Carnegie-Mellon University

NSF Awards · FY 2025 · 2025-08

This project aims to make progress on several areas at the interface of probability, geometry, and combinatorics. Three specific directions will be investigated that range from more theoretical to more applied. The project also involves a research experience for undergraduates. The first direction of the project is that of enabling statistical inference in the realm of probability distributions characterized by geometric constraints. For example, can we efficiently choose a uniformly random partition of a region into geometrically nice pieces? When can we efficiently choose a uniformly random partition of a graph into, say, connected subgraphs? The second directions concerns the minimum lengths of combinatorial structures like spanning trees and Hamilton cycles among "cities" in geometric space. Here, some aims push our knowledge in directions that are more closely motivated by applications, while others probe the limits of our current theoretical approaches. Finally, the third direction aims to explore competitive/online analogs of Radon's theorem for high-dimensional point-sets. This direction places results in geometry and classical machine learning in a new context, while offering new paradigms for inquiry. One example is the notion of a pseudo-randomized trial, in which controlling curse-of-dimensionality effects can replace randomization in rigorous comparison between groups under a linear effects model. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

The broader impacts of this Pathways to Enable Open-Source Ecosystems (POSE) project will be to improve understanding of the human mind by broadening access to a unifying computational framework for modeling human cognition. This framework will then serve as a foundation for cumulative scientific progress across cognitive science, neuroscience, artificial intelligence (AI), and the social sciences. By creating this platform, high-fidelity computational cognitive models of human behavior will become available to a widening group of engineers, programmers, and human factors practitioners. Integrating this technology into current technology infrastructures could positively impact many fields of human activity. Using cognitive models to facilitate human-machine teaming would provide a competitive advantage by enabling a faster, more productive integration of AI tools into the workforce. Unified theories of cognition could extend national leadership in AI by providing the scientific basis for the next generation of robots and intelligent agents. Everyone in society could benefit in the form of smarter systems that work more naturally with humans, resulting in increased satisfaction and productivity. This Pathways to Enable Open-Source Ecosystems (POSE) project aims to develop an open-source ecosystem (OSE) to sustain the development and expansion of the Adaptive Control of Thought-Rational (ACT-R) cognitive architecture. Computational cognitive architectures implementing unified theories of cognition have become increasingly mature and accurate in capturing human behavior. ACT-R is currently recognized as the most successful cognitive architecture; however, its current model for funding, developing, and supporting the user community fundamentally limits its reach. The primary objective of the project is to explore how to expand the user base from cognitive science researchers into a self-sustaining ecosystem comprised of people from all fields interested in human and artificial cognition, including systems neuroscience, AI, robotics, social sciences, and systems design. The ecosystem discovery and scoping activities include one-on-one interviews and a series of interactive workshops with interested scientists and engineers to gauge their interest, understand their needs, and assess relevant risks. The information gathered will be used to develop a new model of organization and governance for the OSE that would support broadening the scientific basis of the ACT-R cognitive architecture and expand its user base beyond its current 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.

NIH Research Projects · FY 2026 · 2025-08

PROJECT SUMMARY Matrix-encapsulated communities of bacteria, called biofilms, are ubiquitous in the environment and are notoriously difficult to eliminate in clinical and industrial settings. Existing within biofilms offers resident cells protection from threats, such as bacteriophage attacks, phagocytosis by the host immune system, and antibiotic treatment. However, the fundamental principles by which bacteria gauge environmental threats to inform biofilm development remain unclear. Recently, my laboratory discovered a novel, threat- agnostic mechanism, by which bacteria sense endangerment and respond by forming protective biofilms. The underlying mechanism is a process we call “lysis sensing,” whereby surviving cells sense a signal released during the death of related species and, in turn, initiate biofilm formation. Going forward, the overarching goal of our research program is to develop a comprehensive understanding of lysis sensing across scales, from molecular mechanisms to population-level dynamics. To realize this vision, we will leverage a multidisciplinary approach, involving molecular microbiology, biochemistry, microscopy, computation, and automation. To reveal how lysis-sensing receptors connect threat assessment to the formation of protective biofilms, we will biochemically characterize the Vibrio cholerae lysis-sensing receptor-signal complex that we recently uncovered. To determine how lysis signaling functions in conditions that approximate microbiomes, we will combine experimental and theoretical approaches to examine lysis sensing in homogeneous, spatially structured, and multi-species communities. To assess the pervasiveness and diversity of lysis sensing, we will identify lysis-sensing pathways in additional clinically relevant bacteria using high-throughput imaging approaches that we have pioneered, and we will follow up to uncover underlying molecular mechanisms. Finally, we will extend our studies beyond regulation of the biofilm lifecycle to explore other defense mechanisms activated by bacteria in response to lysis signals. Together, this work will shed light on fundamental concepts by which bacteria undergo threat assessment to respond to the ongoing challenges faced in their host and environmental niches. Our research could uncover new strategies for manipulating microbial behaviors in health and disease, with potential long-term applications for biocontrol, therapeutics, and microbial ecology.

NIH Research Projects · FY 2025 · 2025-08

Abstract Intracellular delivery of macromolecules into cells outside the body is an essential step in a wide range of processes across biology, biotechnology, and medicine. Developing more flexible, scalable, and effective delivery technologies is the central goal of Dr. Sevenler’s research group. Dr. Sevenler’s research is particularly focused on understanding how the permeability of the plasma membrane to macromolecules can be momentarily increased by applying a brief pulse of stretching force to the membrane, which is also called mechanoporation. Although they and others have shown that mechanoporation can be highly effective for delivering molecules, the specific mechanisms by which stretching leads to increased permeability are poorly understood. For example, the number, size, distribution, lifetime, growth rate, and biomolecular composition of mechanopores have yet to be resolved in living cells. These questions are clinically important, because macromolecule delivery is a critical step in the production of many cell and gene therapies, and current delivery strategies are nonuniform, damaging to cells, and/or not scalable to clinical production volumes. Over the next five years, Dr. Sevenler will quantitatively measure when, where, and how mechanopores can be safely created in the membranes of human cells outside the body and use this information to develop improved delivery strategies. To accomplish this, his group will apply a range of experimental and numerical techniques in microfluidics, viscoelastic fluid mechanics, biomaterials, and optical nano-imaging. Specifically, Dr. Sevenler’s research program will focus on two interconnected areas of investigation to advance our understanding and applications of mechanoporation: (1) quantitative characterization of mechanoporation; and (2) elucidation of membrane permeability and pore structure. To investigate mechanoporation mechanics (1), Dr. Sevenler plans to develop and apply novel imaging and microfluidic technologies to quantify cell deformation and membrane stress across relevant timescales. His group is particularly interested in pulsed holographic imaging techniques and computational modeling of cell mechanics in viscoelastic flows. To elucidate membrane permeability and pore structure over time (2), his group will design and develop innovative analytical methods such as interferometric scattering microscopy and inertial microfluidics-based permeability assays. Progress in these areas will contribute to developing safer, more effective, and scalable intracellular delivery methods by collecting quantitative data about membrane disruption, permeability, and repair in living cells. Our vision is to leverage this information towards a long-term goal of high throughput cell “surgery,” wherein controlled amounts of macromolecules can be gently and efficiently delivered into specific cellular compartments for therapeutic applications.

NSF Awards · FY 2025 · 2025-08

Protons and neutrons, which make up atomic nuclei, are comprised of fundamental particles known as quarks and gluons, whose interactions are described by quantum chromodynamics (QCD). QCD and the electroweak force should describe all nuclear processes, thereby explaining the inner workings of stars and the abundance of hydrogen versus helium and other elements in the universe. QCD also suggests that many other types of composite particles, known as hadrons, should exist in Nature as unstable resonances. Extracting such information from QCD is difficult, requiring large-scale computer simulations. QCD predictions of the structure of the proton and neutron, scattering involving one particular hadron known as a Delta baryon, and other heavier hadronic resonances will be investigated using supercomputing resources. The proposed research lends support to current experiments, such as the GlueX experiment in Hall D at the Thomas Jefferson National Accelerator Facility (JLab). The scattering information involving the Delta baryon will be crucial for current and future experiments studying neutrinos, an important elementary particle that permeates the universe. The PI will mentor a graduate student engaged in this research, and the graduate student will also receive training in the use of state-of-the-art parallel computing resources. The physics of hadron-hadron interactions will be studied by formulating QCD on a space-time lattice so that computer calculations of a variety of QCD correlation functions involving quark and gluon fields can be carried out. Nucleon-pion, nucleon-nucleon, and nucleon-hyperon scattering phase shifts will be investigated, yielding important information on hadron structure. Studies of three-meson systems will be continued to better understand certain nuclear reactions and meson resonances, such as the omega meson, which have significant three-body decay branching ratios. Scattering processes involving nucleon-pion-pion states will be considered in order to understand heavier baryon resonances, such as the Roper resonance. Form factors involving the nucleon and transitions through the Delta baryon will be a particular focus since they are crucial to accelerator-based neutrino experiments, such as the Deep Underground Neutrino Experiment (DUNE). Past work used a computational technique known as the stochastic Laplacian Heaviside (LapH) method, which made possible such computations in lattice QCD for the first time. Recent development of new software exploiting GPU accelerations has opened up the possibility of making exact LapH studies feasible, allowing unprecedented statistical precision of the results obtained. 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 2026 · 2025-08

SUMMARY Substance use disorders, including opioid use disorder, are rising across the United States have devastating consequences on those affected as well as their family and friends. There is growing evidence that addiction and sleep are tightly linked at the genetic, molecular, and behavioral levels. At the genetic level, there have been hundreds of loci confidently linked with addiction phenotypes and sleep phenotypes. Genetic variants associated with both sleep and addiction show enrichment in distal open chromatin regions of many of the same striatal neuron subtypes. Despite these associations, it remains unexplored whether genetic variants that impact the predisposition to addiction or sleep phenotypes also directly impact the circadian regulation of gene expression. The difficulty in understanding the gene regulatory mechanisms underlying candidate addiction- and sleep-associated genetic variants results from an inability to perform controlled experiments in an intact human brain. Here, we leverage a combination of machine learning and genomic technologies to make inferences about how genetic variants impact the circadian regulation of gene expression by altering the combinatorial code of transcription factors at distal enhancer regions. In Aim 1, we conduct controlled single cell genomic experiments in the mouse striatum, which serve as training data for convolutional neural network models. These models then make inferences across the entire human genome about which genetic variants are likely to modulate circadian gene regulation. In Aim 2, we synthesize hundreds of candidate human addiction- and sleep- associated enhancers and then measure their circadian regulation in in vivo massively parallel reporter assays conducted in the mouse striatum. In Aim 3, we apply comparative genomics to find enhancers that are likely to be conserved in function between human and mouse. The result of disrupting those enhancers in a CRISPR-KRAB mouse strain is measured using a variety of behavioral assays for addiction and sleep phenotypes. The result of these aims will be predictions and measurements of the circadian function across hundreds of enhancers that contain addiction- and sleep-associated genetic variants. These experiments will allow us to test the hypothesis that addiction-associated genetic variants impact predisposition by influencing circadian gene regulation. More broadly, we will provide a framework to understand the genetic mechanisms of addiction and sleep across other brain regions and tissues as well as in other disease contexts.

NIH Research Projects · FY 2026 · 2025-08

Virtually all research on brain communication at the level of individual neurons and small populations of neurons has focused on inter-area communication within a hemisphere, While multi-area communication is undoubtedly important for understanding neurological disorders and increasingly studied, such a perspective ignores the more dramatic issue that brain processing occurs across two hemispheres, The mechanisms by which information about the world is stored, processed, and acted upon by both hemispheres are critical to much of higher brain function, Assessing the fundamental principles of interhemispheric coordination has been challenging, However, technical advances in multi-contact electrode technology and conceptual advances in dynamic attractor modeling have allowed us to make significant progress in disentangling competing hypotheses regarding how the prefrontal cortex organizes and represents memory and movements, These advances suggest that prefrontal cortex is more modular and less distributed than previously thought such that each hemisphere has a complete representation of visual space, including both left and right hemifields, The apparent redundancy provides an important avenue for work focused on neurological rehabilitation and restoration, We propose experiments that will capitalize on recent advances and our collective expertise, Our first specific aim will investigate the maintenance of spatial working memory across the hemifields in populations of neurons in prefrontal cortex, Our second specific aim will investigate how memory is transformed into a plan to move the eyes, focusing specifically on how interhemispheric spatial memory influences a single action, The final specific aim is to model spatial working memory networks across the two hemispheres, using the attractor network framework to simulate alternative architectures for brain function, Collectively, these experiments will provide new insights into the mechanisms that underlie critical brain processing circuits that are central to a host of everyday behaviors and neurological disorders,

NSF Awards · FY 2025 · 2025-08

Non-technical summary Crystal growth is an essential process for developing new materials used in technologies ranging from electronics to renewable energy. However, growing high-quality crystals reliably remains a challenge because small changes in temperature, chemical composition, or other conditions can drastically affect the outcome. With this project, supported by NSF’s Office of Strategic Initiatives in the Directorate for Mathematical and Physical Sciences, Prof. Chamorro and his research group improve the control and predictability of a widely used crystal growth technique by developing new tools to monitor and adjust the process in real time. At Carnegie Mellon University they develop a new approach to track crystal growth as it happens, using specialized in-situ monitoring techniques and tools. These tools help identify the best conditions for growing crystals with specific properties. In addition, the team develops artificial intelligence models that can analyze data from crystal growth experiments and suggest ways to optimize the process, potentially leading to future systems that can adjust growth conditions automatically. By improving the precision of crystal growth, this research advances the discovery and production of materials with unique properties, including those useful for next generation computing and energy technologies. This project also plays a pivotal role in training the next generation of scientists and engineers in materials science, chemistry, and, more broadly, data science. By integrating contemporary machine learning techniques with traditional materials science methods, students gain valuable interdisciplinary skills that are increasingly important in today’s technological landscape and job market. Furthermore, the project involves hands-on workshops and interactive demonstrations for K-12 students in the greater Pittsburgh area. The workshops highlight the exciting potential of quantum materials and crystal growth with activities designed to be accessible and engaging, as well as showcasing the real-world applications of STEM and its impact on everyday life. Technical summary This project, supported by NSF’s Office of Strategic Initiatives in the Directorate for Mathematical and Physical Sciences, refines the flux growth technique for synthesizing high-quality single crystals by incorporating real-time, in-situ monitoring and adaptive control. Custom experimental apparatuses are developed to track key growth parameters, providing new insights into the crystallization process of heavy fermion quaternary perovskite compounds (ACu3Ru4O12). Additionally, machine learning models are implemented to analyze growth data and optimize conditions dynamically. By improving control and reproducibility, this work advances the fundamental understanding of crystal growth mechanisms and enables the targeted synthesis of materials with tailored properties. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This Faculty Early Career Development (CAREER) award supports research on the emergent mechanical properties of beadwork as a new class of programmable metamaterials. Beadwork is not only visually captivating but also mechanically fascinating. As a metamaterial, its mechanical properties arise from not just the materials used but also its structure. Fabricated by weaving thread through beads, beadwork combines complementary characteristics of textiles and granular matter into one material. For example, like textiles, beadwork is programmable to various shapes and like granular matter beadwork can rearrange to resist high loads. This work will explore the relationship between beadwork design and its mechanical properties. Insights from this work will not only promote the progress of science but also enable the use of beadwork as material in applications that advance the national health, prosperity, and welfare. For example, new technologies that require materials with varied properties such as soft robotics, wearable tech, and deployable structures. This CAREER award also supports integrated educational activities designed to broaden resources, perspectives, and participation at the intersection of STEAM disciplines of mechanical engineering, physics, and craft. Working towards developing a pipeline for the next generation of STEAM researchers and workforce, these activities include a graduate course on craft mechanics, online STEM-beadwork resources, instructor training, and K-12 STEM programming using beadwork. The award will directly drive these activities as beadwork is implemented in learning modules, hands-on demonstrations, and research opportunities. As a composite material system, beadwork combines mechanics of slender structures and packed grains, enabling them with programmatic design, tensile actuation, and extremely tunable stiffness. This CAREER project will support experiments on beadwork metamaterials to explore these and other emergent properties and develop predictive models that describe beadwork’s key mechanical and geometric features. In particular, the objective is to investigate the beadwork’s mechanical response to internal and external action, expected to be characterized by nonlinear deformation, internal contact and friction, fracture, and instability. In a gradual research approach, simpler, periodic two-dimensional designs to more complex three-dimensional configurations will be considered. Bead level material behavior, kinematics, and forces will be theorized and connected to effective structural response. Homogenization schemes based on lumped masses/energy will be used in developing approximate but efficient predictive models. The desktop-scale experimental platform for testing concepts in this project will be useful in related fields that encode tessellations. For example, beadwork shares concepts of topology and geometry that are used in the computer graphics community for developing meshes as well as in nanoscale chemistry with carbon nanostructures. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Imaging inside the human body can reveal important information for diagnosis and treatment of disease. To obtain a high-quality image, light is usually projected from outside the body using external lenses and microscopes. The imaging depth is limited since light cannot be delivered deep into the tissue from outside. Lens and endoscopes can be inserted into the tissue to mitigate this issue. However, this is invasive and can cause damage. This project will use non-invasive ultrasound waves to guide and focus light through the tissue without the need for inserting a physical lens. This approach will allow for high-resolution imaging in parts of the body, like the brain, that cannot be achieved with MRI, while imaging over a larger volume than possible with a traditional microscope. The project will also help develop the next generation of young scientists and engineers with an interest in ultrasound and optical technologies. This project will develop the first ultrasonically-enhanced swept source optical coherence tomography (ueSS-OCT) system for label-free optical imaging of tissue structure and function with near beam waist limited lateral resolution over an extended depth range. Ultrasound waves can locally change the refractive index profile in tissue to sculpt in situ virtual optical waveguides that can focus and steer light. The project has three main objectives: (1) to utilize the unique interplay between these in situ lenses and SS-OCT detection for beam waist limited resolution over 2 mm of sub-surface structural imaging depth and (2) up to 2 mm of functional phase-resolved imaging with shot noise limited phase resolution with an extended flow velocity range. These advances will be validated by (3) detection of functional activation in the well-established animal model of somatosensory activation in response to whisker stimulation. This project will demonstrate a novel imaging paradigm in which external focusing optics can be replaced by in situ optical waveguides formed within the tissue itself. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Vibrations are everywhere - often in ways too subtle or too fast for human eyes or regular cameras to detect. Floors tremble as people walk, engines hum with motion, air conditioners send ripples of vibrations through the walls, and even human voices or the beat of hearts cause tiny vibrations. These vibrations, though nearly imperceptible, carry a wealth of information about the physical world. By studying these vibrations, it is possible to learn surprising things: how full an opaque liquid container is, whether a machine is malfunctioning, or even where someone is walking in another room. Vibrations can reveal an object’s physical traits - like its weight, stiffness, or internal structure - and can even help systems understand what’s happening in places that are not directly in the line of sight. Today, however, capturing this kind of information is extremely difficult. Most vibration sensing relies on expensive, specialized equipment that can only measure one point at a time. And even if it were possible to gather all that data, making sense of it - especially in real time - is a significant challenge. This project develops technologies for sensing and understanding vibrations. The potential applications of the research are vast and impactful. Firefighters could locate people trapped in buildings. Engineers could detect structural weaknesses in aircraft bodies. First responders could remotely monitor vital signs in disaster zones. Safety inspectors can detect leakage in industrial containers holding hazardous materials from a safe distance. Beyond the lab, the project also focuses on education and outreach. Through hands-on experiences and creative programs, the team is making this cutting-edge science accessible to students and the broader public - helping everyone see (and feel) the invisible world of vibrations. This research project aims to significantly improve sensing and interpretation of vibrations by bringing together three ideas: (i) The research team is developing low-cost “visual-vibration” cameras that use everyday components - like consumer-grade cameras and lasers - to detect vibrations across space and time in complex everyday environments. (ii) By combining physics-based models with modern deep learning models, the researchers are creating algorithms that can detect hidden flaws in objects, estimate how much liquid is inside a sealed bottle, or even track activity in a room that’s out of sight. (iii) The system must not only rely on existing scene vibrations but can also send out controlled vibrations to “probe” the environment, helping it to improve the accuracy and efficiency of the extracted information. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Machine learning (ML) systems have achieved impressive capabilities across domains such as vision, language, and planning. However, even state-of-the-art models can fail dramatically in unpredictable ways, including under minor shifts in data distribution. These failures are exacerbated in the presence of adversaries (i.e., other actors or systems that attempt to undermine or attack the ML system) These vulnerabilities present serious concerns for deploying ML in high-stakes settings. This project aims to develop principled defenses that make robustness a core design property of ML systems rather than an afterthought. The approach is to bridge rigorous analysis with practical experimentation in order to understand, predict, and ultimately fix brittleness in modern models. Through new algorithmic tools and conceptual insights, this work will lead to the development of reliable and trustworthy AI systems that can operate safely and reliably in complex, real-world environments. This project will establish a framework of robustness via analysis for building machine learning systems. This approach interleaves simplified theoretical models with real-world experiments to derive actionable insights that improve robustness in practice. The project proceeds along three technical thrusts. First, it develops robust finetuning techniques for large pretrained models (foundation models) by minimizing forgetting and preserving generalization across domains. Second, it introduces defenses against adversarial attacks on language models, including novel finetuning paradigms that enable models to robustly self-correct and consistently enforce safety guidelines, even under complex and varied jailbreak scenarios (malicious attempts to trick the system into generating undesirable output). Third, the project studies robustness in autonomous AI agents composed of multiple subsystems, identifying weakest links, developing methods to enforce trust hierarchies (i.e. what can be trusted and when), and introducing mechanisms to provide formal safety guarantees across the system. Together, these efforts aim to shift the foundations of robust ML from heuristic patchwork to theoretically informed, systematically validated 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 · 2025-08

One of the foundational challenges in computer science is designing data structures that are both space- and time-efficient. Efficient data structures directly impact the performance, energy use, and cost of computing systems that underlie national-scale infrastructure—from supercomputers to cloud services to embedded systems in defense and transportation. In recent theoretical work, the PIs introduced a new technique, the tiny pointer, with the potential to substantially reduce the memory footprint of pointer-based data structures, enabling performance and efficiency gains across a wide range of systems. This project aims to translate tiny pointers from a theoretical insight into a deployable, foundational systems tool. By making tiny pointers broadly applicable, this research supports the national interest in advancing high-performance, resource-efficient computing critical to economic competitiveness, technological leadership, and secure infrastructure. At a high level, tiny pointers are compressed representations of memory addresses that retain compatibility with modern hardware and software systems. This project aims to develop foundational techniques for implementing tiny pointers, and the project has three primary goals: (1) to establish principles for building practical, scalable, and safe tiny-pointer abstractions; (2) to design and implement a high-performance library and OS module that expose these abstractions to both kernel and user-level applications; and (3) to demonstrate space and performance gains across a broad range of systems. Key applications include classical data structures (e.g., hash tables and trees), OS primitives such as page tables and page cache indexes, and flash translation layers (FTLs) in SSDs. For instance, by compressing pointers in x86-64 page tables, the project aims to increase fanout and reduce walk latency, potentially enabling a shift from 4-level to 3-level page tables. Similar benefits are expected for DRAM-resident mapping tables in SSDs. The project will also explore the impact of tiny pointers on systems such as IOMMUs, key-value stores, file systems, and large language models. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

One of the major approaches modern biologists use for understanding living things is to sequence their “genomes,” which involves reading their genomes and comparing them to known biological databases. A lot of software for analyzing genomes relies on a bag of tricks that scientists have learned to make work using trial and error, but without mathematical proofs that they work. This research will provide rigorous mathematical proofs for when those analysis tricks are guaranteed to work, and also extend those methods to additional biological analysis problems. Specifically, this research will analyze analysis tricks from genome “alignment,” which measure how many differences there are between two genomes, and then apply those tricks to the problems of discovering new variants of proteins and measuring RNA levels in a cell. The broader impact of this work is that researchers will then be able to build faster genomic analysis software, improving our understanding of when living cells produce different forms of proteins. Researchers will perform an average-case analysis of the seed-chain-extend string alignment algorithm to prove bounds on speed and accuracy for sequence alignment and read mapping software. The researchers previously performed such an analysis in the substitution-only error model, but here are extending their analysis to a more biologically-plausible error model including indels, duplications, and subsampling. Some of the probabilistic subsampling techniques will be used to improve RNA-seq quantification and novel isoform discovery. The aim is to improve the speed of RNA-seq quantification by not mapping every individual read and to improve the accuracy of novel isoform discovery by filtering the reads to plausible novel isoform candidates using a subsampling pre-filter prior to mapping. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Non-technical Abstract Understanding how electrons interact with each other in materials is one of the most exciting and challenging areas of modern physics. When these interactions become strong, they can lead to unusual and fascinating states of matter that can enable future technology revolution, such as materials that conduct electricity without resistance or those that could form the basis of future quantum technologies. However, these so-called correlated states are challenging to study due to their complex, many-body nature. This project focuses on exploring these correlated states in a new class of materials made from stacking atomically thin semiconducting layers with a slight twist or mismatch, creating a new periodic "moiré superlattice". This structure can slow down electrons and enhance their interactions, making it possible to observe and control new quantum states. The research team brings together experts in building these delicate materials, shining light on them to understand their optical properties, and using sensitive microscopes to probe their electrical behavior at the nanoscale. The project will explore how different ways of stacking the layers and controlling electron flows can create entirely new states of quantum matter. This project aligns with the goals of the National Quantum Initiative and has the potential to drive innovation in quantum electronics and optoelectronics. Beyond the science, this project will help train the next generation of scientists and engineers in fields of quantum optics, optoelectronics, and nanotechnology. The team will involve students ranging from high school to graduate school, with a strong focus on outreach activities. Activities will also include lab tours, educational modules, and hands-on research experiences, helping to grow and prepare the future quantum workforce. Technical abstract Recent advances in moiré superlattices of graphene and transition metal dichalcogenide (TMD) have demonstrated a promising platform to investigate correlated physics in two dimensions. This project aims to investigate correlated quantum phases in multilayer TMD moiré superlattices, leveraging state-of-the-art techniques in device fabrication, optical spectroscopy, and electrical scanning probe microscopy. The proposal focuses on three major research directions: (1) Study the evolution of excitonic insulator state in moiré superlattices involving a natural bilayer; (2) Explore emerging correlated states enabled by tuning interlayer coupling between moiré and non-moiré states. (3) Investigate correlated states in two coupled moiré superlattices. By controlling the electron tunneling from correlated electrons to designed bands, including moiré flatbands, this proposal plans to systematically explore new quantum correlated states and their unique valley/spin physics. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

The rapid advancements in neuroscience, robotics, and computer systems have underscored the vital interactions between mechanical and computational systems in shaping behavior. In natural systems, such as those found in animals, the brain and body must collaborate effectively for the successful navigation of a complex environment. The brain contributes computational intelligence, while the body provides mechanical intelligence. Integrating these elements—computational and mechanical intelligence—into the concept of mechano-computation represents a frontier in both robotics and neuroscience research. Progress in this field necessitates interdisciplinary communication and collaboration across various scientific domains. To propel this promising field forward, the Mechano-computation for Expanding Scientific Horizons (MESH) Network aims to unite diverse researchers from robotics, mechanics, materials science, neuroscience, information theory, biology, engineering design, and applied mathematics. Through workshops, travel grants, and the facilitation of collaborative projects, this network seeks to stimulate interdisciplinary dialogue, develop rigorous metrics for assessing autonomous systems, train the next generation of researchers, and push the boundaries of research in all areas of mechano-computation. By establishing a centralized resource for sharing findings, benchmarks, and methodologies, this network of researchers can accelerate innovation and position the United States as a leader in this transformative field, laying the groundwork for enhanced robotic systems in healthcare, agriculture, forestry, national security, and beyond. It may be argued that the full potential of robotics will not be realized until an intelligent physical body is purposefully designed from the outset, with careful consideration of both the available computational intelligence and the affordances the body can provide—affordances that, if appropriately leveraged, can offload and simplify computational demands by enabling efficient, embodied solutions to complex tasks. The Mechano-computation for Expanding Scientific Horizons (MESH) Network will bring together leading experts to tackle these critical challenges in autonomous systems through the integration of mechanical and computational intelligence. Creating intentional mechano-computation will enhance the design and control of autonomous systems, making them more efficient and explainable, and it will contribute to the development of innovative materials, mechanisms, and control strategies, pushing the boundaries of current research. Five key outcomes are anticipated as a result of the formation of the MESH Network: (1) A comprehensive theoretical framework and standardized metrics for mechano-computation; (2) Improved interdisciplinary collaboration and communication among researchers; (3) Long-term interactions among network members and early-career researchers, including nurturing graduate students trained at the intersection of disciplines; (4) Sharing of innovative materials, mechanisms, and control strategies; (5) Practical demonstrations by network participants of mechano-computation systems addressing societal and environmental challenges. The network will accomplish these outcomes through tasks that build online repositories of network critical technical and organizational information, in-person events to broaden discussion and collaboration, online communities, and targeted support for bringing in new collaborative research areas. This project is supported by the Dynamics, Control, and System Diagnostics (DCSD), the Engineering Design and Systems Engineering (EDSE) and the Mechanics of Materials (MoMs) programs of the Division of Civil, Mechanical, and Manufacturing Innovation (CMMI) in the Directorate for Engineering. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This CAREER project focuses on assessing the role of atmospheric new particle formation in influencing the pre-industrial climate. A new parameterization of new particle formation that accounts for the role of temperature and atmospheric ions will be developed and its effectiveness will be evaluated through climate models. This research will help to improve the understanding of aerosol-cloud interactions and reduce the uncertainties in the role of aerosol in climate evolution. Currently, NPF and aerosol growth is poorly represented in most global climate models. The newly developed parameterizations of new particle formation mechanisms will improve model predictions of cloud condensation nuclei (CCN) and cloud droplet concentrations (Nd). The chemical production of organic material from biogenic volatile organic compounds (BVOCs) that plays a role in NPF will be investigated with and without sulfuric acid, ammonia and atmospheric ions. A land surface model including prognostic fires will be used to determine the role of NPF and BVOCs in buffering aerosol variability due to biomass burning. The combination of empirical data from the CERN CLOUD experiment and extensive field measurements will enhance the accuracy and relevance of the climate modeling. The models will be used to improve constraints on pre-industrial cloud droplet concentrations and aerosol radiative forcing, by considering aerosol sources and sinks in different meteorological regimes and simulated changes from 1850 to 1950. The models also will be tested on the current-day atmosphere under a variety of conditions that enable an assessment of the ability to recreate the preindustrial atmosphere. Working with The Citizen Science Lab, an after-school pipeline program will be developed for high-school students in the region to gain experience in field measurements and scientific data analysis through demonstrations and hands-on activities. For example, students will make and interpret smartphone photographs of clouds using NASA’s Globe Clouds app and will be introduced to environmental history over the industrial period via case studies in the Pittsburgh region. 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 · 2025-07

Autonomous control systems in many domains, such as transportation, healthcare, assistance, and manufacturing, operate with humans in increasingly uncertain and interactive environments. Their design introduces unprecedented challenges in ensuring performance, safety, and human alignment due to a broad range of risks. First, existing techniques in risk quantification and control often assume observable state, system dynamics, and in-distribution data with sufficient risk events. However, these assumptions frequently break down in real-world scenarios. Second, inferring and reducing human-perceived risk requires characterizing safe actions from sparse feedback, but existing techniques often cannot handle this complexity. Finally, in nonstationary interactions, control methods that ignore the opponents’ adaptation can unintentionally encourage aggressive or exploitative behaviors by humans. In this proposal, we study how to provide long-term, lifelong assurances against various risks in the control of autonomous systems operating in uncertain and interactive environments. The outcome of this research will be broadly disseminated through seminars and tutorials and integrated into the PI’s classes and K-12 outreach programs. To enrich the learning experience, we will create a virtual game that allows students to understand key concepts through interacting with others. These efforts will train future researchers and engineers, and facilitate the transfer of insights across domains. The goal of this proposal is to develop techniques that mitigate various risks—visible or latent, actual or perceived—across single and repeated interactions. In Thrust 1, we quantify long-term risks despite latent variables and limited data. In Thrust 2, we develop efficient control techniques that provide long-term assurance against latent and human-perceived risks. One of our core innovations is probabilistic invariance, which flexibly handles unknowns, improves assurance time horizon vs. computation tradeoffs, and allows many risk inference methods to be leveraged in real-time control. In Thrust 3, we study the designs that avoid adverse adaptation or facilitate desirable behaviors in nonstationary interactions. Thrusts 1 and 2 will offer complementary insights into how short-term observations can be used to infer long-term risk and how controllers with short planning horizons can mitigate long-term risks. Thrusts 1-3 will provide a unified perspective across two pairs of topics that are often studied separately—control within a single interaction vs. adaptation across repeated interactions and alignment vs. deliberate misalignment. This unified perspective will enable us to proactively harness uncertainty, adaptation, and (mis)alignment to design safer, high-performing, and more human-aligned autonomous control 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 · 2025-07

Many scientific fields rely on high-performance computing in order to accurately simulate complex physical phenomena involving high dimensional tensors (multi-dimensional arrays). In particular, implementations of algorithms in the domain of computational quantum chemistry follow one of two paths: excruciatingly slow and error-prone expert hand-coding which results in very fast but inflexible code, or general, library-based (or code generation-based) development, which reduces development effort but often leaves significant performance on the table. The project's impact is to enable provably correct (correct by construction) code generation of key computational routines in a high-performance manner that can be incorporated into larger code bases. This contributes to the overall goal of whole-program verification of scientific applications. The project’s novelties are 1) new notations for describing operations involving structured matrices and tensors, 2) new insights to identify high-performance algorithms from their specification, and 3) integrating complex data movement into the specification of the algorithm to better identify and exploit optimization opportunities at various abstraction levels. Techniques and algorithms developed by the project will broadly impact the theory and practice of computational science, data science, and machine learning. All aspects of the work will involve the training of young scientists at the graduate and undergraduate student levels. This project leverages and extends the Formal Linear Algebra Methods Environment (FLAME), a proven methodology for the correct-by-construction development of dense linear algebra algorithms, to structured tensors and novel high-level linear algebra operations which commonly arise in computational chemistry. Real-world examples of linear algebra and tensor operations in computational chemistry applications will be used to develop new notations and abstraction levels. Desirable performance characteristics will be captured using the developed notations in the specification and at the loop invariant level to identify and derive high performance algorithms. The performance of the resulting implementations will be benchmarked against those in standard packages. 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 · 2025-07

Real-time artificial intelligence (AI) has become increasingly popular due to its ability to increase the accuracy of those tasks that need to be executed quickly, like the decision made by self-driving vehicles. In recent years, more complex use cases in different areas of science and beyond are emerging, where even higher accuracy is needed for sub-milliseconds tasks. In these conditions, more complex architectures need to be accelerated with dedicated hardware, by deploying them, for example, on specialized chips like field-programmable gate arrays (FPGAs). This award will develop a FPGA-accelerated graph neural networks (GNNs) to improve the real-time data filtering systems employed by high energy particle physics experiments. This work will not only promote the progress of science, but its impact can potentially transcend the field of high-energy particle physics, with potential applications in quantum computing, where it can improve the readout and control of qubits, or in autonomous navigation, where hardware-accelerated GNNs can improve the simultaneous localization and mapping of drones used for search and rescue. The Large Hadron Collider (LHC) at CERN will undergo a high-luminosity (HL) upgrade in 2030. It will deliver denser collisions, which will result in a dataset ten times larger, suitable for searches for rarer physics processes, as well as for higher precision measurement of particle properties. However, more particles per collision will be produced with increasing radiation. To cope with this, the Compact Muon Solenoid (CMS) experiment will upgrade its detector by installing a new radiation-hard high-granularity calorimeter (HGCAL). The HGCAL will have six million read-out channels and will produce hundreds of terabytes of data per second. Therefore, new techniques are required to reconstruct and select in real-time the most physics-sensitive collisions. This award will build a research program to improve the online HGCAL particle reconstruction used in the CMS real-time data filtering system, known as the ``trigger'', with cutting-edge GNNs deployed on FPGAs. By processing detector "images" as compressed by an auto-encoder, using directly the latent space and avoiding the decoding, the GNNs will perform a reconstruction that fits within the latency budget while, at the same time, making use of the information from the full detector. To achieve this, lightweight learning models that can balance simplicity and efficiency will be produced. In particular, for this project, virtual nodes augmented GNNs will be used to effectively capture both short and long-range interactions in spatially extended cell groups. Lightweight models for predicting particle groups will be explored, by for example, fast clustering in the latent space. The project will develop an ultra-low-latency FPGA accelerator to implement these new GNNs, which is also generic to enable algorithm/hardware co-design. The project will utilize Smart Network Interface Cards (SmartNIC) to perform GNN-specific edge embedding computation together with cross-machine communication, to reduce system-level latency and to improve scalability. 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 · 2025-07

The objective of this project is to support research on a novel paradigm for estimating, predicting, and managing traffic in response to cascading failures. Cascading failures occur when disruptions in critical components (e.g., bridge collapses, highway failures, or major public transit disruptions) trigger widespread ripple effects across inter-connected infrastructure and service networks. Leveraging low-cost, ubiquitous, system-level data, this project seeks to understand how and why travelers alter their routes and modes during extreme events. The theories and models are further validated with case studies in Pittsburgh, Pennsylvania, and Baltimore, Maryland. This project has the potential to enhance resilience of the nation’s critical infrastructure and lifeline services, ultimately saving lives and reducing economic losses. It supports development of efficient strategies for emergency response, daily operations, and long-term planning. The project promotes interdisciplinary education by integrating findings into undergraduate and graduate curricula, offering an online short course, and mentoring researchers. Project outcomes are broadly disseminated through open-source software, tutorials, international conferences, and policy briefs - ensuring benefits for government agencies, industry stakeholders, and the public. Research funded by this project seeks to develop a theoretical and computational framework to model spatiotemporal vehicular and passenger flow under cascading disruptions. It leverages high-granularity, ubiquitous, multi-year, system-level data (e.g., 5-min traffic speeds, transit vehicle locations, passenger counts, and high-resolution satellite imagery) to develop travel behavior models that capture travelers’ stochastic choices in mode, route, and departure time. By integrating both within-day dynamics and day-to-day adaptations, the project looks to fuse behavioral models with mesoscopic, multimodal network flow simulations. This integration is enabled by a novel computational graph (CG) design, which learns high-dimensional parameters to replicate observed flows, across large-scale networks and different disruption stages - immediate impact, partial re-opening, and full recovery. This framework seeks to facilitate the discovery of fundamental knowledge in how travelers respond to large-scale infrastructure and service failures, establishing a methodological foundation for traffic management under cascading disruptions. Replicable and transferable machine-learning based computational tools are provided to inform public agencies to design and implement mitigation measures. 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 · 2025-07

The recent growth in data-enabled science and engineering has ushered in a promising new generation of cyberinfrastructure (CI) technologies. These technologies use sophisticated features that have propelled the adoption and advancement of artificial intelligence (AI) in scientific and engineering research and discovery. While these technologies hold enormous potential to accelerate scientific discovery, their novelty and complexity often present significant barriers to effective use by researchers. This project, ByteBoost 2.0, offers a community-driven unified training platform where interested researchers and instructors can learn to use state-of-the-art sophisticated technologies. The program offers a series of specialized online training events followed by an in-person hands-on researcher-training workshop. The program enables researchers to gain the skills and strategic understanding required to effectively place scientific workflows on cutting-edge CI resources and train the new generation of researchers, accelerating the process of scientific discovery and innovation. Building on the lessons learned from a successful Pilot program, the ByteBoost 2.0 training platform advances the goals of familiarizing researchers and educators with advanced novel computing technologies. The program is modular and focuses on common challenges faced by researchers rather than a specific accelerator technology or discipline. The program offers an environment that accommodates interested campus, regional, and national computing technology testbed facilities including NSF-funded advanced cyberinfrastructure resources. The program begins with a webinar series open to a broad audience of computational researchers and instructors. These sessions introduce foundational topics, highlight key distinctions among computing technologies, and incorporate hands-on exercises. All training materials will be maintained in a learning management system. Participating researchers will also get access to existing asynchronous CI-courses to establish a baseline level of knowledge and CI skills. After the webinars, participants will apply to attend a five-day, in-person “Bring Your Own Science” workshop. At the workshop, they will work on a capstone project related to their research. The workshop will include educators who will develop curricular programs in collaboration with educational experts. During the workshop, all participants will work on their chosen research problems using the innovative technologies. Trained peer-facilitators and experienced scientists will offer recommendations on how to effectively utilize these systems. The community-driven focus helps integrate the innovative CI technologies into the growing research and instructional fabric. Working with active CI-engagement groups, ByteBoost 2.0 will broaden adoption of computation and AI at two- and four-year institutions, contributing to the goals of national AI research, entrepreneurial, and educational initiatives. The program will be offered on an annual basis. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-07

PROJECT SUMMARY During early pregnancy, the maternal uterine vasculature, particularly the spiral arteries, must undergo significant remodeling to establish uteroplacental circulation and create a hemodynamic environment suitable for placental and fetal development. Defective or incomplete spiral artery remodeling has been reported as a common characteristic of severe pregnancy disorders related to abnormal placental formation. Despite their importance, the mechanistic links between uteroplacental hemodynamics and placentation and their implications in fetal development have not been fully elucidated. Given the limitations of current technology to obtain in vivo data in the early stages of pregnancy, computational models provide a promising approach to investigate mechanisms of placentation and identify metrics for early diagnosis of pregnancy complications. Although models have already been proposed to study this process, they are not without limitations. Our preliminary data have shown that uncertainties related to placental anatomy and physiology can substantially affect simulation results. In addition, the effect of blood rheology or other assumptions related to uterine vascular remodeling and mechanical properties of the placenta and its vascular structures could significantly change local hemodynamics in this organ. Therefore, a systematic analysis of the effect of these factors to establish a robust platform to study pregnancy hemodynamics and vascular remodeling is much needed. We propose a computational study that leverages our expertise in multiphysics simulations and cardiovascular multiscale modeling toward maternal-fetal health applications. This project will focus on developing and validating a novel computational framework that can simulate uteroplacental hemodynamics and critical aspects of placentation and investigate the potential for applying these models to the early detection of pregnancy complications. We will do this through the following Specific Aims: 1) Simulate uteroplacental hemodynamics and determine the implications of spiral artery remodeling in the development of the placenta; 2) Determine metrics to measure the risk of adverse placental development by combining standard-of-care technology (Doppler ultrasound) and data from uteroplacental hemodynamics simulations. Successful completion of this exploratory project will establish a modeling framework for the simulation of the maternal-placental interface that would lay the groundwork for subject-specific modeling studies through an R01 award. More importantly, this model will offer a window into the early stages of pregnancy, which is currently challenging to study in humans, and will help generate data to inform future hypothesis-driven — experimental or computational — research.

NSF Awards · FY 2025 · 2025-06

Artificial intelligence (AI) is rapidly transforming workplaces, presenting both opportunities and challenges for millions of workers. While AI can automate repetitive tasks and enhance productivity, some fear that these technologies may threaten job quality. This project investigates how to design AI systems that prioritize workers' input and needs. The research explores three areas in the food industry: frontline service such as AI-powered ordering systems, food preparation such as robotics in kitchens, and cleaning such as autonomous sanitation technologies. The research team will engage workers as active contributors to the design and implementation of AI technologies. Findings will inform practices that balance technological innovation with the needs of workers, offering a blueprint for a prosperous future of work for all. The research employs a multi-method participatory design approach. Through ethnographic studies at several sites, the research team will document how AI technologies are integrated into workplaces and how workers adapt to these changes. Observation at industry events and interviews with workers, managers, and technology developers will examine the impact of AI on work practices, including the development of new skills and roles, and assess implications for workplace safety and job satisfaction. Participatory design workshops and iterative prototyping sessions will engage workers in co-designing prototypes for AI technologies. Strategies to include worker perspectives will be explored at every stage of the technology lifecycle: design, deployment, and oversight. The prototypes will emphasize collaboration, autonomy, and job satisfaction, fostering sustainable workplace transformations. The project will result in (1) empirical insights into worker-AI interactions, (2) novel worker-driven prototypes, and (3) actionable strategies for incorporating worker input throughout the AI technology lifecycle. These findings aim to guide the design of AI systems that balance the needs of workers and organizations, contributing to broader societal and economic resilience. 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 · 2025-06

This award supports the participation of students from U.S.-based institutions in the IEEE International Conference on Computational Photography (ICCP) 2025, taking place in person from July 19–23, 2025, at the University of Toronto in Toronto, Ontario, Canada. Now in its seventeenth year, ICCP is the premier annual conference dedicated to computational photography, bringing together researchers from optics, image and signal processing, computer vision and graphics, and sensor and electronics communities. The ICCP 2025 program features keynote and invited talks, technical paper presentations, poster and demo sessions, and networking opportunities—fostering interdisciplinary collaboration and providing an engaging environment for both junior and senior researchers to exchange ideas and mentorship. This award offers travel grants to 20 U.S.-based students, selected through a competitive process by a committee of ICCP 2025 organizers. Grants will partially cover travel, lodging, and registration expenses. Selected students are expected to fully participate in the ICCP program—including all talks, poster/demo sessions, and networking events—and will present their own research in a poster session. They will also take part in a mentoring event with academic and industry professionals offering guidance on research and career development. Additionally, recipients will attend the inaugural ICCP Summer School on Computational Imaging, featuring lectures by leading experts on active research topics in the field. Participation in ICCP 2025 will support students’ professional growth, enhance their research skills, and offer valuable networking opportunities. This initiative aims to strengthen the future STEM workforce by encouraging students to pursue careers in science and engineering, while simultaneously enriching the broader computational photography and computing research communities. Ultimately, this program contributes to increasing the number of active researchers and educators in STEM, thereby advancing the development of impactful technologies for society. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.