pressing need for comprehensive cross-layer training materials to educate the research workforce on these advanced technologies. The primary objective of this pilot project is to address this need by providing comprehensive cross-layer HPC-DL training to a wide range of cyberinfrastructure (CI) users. The target audience includes undergraduate and graduate students, postdocs, faculty, and research staff who can benefit from enhanced knowledge and skills in utilizing HPC-DL CI technologies and resources. By equipping them with the necessary training, the project aims to improve their research efficiency and maximize the potential of HPC-DL in their respective fields. In addition, the project has a specific focus on fostering inclusivity and expanding opportunities for underrepresented communities in the Central Valley area of California. This will contribute to the national interest by empowering individuals with the knowledge and skills necessary to excel in the HPC-DL field. This project addresses the critical training needs of the converged HPC-DL field by developing comprehensive training materials, fostering peer consultant programs, conducting workshops, and building an inclusive learning culture. It includes an integration of scientific applications, HPC technologies, and DL in a cross-layer approach. The training program covers several important CI topics, including Remote Direct Memory Access (RDMA), GPU-based distributed computing, Slurm, MPI, and NCCL, which are critical to achieving high performance for HPC-DL workloads. The training will also dive into distributed DL training frameworks such as PyTorch, TensorFlow, and Horovod, enabling participants to effectively leverage these tools for their research. Moreover, the training incorporates practical DL application case studies, offering real-world examples and insights. The short-term goal is to empower individuals with HPC-DL knowledge and cross-layer optimization skills to maximize the utilization of HPC-DL CI resources and improve research efficiency. This project will also examine the effectiveness of practice-central models and HPC-DL-centered workshops in promoting HPC-DL adoption in underrepresented communities. The project's long-term aim is to cultivate a robust research workforce with a deep understanding of HPC-DL CIs. By establishing a learning culture and targeting a significant number of CI users, this project addresses workforce shortages and extends its impact beyond the Central Valley. Through collaborations and the dissemination of open-source training materials, it will contribute to advancing compute- and data-intensive scientific simulations and knowledge discovery. 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

Efficient data movement is a critical challenge in high-performance computing (HPC) and artificial intelligence (AI) cyberinfrastructures due to the massive volumes of data generated by modern data-intensive applications. Existing methods often struggle with performance bottlenecks, particularly when transferring data across parallel and distributed computing environments. To address these limitations, this project -- the Open DPU-Offloading data Transfer Architecture (OpenDOTA) -- provides a framework that leverages Data Processing Units (DPUs) to accelerate data movement. By enhancing efficiency in DPU-powered systems, OpenDOTA aims to advance scientific simulations, drive AI advancements, and strengthen computational research infrastructure. The project fosters collaboration and contributes to the evolution of state-of-the-art data movement technologies, benefiting a wide range of users in academia and industry. This project focuses on designing OpenDOTA as a high-performance, scalable framework for DPU-offloaded data movement in HPC and AI cyberinfrastructures. The research is structured around three key thrusts: (1) Adaptive point-to-point data movement, which employs diverse offloading strategies to optimize data transfer over DPUs; (2) Accelerating collective communication by leveraging advanced DPU offloading techniques to improve scalability; and (3) Deep reinforcement learning (DRL)-based optimization, which dynamically adapts data movement strategies for maximum performance. By integrating these approaches, OpenDOTA offers a comprehensive solution to existing data movement challenges, paving the way for scalable, high-performance applications across HPC and AI domains. 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

In an era where High-Performance Computing (HPC) cyberinfrastructure systems are undergoing rapid and transformative evolution, marked by a surge in heterogeneity and scale, this project stands at the forefront of transformative innovation in HPC cyberinfrastructure, where rapid development in heterogeneity and scale presents both opportunities and challenges. Its significance lies in its unwavering commitment to advancing the usability, efficiency, and scalability of HPC systems and applications. By pioneering Heterogeneity-Enriched Communication designs and software, this project not only propels the field forward but also resonates with the broader mission of the National Science Foundation (NSF), which aspires to promote the progress of science, advance national welfare, and secure the national defense. Beyond its scientific impact, this endeavor prioritizes education and diversity, fostering a learning culture and addressing workforce shortages in critical STEM (Science, Technology, Engineering, and Mathematics) fields. It seeks to democratize knowledge and empower the scientific community through open-source principles, amplifying its potential benefits for data-intensive scientific research. The primary objective of this project is to pioneer innovative heterogeneous-enriched communication designs for modern and next-generation HPC systems and applications, significantly enhancing their usability, efficiency, and scalability in the face of rapid evolution characterized by increased heterogeneity and scale. This will be achieved through a multi-pronged approach involving analytical modeling and architectural performance optimization. This research project targets three key challenges: (a) achieving high usability by developing precise cost modeling, prediction, and simulation for heterogeneity-enriched communication; (b) addressing efficiency through the composition of innovative communication schemes and adaptive orchestration; and (c) ensuring scalability through optimization techniques that reduce message and connection overhead. These designs, released as open-source tools and integrated into existing HPC libraries, can revolutionize HPC and Machine Learning (ML) applications across domains such as agriculture, industrial workloads, biostatistics, and geospatial information science. Furthermore, this project aligns with the NSF's core mission to advance the national welfare by actively fostering inclusivity, education, and underrepresented groups in STEM. It aims to drive innovation and leave a lasting impact on the scientific community. 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

PART 1: NON-TECHNICAL SUMMARY Living cells constantly communicate and respond to their surroundings using complex molecular interactions on their surfaces. Many life-saving immunotherapies, such as engineered T cells used to fight cancer, rely on these surface interactions to activate the immune system. However, building artificial systems that can reliably mimic these natural signals has been very difficult. Current synthetic materials cannot easily control how proteins cluster, move, and organize on a surface, even though these features are essential for proper immune activation. This project supports fundamental research that creates new ways to design protein-based building blocks that self-assemble into soft, cell membrane-like vesicles. These vesicles behave like simplified versions of natural cell membranes, allowing researchers to program how surface proteins move and interact. By learning how to control the spacing, mobility, and density of these proteins, the research lays a foundation for next-generation artificial antigen-presenting cells that may one day improve cancer immunotherapy and other biomedical technologies. The broader impacts of this work include developing hands-on research experiences for undergraduate and high-school students. The project will integrate laboratory modules into courses, provide mentoring opportunities, and create outreach activities that introduce young learners to biomolecular engineering and synthetic biology. Together, these efforts support the national interest by advancing scientific understanding, improving human health, and strengthening the future STEM workforce. PART 2: TECHNICAL SUMMARY This project seeks to establish fundamental design principles governing the self-assembly, membrane organization, and immunological function of recombinant protein vesicles that mimic essential features of natural antigen-presenting cells (APCs). The research uses modular fusion proteins composed of three domains: a folded globular protein for functional display, a heterodimerizing ZE/ZR coiled-coil pair that specifies molecular stoichiometry, and an elastin-like polypeptide (ELP) segment that provides amphiphilicity, membrane formation, and tunable mechanical properties. These components self-assemble into immunomodulatory protein vesicles (iPVs) whose deformability, lateral mobility, and protein valency can be precisely modulated through sequence design and controlled mixing ratios. The central objective is to determine how protein architecture, stoichiometric loading, membrane mechanics, and ligand mobility collectively regulate mesoscale spatial patterning and receptor engagement at synthetic cell-mimetic interfaces. Quantitative biophysical methods will define how these parameters influence membrane fluidity, protein clustering, and dynamic reorganization during contact with T cells. A major focus is understanding how iPVs present peptide–MHC complexes, costimulatory antibodies, and cytokines to cytotoxic CD8+ T cells. The platform enables systematic interrogation of early signaling events such as receptor activation thresholds, microcluster formation, immunological synapse stabilization, and downstream proliferative responses. Overall, this research will define foundational principles for engineering synthetic biomolecular membranes capable of coordinating immune recognition. The resulting framework will guide the construction of next-generation artificial antigen-presenting cells built entirely from recombinant proteins, enabling scalable, tunable, and safe immunomodulatory materials. The principles discovered will also support broader efforts in synthetic cell design, biomaterials development, and bottom-up synthetic biology. 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

Strengthening wildfire resilience requires accurate modeling and a deep understanding of collective human behavior during wildfire evacuations. In particular, there is a critical need for simulation models that can realistically capture how civilians, incident commanders, and public safety officials make protective action decisions during wildfires. However, existing simulation models face fundamental limitations that often cause low prediction accuracy and insufficient capacity to support effective decision-making during wildfire response. Therefore, this project aims to develop a convergent framework for next-generation wildfire evacuation simulation that features realistic Artificial Intelligence (AI) agents powered by psychological theory-informed large language models (LLMs), reinforcement learning, and multi-modal datasets. This research seeks to be a transformative step toward improving the behavioral realism, prediction accuracy, and decision-support capability of wildfire evacuation simulation models. This project intends to lead to generalizable simulation methods, promote teaching, training, and learning, strengthen partnerships, and support wildfire resilience through broad dissemination and open-access tools. Despite progress in wildfire evacuation simulation models, key challenges remain in accurately modeling and understanding the decision-making processes by incident commanders and public safety officials, realistically modeling human behavior in wildfire evacuations, and adequately representing diverse populations and their dynamic, complex interactions. LLM-based agents could help address some of these limitations, though they bring their own issues with hallucination and generalizability. To tackle the above research challenges, this project looks to develop a novel convergent framework for learning-based simulation of collective human behavior during wildfires. Specifically, the objectives of this research are to: 1) extend and enrich the Protective Action Decision Model (PADM) for civilians as well as incident commanders and public safety officials; 2) develop psychological theory-informed LLM agents for protective action modeling; 3) generate a realistic, context-aware synthetic population to serve as the critical input for the simulation platform; 4) develop the learning-based simulation platform to integrate complex interactions among various agents and predict collective human behavior at the community level under various scenarios (e.g., fire spread, warning, infrastructure damage); and 5) test and validate the convergent simulation framework with various case studies across the U.S. The research outcomes intend to enable wildland-urban interface (WUI) communities to better predict wildfire impacts, manage risks, and develop life-saving strategies. 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

Integrated Sensing and Communication (ISAC) is a transformative technology that unifies sensing and communication functions within a single system, significantly enhancing efficiency, cost-effectiveness, and performance. By sharing key resources such as spectrum, power, and hardware, ISAC not only reduces infrastructure costs but also improves spectrum utilization, minimizes interference, and alleviates congestion in increasingly crowded wireless environments. This integration enables real-time environmental awareness and faster decision-making, which are essential for applications such as autonomous vehicles, smart cities, Internet of Things networks, and industrial automation. Despite its promise, ISAC systems face major challenges due to the dynamic and complex nature of the wireless medium, particularly in multi-user scenarios, and the limitations of Radio-Frequency (RF) circuitry, which impact both sensing accuracy and communication reliability. This project introduces a novel ISAC system, Integrated Sensing and Telecommunications for Intelligent Connection and Transmission (INSTINCT), to address key challenges in joint communication and sensing. It advances multi-dimensional signal processing (MSP) across the delay, Doppler, and wavenumber domains, while ensuring compatibility with standard wireless protocols. A central innovation is the use of wavenumber-delay-Doppler domain signal processing, where range and velocity information naturally reside. INSTINCT further enables continuous-aperture phased multiple-input multiple-output (CAP)-MIMO, a reconfigurable sub-aperture architecture that enables dynamic, simultaneous communication and sensing capabilities. Key research contributions of this project include: (1) Developing reconfigurable RF hardware that seamlessly integrates communication and sensing via spatially adaptive apertures; (2) Creating electromagnetic-informed channel models and optimal signaling strategies tailored for CAP-MIMO systems; (3) Designing multi-dimensional waveforms and analyzing the impact of synchronization errors, RF impairments, and multi-user interference in the wavenumber-delay-Doppler domain; (4) Developing domain-informed progressive neural network architectures for joint beamforming and phase shift design in communication and radar sensing using reconfigurable hardware; (5) Designing a practical dynamic spectrum sharing framework using learning-based spectrum activity sensing for INSTINCT. 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

Barrier island breaches have occurred during many tropical storms, constituting a major mechanism for tidal inlet formation, dune and beach erosion and development. Thus, they represent a major challenge to coastal management. The current understanding of the fate, physical processes, and impacts surrounding new and evolving breaches is limited due to the lack of comprehensive longitudinal studies capturing the breaching event and post-breaching evolution on monthly and annual time scales in a holistic and transdisciplinary manner. To address the current gaps in knowledge and data, this EArly-concept Grant for Experimental Research (EAGER) study investigates the development of two barrier island breaches from their original formation over multiple seasons and years, and their potential impacts on coastal management and infrastructure systems. The "high-risk high-outcome" study is expected to reveal new insights into the roles of hydrodynamics, land coverage, and geomechanical sediment properties on barrier island breach evolution, as well as into the impacts of these newly formed inlets on coastal infrastructure systems. It looks to unravel the importance of barrier island breach data collection for informed coastal management, planning, engineering design, and decision-making in coastal regions affected by storms. The data are expected to become a benchmark data set that will serve the wider coastal research community for calibration and validation of numerical and physical models and the development of new concepts, relationships, and theories regarding the geomorphological evolution of storm-induced barrier island breaches, local hydrodynamics, surrounding sediment and land-use conditions, coastal infrastructure, and the built environment. Midnight Pass breach in Venice, Florida, and Milton Pass breach in Englewood, Florida, opened during the 2024 sequence of Hurricanes Helene and Milton and are located in the same geological and meteorological region. The two inlets will be investigated with focus on post-breach geomorphodynamics driven by small-scale variability in hydrodynamics, sediment properties, geomorphology, vegetation, and anthropogenic influences from engineering actions and land use. The study seeks to leverage and extend the interdisciplinary field data collections following the storms and in 2025, complementing the effort with analyses and initial application to existing numerical models. The project intends to also test and assess newly emerging instrumentation and cross-disciplinary data collection strategies for storm-related geomorphodynamics and infrastructure system performance research. The study seeks to build on and strengthens an interdisciplinary network of natural hazards sciences and engineering researchers. 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

High-resolution geoscience data are essential for understanding and predicting extreme weather events, yet producing such data remains a major challenge due to limitations in observational infrastructure and computational cost. This project introduces a transformative AI-based framework to overcome these barriers by generating high-fidelity, physically consistent, and uncertainty-calibrated geoscience data. These enhanced datasets will empower better decision-making in disaster preparedness, emergency response, and infrastructure planning. The project’s broader societal impacts include advancing tools for tropical cyclone prediction and wildfire detection, training a new generation of interdisciplinary scientists in AI and geosciences, and releasing open-source software for broad accessibility. By integrating explainable AI with physical principles and expert knowledge, the research aims to improve public trust in scientific models and provide actionable insights for meteorologists, policymakers, and emergency managers. Outreach efforts and mentoring initiatives will promote participation in STEM and foster the development of future leaders in climate resilience and AI for natural hazards. This project develops a next-generation generative downscaling framework that combines diffusion-based generative models with physical constraints, domain expertise, and probabilistic uncertainty quantification. Key innovations include physics-guided loss functions to enforce geophysical realism, text-prompted guidance from expert knowledge for tailored downscaling, and conformal prediction techniques to provide rigorous confidence intervals for model outputs. The approach will be validated using diverse, high-impact datasets such as IMERG, CMORPH, and radar observations, with specific applications focused on improving precipitation estimation in tropical cyclones and enhancing wildfire risk detection. By unifying advances in machine learning, statistical modeling, and atmospheric science, the project will establish a new foundation for trustworthy AI-driven downscaling and support critical scientific and societal needs in a changing climate. 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

Learning-enabled autonomous systems operating in unfamiliar or unprecedented environments pose new foundational challenges for their safety assessment and subsequent risk management. In this context, the system-level safety means the complicated behaviors created by the interactions between multiple learning components and the physical world satisfy the safety requirements, protecting the system from accidental failures to avoid hazards such as collisions to other vehicles, bicycles and pedestrians. The qualitative and quantitative methodologies envisioned to complement each other by providing both 'yes' or 'no' binary decisions and numerical measures of safety, which allow for a thorough understanding of safety concerns and enable effective safety verification in uncertain environments. This project targets the foundational challenges of developing qualitative and quantitative safety assessment methods capable of capturing uncertainties from environments and providing timely, comprehensive, and accurate safety evaluations at the system level. The outcomes are expected to boost the trustworthiness and adaptability of learning-enabled systems to the unknown world and facilitate their safe integration into various domains, such as autonomous vehicles, robotics, or industrial automation. Educational and outreach activities are well-integrated into the research, including curriculum development, K-12 STEM outreach, and industrial engagement activities. The designed activities are uniquely positioned to promote diversity throughout this project by giving priority consideration, mentoring, and working with students in underrepresented minority groups. The proposed research efforts will be directed toward building the foundations of end-to-end qualitative and quantitative safety assessment of learning-enabled autonomous systems. This project will develop the probabilistic star temporal logic specification language. The new specification language offers a formalism for expressive modeling of learning process uncertainty and complex temporal behaviors, and supports both qualitative and quantitative reasoning. Efficient computation methods and tools will be developed to verify probabilistic star temporal logic specifications for learning-enabled deep neural network components. The verification methods and tools are centered on enhancing their scalability and resource efficiency. This project will develop system-level qualitative and quantitative safety assessment methods and tools that can handle the interplay of various learning-enabled components in a system under different availability of environment information. Learning-enabled F1Tenth testbed, a small-scale system of real autonomous vehicles and its simulator, will be used to create multiple real-world autonomous driving scenarios to validate and evaluate the applicability, scalability and reliability of the proposed methods and tools. This research is supported by a partnership between the National Science Foundation and Open Philanthropy. 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

Machine learning (ML) is projected to be essential in future autonomy development. However, ML components such as deep neural networks (DNNs) may react unexpectedly to even tiny input variations. Recent incidents in ML-powered systems, such as Tesla and Uber autonomous vehicles, raise an urgent need for techniques and tools to formally verify the safety and robustness of DNNs before utilizing them in safety-critical applications. The state-of-the-art research, focusing on the safety, robustness, and fairness of deep neural networks and neural network control systems, is powerful and promising for some parts of ML-powered autonomy development. However, enabling trustworthy, complex, learning-enabled autonomy is still challenging due to the need for verification technologies for system-level reactive behaviors involving complex interactions between multiple components. This project's novelty is in creating new formal method foundations in modeling, specification, verification, and toolchains to address this grand challenge research problem beyond the state-of-the-art. The project's impact is the substantial enhancements of safety, reliability, and explainability of various learning-enabled unmanned systems. Additionally, the project will strengthen the research and study in autonomy-focused topics in Nebraska by recruiting undergraduate research assistants and integrating research findings into ML and autonomy verification courses. It will also increase the interest and engagement of K-12 group students in STEM majors and science literacy through outreach events. The project team will also collaborate with the University of Nebraska-Lincoln Osher Lifelong Learning Institute to give lectures and discussions for adults on how autonomy concepts and technologies may impact their lives. This project involves two foundational research thrusts, modeling and specification and quantitative verification, along with software and trustworthy autonomy testbed development and rigorous evaluation. The project's expected research outcomes include 1) a new generic graph-based modeling approach for complex, learning-enabled autonomy (CLeA) in which CLeA's components and their interaction are represented using nodes and edges, respectively, 2) a new set-based algebra, built upon the concepts of probilistic star (or shortly ProbStar, a new variant of the well-known star set) and containing a collection of mathematical propositions and important operators such as parallel composition, decomposition, Minkowski sum, etc., that allow users to discover and keep track of the dependency between multiple reachable sets produced by different components in a CLeA and compose/decompose precise inputs for these components in the analysis, 3) a new set-based specification language to specify CLeA's temporal behaviors based on the concepts of ProbStar set representations, 4) a suite of algebra-based depth first search (DFS) reachability algorithms to construct the reachable set traces of all components in CLeA over multiple steps. and 5) a suite of scalable quantitative verification algorithms to quantify the satisfaction of CLeA's temporal properties under uncertainties. Results from proposed research thrusts will be integrated into StarV, a new quantitative verification tool, to enhance its verification capacity for various applications. Indoor and outdoor learning-based autonomous driving testbeds using the F1Tenth platform and Robify robot will be developed to evaluate the proposed verification framework. CARLA simulator, SCENIC, and hardware-in-the-loop (HIL) simulation will also be used to assess the proposed research in various project phases. 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 objective of this research project is to establish and operate a research network for enhancing airport resilience. It brings together experts from academia, industry, and government to develop a novel framework for analyzing airport resilience to natural hazards and operational disruptions. The network aims to quantify the resilience of airport operations, synthesize best practices for recovery, and explore the integration of emerging technologies to simulate and optimize airport responses to external shocks. Through an interdisciplinary approach, the network intends to addresses complex resilience challenges in the aviation sector, enabling data-driven decision-making processes, anticipating evolving threats, thereby advancing both the science and practice of infrastructure resilience. Airports are highly complex systems with the primary function of maintaining flight schedules while ensuring safety, sustainability, and economic viability. The research network looks to produce conceptual models informed by resilience science to help airport managers identify critical functions across interconnected sub-systems, quantify the impacts of disruptions, evaluate resilience against both known and emerging threats, and optimize resource allocation for performance improvements. This goal intends to be achieved through three key objectives: 1) evaluating how resilience indicators translate into airport operational impacts and vary across different types of disruption; 2) identifying challenges related to data access, management, privacy, security, and interoperability while exploring pathways to improve data integration; and 3) proposing new paradigms for infrastructure resilience research by applying resilience science concepts to airport operations. Using Dallas Fort Worth International Airport as testbed, the project seeks to address significant gaps in resilience measurement and evaluation and generate robust and generalizable knowledge for transportation systems. Most importantly, it looks to forge a self-sustaining, collaborative platform for expert engineers, scientists, and practitioners to share insights that will improve airport resilience strategies, inform broader efforts in standardizing aviation resilience metrics, and set the foundation for long-term competitiveness of US transportation sector. 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 wireless research community continues to face major challenges in conducting rigorous, repeatable experiments to evaluate next-generation wireless networks and Internet of Things (IoT) systems. Existing testbeds are limited in availability and often fixed to specific environments, making it difficult for researchers to test how their innovations perform under different conditions. This project addresses these challenges by creating UnionLabs, a new cloud-based federation of wireless testbeds that aims to democratize access to experimental research resources. By enabling seamless remote access to testbeds distributed across multiple U.S. institutions, UnionLabs promotes wider participation in wireless research. It also helps accelerate research in areas such as Artificial Intelligence and Machine Learning (AI/ML) for autonomous systems, mobile edge computing, and spectrum sharing. Through integration into university courses, hands-on training opportunities, and public workshops, UnionLabs supports both workforce development and broader engagement. The project establishes an innovative two-tier infrastructure that combines a centralized public cloud platform with edge computing resources located at individual testbed sites. A federation plane hosted on Amazon Web Services (AWS) enables seamless integration and remote access to geographically distributed experimental facilities through a unified web-based interface. To validate and demonstrate the scalability of this platform, UnionLabs will federate testbeds across four institutions with complementary strengths including University at Buffalo (5G and UAV systems), University of Florida and University of Utah (IoT technologies), and Northeastern University (programmable 5G and beyond systems). The platform’s standardized federation APIs will support easy onboarding of additional grassroots testbeds over time, laying the foundation for a dynamic and sustainable national research infrastructure. 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

Non-technical abstract: This project will accelerate the discovery of new superconducting materials through a transformative approach that combines artificial intelligence (AI), quantum theory, and experimental synthesis. Superconductors are essential for technologies ranging from MRI systems and high-field magnets to quantum computing and sustainable energy. Yet analysis of known compounds suggests that only a small fraction of potential superconductors may have been discovered. This project aims to significantly expand the number of known superconductors and identify materials optimized for practical applications—specifically those with high critical temperatures and magnetic fields, ductility for wire fabrication, and three-dimensional electronic structures for enhanced performance. A core educational mission will train a group of students in AI-driven materials research and develop hands-on experiment kits for K–12 classrooms to promote STEM engagement. Partnerships with national laboratories, industry, and international collaborators will ensure timely and impactful transition of discoveries to real-world applications. Technical abstract: The research integrates two complementary AI methods to accelerate superconductor discovery. Property prediction models based on graph neural networks will estimate superconducting characteristics—such as electron-phonon spectral functions, critical temperatures, and critical fields—directly from crystal structures. In parallel, generative AI models using stochastic flow matching will design novel, synthesizable materials with targeted superconducting and mechanical properties. Predictions from both models will be evaluated using density functional theory (DFT) to assess thermodynamic stability, electronic structure, and superconducting potential. Selected candidates will undergo targeted synthesis and experimental characterization using high-throughput techniques. Both AI models will be iteratively refined using experimental feedback, forming a closed discovery loop that integrates theory, simulation, and validation. This approach will yield a comprehensive, open-access dataset of successful and unsuccessful candidates, providing a foundation for future AI-guided materials discovery. 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

When land is cleared for human activities, any remaining habitat is often found in isolated patches surrounded by very distinct land cover. This process, commonly referred to as ‘habitat fragmentation’, is hypothesized to be a major threat to the integrity of ecosystems worldwide. Studies to date have documented a wide variety of biological and environmental changes in habitat fragments, which is why many have concluded that the ecological consequences of habitat fragmentation will be severe, detrimental, and long-lasting. Whether the effects of habitat fragmentation on biodiversity are indeed detrimental, however, has emerged as one of the most contentious contemporary debates in ecology. At the core of this debate is that while habitat loss and fragmentation often occur simultaneously, they are actually distinct processes with potentially unique impacts on biodiversity. This means that detrimental effects attributed to fragmentation could instead be due to habitat loss. To address this contentious topic, the research team will synthesize multiple data sets collected over the course of two decades on the biology of the Amazonian understory plant Heliconia acuminata. In addition to generating a novel understanding of the relative impacts of habitat loss and fragmentation on biodiversity, it will also motivate future population-focused research on how these impacts might differ between functional groups. In doing so the researchers will also be generating publicly available data sets and computer code for use by the research community and peer-reviewed educational materials on the effects of tropical deforestation for use in undergraduate STEM courses. To isolate the effects of habitat loss and fragmentation on population viability, the researchers will model H. acuminata population dynamics in simulated landscapes that vary in the total amount of habitat and the size and arrangement of habitat fragments. The researchers will determine if there is a relationship between spatiotemporal variation in environmental conditions and demographic vital rates; the nature of this relationship will inform the type of Integral Projection Model used to simulate population dynamics in patches connected by seed-dispersing birds. These spatially-explicit demographic models will be used to determine (1) if there is a threshold amount of habitat below which patch configuration fails to increase landscape-level viability, (2) if there is a threshold amount of habitat above which landscape-level viability is assured regardless of how habitat patches are configured, and (3) how habitat fragmentation mediates the role of habitat configuration in landscapes with the same amount of total habitat. 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

Deep learning-based perception and control are increasingly popular in recent autonomous systems. Unfortunately, deep learning is vulnerable to various changes in the visual environment, known as visual shifts, which threaten the system’s performance and safety. This project aims to ensure the safety and performance of vision-based autonomous systems subject to visual shifts, including sun glares and seasonal changes such as snow-covered terrain. We will build specialized modeling, training, and adaptation techniques to overcome visual shifts based on the core idea that visual uncertainty should be treated by balancing informativeness with conservatism. This balance can be achieved by focusing on the most surprising visual phenomena to make safe choices in unforeseen circumstances. The intellectual merit of this project includes developing theories and algorithms at the intersection of formal verification and information theory that will endow vision-based autonomous systems with high-performance behaviors and previously unavailable theoretical guarantees. The broader impacts of this project include making vision-based autonomy safer and more reliable, particularly in the automotive sector, as well as transitioning insights gained from the project to practice by collaborating with researchers in the auto industry, publicly releasing our data and code, and providing university students with hands-on experience with safe perception and control. This project will develop an end-to-end methodology that leverages information-theoretic and statistical techniques in modeling, analysis, training, control, and adaptation. At the foundation of the proposed methodology is a robust framework for probabilistic verification and control synthesis, which will provide conservative models and safety estimates under latent shifts. Building on these models is a neuro-symbolic training process that bridges the gap between visual perception, safety, and control. Finally, to protect the system from overreacting to unseen shifts, this project will develop an online adaptation framework based on quick change detection and perception/control switching. The proposed methodology will be validated on a variety of autonomous systems from different domains, including physical experiments of small-scale autonomous racing. This research is expected to provide tight complementary connections between information-theoretic learning and formal techniques for safe control. 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

Metal additive manufacturing (AM) such as laser powder-bed fusion (LPBF) has been increasingly explored not only for product innovation, but also shop-floor production, demonstrated by growing success from a variety of industries. However, the lack of knowledge in both fatigue failure and the performance uncertainty of LPBF parts poses a significant challenge and undermines the potential of deploying LPBF for high-consequence applications. This Faculty Early Career Development (CAREER) award supports fundamental research to understand the effects of LPBF processing on defects and subsequent fatigue behavior, advance the knowledge of fatigue scattering of LPBF parts that are complex in geometry and subject to multiaxial loading. The effort will establish a physics-centric, machine learning framework for fatigue life predictions, serving as a technological foundation for future metal AM production of dynamic load-bearing applications, and thus, enhance the competitiveness of U.S. industry. This CAREER project will also integrate education and outreach programs designed to broaden the participation from underrepresented groups through actively engaging K-12 students for STEM education and recruiting women and minorities into research, priming future generations of diverse engineers with the knowledge and skills indispensable in the age of manufacturing innovation and big data. The ultimate goal of this early career effort is to understand fatigue failures of complex LPBF parts under multiaxial loading for data-driven fatigue life predictions. The research will investigate the nature of fatigue failures from plastic deformation and crack initiation at the highest stress concentrations and translate fatigue life predictions into evaluating the crack growth at the vulnerable zones using a multiscale approach. On the micro-scale, critical defects with crack-initiating features (by x-ray computed tomography or optical profilometry) will be identified based on the correlation with fatigue failures; both the effects of critical defects and their spatial interactions on crack growth will be examined using fracture mechanics and data-intense statistics. On the part scale, the weak regions of the highest stress concentrations will be examined by finite element modeling of stress and strain behaviors through decoupling multiaxial loading. The effects of critical defects and the principal stresses at vulnerable localities will then be incorporated into a hierarchical graph convolutional network of deep learning to model their synergistic impacts on crack growth and calculate the fatigue life of LPBF parts with advanced data analytics. The findings are expected to generate new knowledge of defect formation relevant to fatigue performance of LPBF parts, uncover the synergistic impacts of multiscale factors on fatigue fractures, and further LPBF adoption for high-consequence 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.