Carnegie-Mellon University

NSF Awards · FY 2025 · 2025-06

Markus Deserno of Carnegie Mellon University is supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to investigate how the asymmetry of cell membranes affects the phase behavior of their lipid constituents. While it is well established that the two leaflets of most biological membranes differ in their lipid composition, recent evidence shows that this asymmetry extends to many other properties, such as elastic material parameters or mechanical stresses. Prof. Deserno will develop theoretical and computational techniques to elucidate how these new facets of asymmetry alter lipid mixing in each leaflet. He will focus on three-component model mixtures (saturated lipids, unsaturated lipids, and cholesterol), whose thermodynamics is well characterized, at least in the symmetric case. Prof. Deserno will predict how such membranes react to controlled perturbations (e.g. changing cholesterol content), which will be tested in collaboration with experimental groups. By uncovering a new dimension of membrane organization, this award will advance fundamental understanding of vital physiological processes. At the same time, it affords academic training for graduate and undergraduate students and incorporates educational activities involving a local high school, as well as an ongoing physics teachers training program. Prof. Deserno will develop the theoretical framework and computational tools necessary to integrate differential stress into predictions for the phase behavior and mechanical stability of asymmetric ternary lipid bilayers. He will examine how trans-leaflet cholesterol partitioning and mechanical stress/torque balance constrain the choice of proper thermodynamic variables for such membranes. Elastic and thermodynamic drivers for cholesterol distribution and leaflet-specific non-ideal lipid mixing will be captured in a continuum-level free energy model, further tested and refined using coarse-grained simulations. Perturbations that move leaflets between two-phase coexistence regions, and their read-out in terms of cross-leaflet phase imprinting and curvature creation, will be analyzed and compared against experiments. Since transient scrambling has been proposed as a means to create raft-like domains in asymmetric but leaflet-wise well-mixed membranes, Prof. Deserno will examine the viability of such a “transient raft hypothesis” by determining what bounds the underlying dynamics of scrambling versus lipid diffusion. This project will also advance science education across multiple constituencies, ranging from graduate and undergraduate students conducting this research, to tailored outreach events given at a local high school, and “Physics Teachers Workshops” that have been run at CMU since 2021. 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

We live in a 3D world, and our perception systems, despite only receiving 2D retinal percepts, can effortlessly understand the underlying 3D structure, e.g., observing a set of stacked dishes, we understand how they support one another. Even more remarkably, we understand not just what it is, but also what can be, e.g., removing a dish in the middle can cause the tower to collapse. While the computer vision community has made impressive advances in developing computational systems that can reconstruct the underlying 3D from visual input, these systems do not understand about actions and their effects in this context. This project will bridge this gap and build perception systems which have an actionable understanding of the 3D world they observe. Such systems can be broadly useful across applications in computer vision, robotics, and mixed reality, e.g., allowing robots to intelligently act and efficiently learn in generic scenarios and helping virtual assistants better understand and guide their user's actions. This project will also contribute to the development of undergraduate and graduate students via research engagements and development of a specialized course on embodied agents, as well as benefit the community at large through dissemination of research and organization of tutorials. To achieve its goal, this project will make research contributions along three thrusts: a) developing approaches to learn about 2D affordances (what actions can be performed) and world models (understanding the effect an action may have), b) investigating techniques for similarly scalable learning for 3D, and c) leveraging the learned ‘world models’ for closing the perception-action loop for real-world manipulation. Specifically, this project will introduce mechanisms for learning expressive 2D affordance and world models from large-scale internet data, while exploring varying parametrizations for actions and states. To enable learning for 3D representations, this project will tackle the lack of training data via a framework for 3D reconstruction of generic interaction videos as well as novel a “reprojection consistency” formulation that side-steps the requirement of 3D supervision by leveraging generic 2D models. Lastly, current perception systems that learn about interactions often rely on passive (human or robot) data, and this project will develop mechanisms for actively using and improving the learned perception systems in context of real-world manipulation. 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

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.

NIH Research Projects · FY 2025 · 2025-06

PROJECT SUMMARY / ABSTRACT The NIH Common Fund supports transformative programs that generate large-scale datasets crucial for the broad biomedical research community. The emerging spatial transcriptomics in Common Fund programs provide invaluable insights into cellular diversity and tissue organization. However, integrating and analyzing these diverse datasets remains a significant challenge. To address this gap, we propose developing new machine learning methods to analyze spatial transcriptome data from multiple NIH Common Fund programs, in particular, HuBMAP and SenNet. First, we will develop machine learning methods for integrative analysis of multiple spatial transcriptome datasets. This will include creating a scalable, platform-agnostic framework using pretrained single-cell RNA-seq foundation models to integrate diverse spatial transcriptome datasets and a method to analyze spatial gene programs across multiple samples. Second, we will develop machine learning methods to reveal the underlying mechanisms of spatial gene expression. This will involve creating a transformer-inspired model to disentangle intrinsic cellular factors from intercellular interactions and a framework to identify transcription factors modulating spatial gene programs. Collectively, the methods and tools developed in this project will significantly enhance the utility of existing NIH Common Fund spatial transcriptomics datasets and facilitate integration with other related programs. This will drive forward our understanding of spatial gene regulation and its role in health and disease, aligning with the NIH Common Fund’s mission to support transformative research with broad scientific impact.

NIH Research Projects · FY 2026 · 2025-06

Project Summary Cryo-electron tomography (Cryo-ET) has emerged as a major tool for in situ biology which enabled the system- atic 3D visualization of subcellular objects and their spatial organization. Cryo-ET enables recovering entire subcellular morphologies and organizations across different cell populations, bringing significant potential for downstream biomedical tasks of phenotype-genotype analysis, disease diagnosis, and drug design. A major challenge in utilizing the potential of Cryo-ET for such downstream tasks is the lack of methods to systematically study the cell-population-wide differences in subcellular object morphology and their spatial organization. The goal of this program is to develop novel high-throughput machine-learning approaches to study cell-population- wide differences in subcellular object morphology, conformations, and organization. We set out to solve multiple computational problems toward this goal. The current deep learning-based cellular cryo-ET tomogram analysis approaches have limited data-efficiency and apply to one or a few tomograms having similar data domains. Consequently, they are unsuitable for large-scale analysis of tomograms of diverse data domains that can be used to image a cell population. In our first module, we will develop domain generalization methods to im- prove the data efficiency of existing deep learning-based cryo-ET image analysis methods. Our next module will develop high-throughput methods for coarse-to-fine segmentation of multi-scale subcellular objects in cryo- ET tomograms with unsupervised and weakly-supervised learning. In the following module, we will develop novel representation learning methods for statistical shape and morphology analysis of the subcellular objects. Finally, we will develop interpretable methods to study cell population-wide shape and spatial traits of multi- scale subcellular objects with multiple biological co-variates. In addition to the novel algorithms, we will create user-friendly software with a Graphical User Interface (GUI) incorporating the algorithms that will ensure the biologists can leverage cryo-ET for cell-population analysis.

NSF Awards · FY 2025 · 2025-06

This project concerns mathematical problems arising in the modeling of systems composed of individual particles, such as the atoms/molecules in a gas confined by a container. Particles repel or attract one another, and importantly, this repulsion or attraction may become increasingly strong as particles become increasingly close together. In principle, the equations of physics, such as Newton’s laws of motion, allow to completely determine the behavior of each particle in the system for arbitrary periods of time. However, in practice, the large number of particles, and therefore the complexity of the system, far exceeds even the best computing resources. To answer questions of interest, one needs a simpler effective description that is a good approximation. Following the prescriptions of statistical mechanics, much effort over the years has been devoted to achieving a reduction in complexity through a point of view focused on the probability of finding at a given time a particle in the system at a certain position in space and moving with a certain velocity. A first approximation, known as the mean-field limit, effectively captures the “typical behavior” of a particle in the system through a solution of a single equation, which is much simpler than the equations governing the exact motion of all the individual particles. However, our understanding of the mean-field limit over the timescales relevant for the system’s relaxation to equilibrium is quite poor. Additionally, the mean-field description ignores correlations between particles, created by their interactions and the thermal effects of the particles’ environment, which capture the disorder of the system. Little is known about this disorder, which is encoded in the probability of deviations from the mean-field limit. The distribution of these deviations is the next-order approximation beyond the mean-field. The project will advance our understanding of how the system disorder grows with the number of particles and changes over time. The results are of direct application for the modeling of states of matter, such as plasmas, as well as problems with particle-like behavior, such as vortices in fluids or superconductors and the training of neural networks. The methods developed to tackle these problems, based on measuring how the energy or entropy of a system varies in space and time, will establish new connections between different areas of mathematics of independent interest. A key feature of the project is the integration of research and educational activities, promoting the goals of the Federal 5-Year STEM Education Strategic Plan. Through summer courses for advanced high-school students, new undergraduate curricula focused on mathematical communication, working groups, and supervising undergraduate and graduate student projects, the research program will provide opportunities for the mentorship and training of a new generation of researchers in the U.S. at this intersection of mathematics, physics, and statistics. Building on the investigator’s extensive work, the project addresses fundamental, open questions for the large time, beyond leading order description of Coulomb/Riesz gases. The two main lines of investigation are (1) central limit theorems for fluctuations around the mean-field limit and cumulant expansions that capture the system disorder and corrections to the mean-field limit; (2) the interplay between generation of chaos in these systems (the phenomenon of the particles becoming independent and identically distributed in the joint large time and system size limit), log Sobolev inequalities (LSIs), and relaxation estimates for gradient flows. The project develops both the physical and mathematical understanding of Coulomb/Riesz gases. (1) advances several new conjectures for the scaling of fluctuations and cumulants, their limiting dynamics, and the optimal rate of convergence to Gaussianity, quantifying the disorder as a function of the system size, length scale of observation, temperature, and singularity of interaction. These questions are all open except in limited cases, and the project stands to shape the future direction of the field. (2) develops the new connection between generation of chaos and LSIs for Gibbs measures, which has wide application beyond Coulomb/Riesz gases, with emphasis on LSIs for nonconvex energies, which is a major challenge in the field. The project develops new mathematical tools, such as commutator and gradient flow relaxation estimates, which have potential applications across analysis/PDE and probability. Lastly, Coulomb/Riesz gases are an important testing ground for the longer-term goal of answering fundamental questions of statistical mechanics for more general systems arising in machine learning, physics, and statistics. 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

The human brain consumes millions of times less energy than artificial intelligence (AI) systems, yet it remains more flexible, versatile, and effective at solving complex problems. A key reason lies in the brain’s ability to learn abstract concepts and their relationships, and to construct internal models of the world--rather than simply memorizing and retrieving patterns from massive datasets. This project aims to investigate the computational mechanisms underlying a recently discovered neural process in the brain: the ability of neurons in the early visual cortex to rapidly form local recurrent circuits, enabling them to quickly encode relationships between concepts. Although these neurons primarily act as local feature detectors, they can dynamically adjust their responses based on global context through these recurrent connections. However, the computational rationale behind this mechanism remains poorly understood. The project seeks to develop a machine learning framework to formalize this rapid recurrent plasticity. The central hypothesis is that these recurrent circuits do more than encode global context as attractors--they also perform manifold transformations that bring semantically related concepts closer together in the neural activity latent space for encoding dependency among concepts, while compressing irrelevant dimensions of variation. The investigator will test this theory in tasks such as associative learning, de-noising, and pattern completion, combining computational modeling with neurophysiological experiments. The project will evaluate the predictive power, limitations, and practical utility of these biologically grounded models in real-world computer vision applications. This research will offer a new conceptual framework for understanding cortical recurrent circuits and pave the way for a new generation of biologically inspired AI systems--systems that are more efficient, generalizable, and aligned with the flexible intelligence of the brain, with broad potential impact on neuroscience, technology, and national security. 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 · 2025-05

Abstract Small molecule kinase inhibitors are a class of targeted therapeutic anti-cancer agents that selectively target specific kinases associated with a given cancer and are generally less toxic than general-purpose chemotherapies. Due to the highly conserved nature of the ATP-binding domain, kinase inhibitor off-target activity is often observed. This off-target activity can be beneficial--for instance, when the kinase inhibitor is active against multiple drug targets--or it can be deleterious, leading to toxic side effects. Thus, knowledge of off-target activity is essential during the early stages of the drug discovery pipeline; however, as experimental screening is costly and time consuming, off-target activity is often only explored in the final stages of lead optimization. Free energy-based in silico methods such as alchemical relative binding free energy (RBFE) simulations, which are becoming prevalent in traditional single target drug discovery efforts, afford a cost-effective solution; but, to date, no one has attempted their use for predicting off-target activity due to the cost and complexity of simulating multiple protein-ligand complexes for many ligands. To address this shortcoming, we propose to combine novel methods for reducing the computational expense of each simulation with multi-objective active learning to limit the overall number of requisite simulations to explore a chemical space. In Aim 1, we propose a novel method for the optimal stratification of the phase space for RBFE simulations to reduce computational expense while maintaining requisite accuracy. Combined with our recently developed on-the-fly optimization method, we will greatly reduce the computational cost of RBFE simulations. In Aims 2 and 3, we will combine these methods with our group’s recently developed free energy-based active learning workflow implemented for multi-objective optimization of ligand binding affinity within a chemical space, representing the first use of free energy-based methods for the prediction of affinity to multiple proteins in a drug discovery setting. In Aim 2, we will utilize this workflow to optimize ligands for selectivity to a kinase target and against activity of known off-target kinases. We will test this workflow by targeting the activin-like receptor kinase 2, mutations of which are associated with the fatal pediatric brainstem tumor diffuse intrinsic pontine glioma, while penalizing activity against other activin-like receptor kinases. In Aim 3, we will utilize this workflow to optimize multikinase inhibitors for activity against multiple known drug targets. We will test this workflow by optimizing the hit compound which led to the development of the multikinase inhibitor Entrectinib against anaplastic lymphoma kinase, c-ros oncogene 1 kinase, and tropomyosin receptor kinases A-C. By demonstrating the feasibility of free energy-based multi- objective optimization for affinity to and for selectivity over multiple kinases, we will greatly increase the efficiency of the early drug development pipeline for targeted therapeutics.

NSF Awards · FY 2025 · 2025-05

Robotic hands with human-like dexterity are urgently needed to address multiple challenges facing our nation and society, including enhancing productivity in manufacturing and improving quality of life. According to Deloitte, by 2028 labor shortages could jeopardize more than $450 billion of U.S. GDP in the manufacturing sector. Meanwhile, over 400,000 Americans live with upper limb loss, and more than 20 million people struggle with daily activities due to limited mobility. While automation has the potential to alleviate these pressures, current robotic solutions have not been effective. Robotic hands have yet to gain widespread adoption in factories, and upper limb prosthetics face the highest rates of abandonment of any assistive device. To address these challenges, this project brings together a diverse team of university, industry, and government participants to develop, pilot, and create sustainable pathways for robotic and prosthetic hands that are bio-inspired, feature all-over sensing and intelligent control, and are customized to the needs of the end-user. This project has four specific goals for impact: (A) ease labor shortages in a manner that benefits both manufacturers and employees, (B) prepare new learners for a future of work that includes dexterous robotics, (C) ensure that small and medium manufacturers have access to dexterous robot hands for high mix / low volume applications, and (D) enable individuals with disabilities to have greater independence to meet their personal goals. Research thrusts to achieve these goals emphasize (1) bioinspired sensors, skeleton, and skin designed for customizability, reliability, and manufacturability, (2) bioinspired fine control with intrinsic muscles for dexterity in a lightweight and compact package, and (3) a software pipeline for custom robot hand design for best end-user experience, performance, and reliability. Research thrusts are accompanied by development efforts, customer outreach, and educational activities to ensure that impact goals can be met with a sustainable solution. A hand that is humanlike and dexterous can enable people to do dangerous work from a safe, remote location or to provide expert assistance anywhere in the world. A robust solution to robot dexterity will see application in homes, workplaces, farms, helping individuals to age gracefully in place, small business owners manage labor shortages and upscale jobs to improve quality of life. 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 · 2025-05

This Faculty Early Career Development (CAREER) award supports research that aims to investigate an innovative power field control strategy to achieve prescribed thermal histories throughout a part in powder bed fusion additive manufacturing. The metal powder bed fusion market, projected to grow by 683 million USD by 2027, has about 25 percent of its market share in the aerospace and defense sectors, where qualification and certification are critical factors for its adoption in manufacturing. The qualification and certification approach typically takes years and costs millions of dollars. A key obstacle in the workflow is the presence of a process-induced location-specific variation of the thermal history during conventional additive manufacturing processing. This affects the microstructure, defects, and properties of the material. The findings from this project are expected to enable the design of novel processing pathways to prescribe desired thermal history in the part and tailor properties. If successful, this approach will fully utilize the processing possibilities offered by the open architecture powder bed fusion additive manufacturing machines and break the conventional processing paradigms. The integrated education and outreach plan aims to address the current, emerging, and future manufacturing workforce needs through training machine operators, integrating research findings into graduate and advanced undergraduate curricula, and via summer camps for K-12 students. This CAREER research project aims to investigate an innovative power field control strategy to achieve prescribed thermal histories throughout the part in powder bed fusion additive manufacturing. This will be accomplished via research that attempts to address the following objectives: (1) Understand the effects of power field control on porosity and microstructure, including solidification and solid-state transformations, and (2) Evaluate the effectiveness of power field control in decoupling thermal history from part geometry. The experimental tasks leverage powder bed fusion-electron beam process to demonstrate control over final microstructure and mechanical property in the part. The tasks are organized to design power fields for achieving a target thermal history by utilizing in-situ process monitoring and characterization to evaluate the effectiveness in maintaining consistent part quality irrespective of the part geometry. The generated knowledge could advance other fields including, manufacturing process planning and control, process monitoring, and multi-scale material characterization. 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 · 2025-05

The award will provide partial support for a three-day workshop on Quantum Sensing for Biomedical Applications planned for August 4-6, 2025 at Carnegie Mellon University, Pittsburgh, PA. Two days will be devoted to intensive lectures on mathematics essential for understanding quantum mechanics, followed by foundational quantum physics necessary for comprehending quantum sensing. The last day of the workshop will expose workshop participants to leading-edge research presentations delivered by invited guest lecturers. Participants for the workshop will be selected from junior and senior undergraduate and first-year graduate students, and young professionals. The majority of about 55 students attending the workshop will be recruited from Pittsburgh area universities that include Carnegie Mellon University, Duquesne University, and the University of Pittsburgh, and six will be recruited from other universities. Quantum technology, and quantum sensing in particular, are expected to have broad and significant impact. The cross-disciplinary nature of the field requires foundational knowledge in various areas of mathematics and physics; these are in addition to knowledge in specific application areas. The purpose of the workshop is (i) to introduce junior and senior undergraduate and first-year graduate students, and young professionals to essential foundational knowledge on quantum sensing, and (ii) to expose them to leading-edge research in the field. This activity will benefit participants by making them aware of potential opportunities in an emerging field early in their technical education. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-05

Nontechnical Description: Nanotechnology plays a critical role in fueling advances in information technology. Miniaturization and scaling of the basic devices that underpin all leading technologies in computing, communication, and sensing are based on innovations, and therefore research advances, in nanotechnology. This project will support the 19th Nanotechnology Forum between the United States and South Korea, which will be held July 3-4, 2025, in Kintex, Gyeonggi-do, South Korea. Past nanotechnology fora between the two countries have addressed leading-edge topics at the time of each forum. The present forum will continue this tradition by focusing on nanotechnology issues related to Neuromorphic Computing, Quantum Sensing, and Sustainability in Semiconductor Manufacturing by Design. The key objective of past fora, including this one, is to facilitate networking among members of the research communities and funding agencies of the two countries. The forums have always enabled exchange of technical information and exploration of opportunities for collaborations. This was the motivation that led to the convening of the 1st U.S.-Korea Forum on Nanotechnology that was jointly supported by the U.S. National Science Foundation and its Korean counterpart. Technical Description The 19th U.S.-Korea Nanotechnology Forum will bring together a bi-national community of expert researchers and innovators, working on leading-edge, next-generation neuromorphic systems, novel quantum sensing schemes, all based on sustainable semiconductor manufacturing processes. This forum will stimulate efforts to promote these areas. Expected outcomes of the forum will lead to definition of new milestones and vigorous research collaborative opportunities in both countries, thus advancing investigations of scientific and technological questions with potential to make significant economic development, welfare, and security. The exchange of ideas among leading scientists and engineers in the neuromorphic, quantum sensing, and semiconductor technology areas, will spur new paradigms of investigations in these areas, thus promoting collaborations in them. In addition to individual collaborations, the forum is expected to nurture development new research frameworks by the funding agencies of the two countries. The discussion will enable the U.S. to pivot towards transformative approaches for novel semiconductor manufacturing and sensor development. 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 · 2025-05

As robotics technology becomes more widely adopted, including warehouse robots, autonomous vehicles, and delivery drones, there is a critical need to address the challenges of coordinating large teams of autonomous agents in congested environments. These challenges arise from the complexity of navigating shared spaces, where multiple agents must work simultaneously while completing their tasks quickly and safely, all while avoiding collisions and reducing congestion. While considerable efforts have focused on developing advanced planning algorithms to generate high-quality, collision-free paths for these mobile agents, there has been limited focus on improving their performance through environment optimization. In this context, "environment" refers to the various factors that influence how autonomous agents operate. This includes the physical layouts, such as the arrangement of shelves in a warehouse or the design of roadways, which dictate how agents navigate through space. Additionally, it encompasses virtual elements, such as traffic rules and operational protocols, that guide the behavior of these agents. Furthermore, current conventions for designing layouts and traffic rules in environments such as warehouses and road networks are primarily geared toward human operators and human-driven vehicles. This focus can lead to suboptimal conditions for autonomous agents, which exhibit different behavior patterns and operational needs. Consequently, this project has two main objectives: (1) establish a foundational understanding of the importance of environment optimization for multi-agent coordination, and (2) develop innovative and potentially unconventional environment designs that are specifically tailored for large-scale multi-agent coordination across various applications. This project aims to establish a comprehensive, versatile framework for optimizing environments tailored to large-scale Multi-Agent Path Finding (MAPF), centering on three foundational research areas: models, methods, and applications. It studies how to model MAPF environments, which involves optimizing physical layouts, constructing graphs optimized for MAPF, and introducing MAPF traffic rules. It explores different methods to optimize these environment models, which involves not only methods that directly optimize the models but also methods that optimize environment generators and updaters. Environment generators, after training, can generate environments with similar setups on potentially larger scales, and environment updaters can adapt environments in real time to accommodate dynamic changes in agents' tasks and other variables. The project focuses on three distinctive case studies, namely automated warehouses for mobile robots, unmanned traffic systems for drones, and decentralized MAPF with guarantees, to exemplify the broad applicability and efficacy of the established frameworks across various domains. By significantly enhancing the coordination of autonomous agents, this project has the potential to greatly increase efficiency and safety in warehouse and delivery operations, leading to faster processing times and safer navigation in congested environments. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-05

This award provides travel support for selected student participants of the Doctoral Consortium (DC) at the 18th International Symposium on Combinatorial Search (SoCS-2025), which will take place in University of Glasgow in Scotland, United Kingdom from August 12 to August 15, 2025. SoCS-2025 is an international conference in artificial intelligence (AI) focusing on combinatorial search methods, which are techniques to effectively explore and find the best solutions from a combinatorially large set of possibilities. The doctoral consortium at SoCS-2025 will provide a platform for the students to present their work, receive constructive feedback, and engage in insightful discussions with their peers and established researchers. In addition, students in the doctoral consortium will also have the opportunity to attend and participate in the main technical program of SoCS-2025. This will allow the students to gain exposure and develop a deeper understanding of cutting-edge advances in the field of combinatorial search. 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 · 2025-04

This grant supports student travel to the Eighth Conference on Machine Learning and Systems (MLSys 2025), which will take place at the Santa Clara Convention Center in California, United States, from May 12th to 15th, 2025. The conference focuses on research at the intersection of machine learning and systems, an area that hasn't been extensively explored by traditional Machine Learning or Systems conferences. By bringing together these fields, MLSys aims to foster new connections, identify best practices, and establish design principles for learning systems. It also seeks to develop innovative learning methods and theories tailored to practical machine learning workflows. The MLSys community has enjoyed a close interaction between academic researchers and industrial designers and aims to continue this tradition at MLSys 2025. Graduate and undergraduate student attendance is vital to the growth of the MLSys community. Students receiving travel support will benefit from the opportunity to engage in the technical, professional, educational and social exchanges that MLSys fosters. The MLSys conference serves as an open forum to exchange research results and develop new ideas, as well as an educational platform to expose students to new research methodologies, tools, and infrastructures, and as a social event to network and prepare students as future researchers and leaders. The conference will also feature a Young Professionals Symposium that will especially benefit students and young researchers. This grant will enable students who might otherwise be unable to attend to participate fully, ensuring they can benefit from the mentorship, interactions, and networking opportunities available at MLSys 2025. 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 · 2025-04

In software engineering, date and time are fundamental concepts. Whether it is scheduling flights, processing bank transactions, computing payroll wages, validating digital certificates, controlling industrial processes, or logging operational data—software relies heavily on making accurate calculations related to dates and times. However, code related to date and time calculations is often also a source of human error, due to the inherent complexity of concepts such as calendars, time zones, and daylight savings, as well as the wide variety of textual representations and international conventions used across different industries. Moreover, most software makes use of third-party components, called libraries, for providing functionality related to date and time. Errors in any of these components can potentially impact the reliability of a huge number of software systems which depend on their correctness for day-to-day operations. This project aims to first systematically study past software bugs related to date and time computation by analyzing source code repositories, and then, to develop techniques for automatically uncovering new date and time related errors that may be present in existing software. Successful completion of this project will improve software systems that our society critically relies on against a pervasive class of errors. The research project will also result in materials that will be incorporated in software engineering courses. At the same time, the project will provide research opportunities for undergraduate students and students from underrepresented groups in computing. Consistently performing date/time computations accurately is challenging due to: (i) fundamental complexities with the domain such as dealing with leap years, time zones, daylight savings, clock drifts, diverse data representations and string formats, etc.; and (ii) the heterogeneous landscape of date/time interfaces across different programming languages and third-party libraries, which all provide similar functionality but use subtly different representations, conventions, and default behavior. This project aims to strengthen the correctness of date and time computations performed in software systems via: (i) a systematic study of date/time-related issues in open-source repositories; (ii) the development of static and dynamic program analysis techniques to uncover date and time bugs in open-source date/time libraries, as well as client software that makes use of such libraries; and (ii) enhancing support for formal reasoning of date-based constraints in theorem provers. 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 · 2025-04

The Center for Materials Data Science for Reliability and Degradation (MDS-Rely) is adding a new site at Carnegie Mellon University (CMU). The center is dedicated to enhancing the reliability and longevity of essential materials through data science. By leveraging advancements in computing and data analysis, MDS-Rely addresses critical issues in materials degradation, which is vital for a broad spectrum of industrial sectors. MDS-Rely could transform electronics, energy, aerospace, and manufacturing by enabling better materials selection, failure prediction, and performance optimization. It will also contribute to next-generation electronics, printed sensors, and coating technologies. The center's research aims to develop innovative solutions to improve material performance and reliability, thereby supporting the U.S. economy and national interests. Some examples of these interests include more reliable processes for additive manufacturing of parts, developing formulations for consumer products that reduce degradation, and better methods for quantifying degradation in products and processes. MDS-Rely's interdisciplinary approach brings together industry leaders, government labs, and academic experts to tackle grand challenges in materials science. The center's work will advance scientific knowledge and foster workforce development by training students and professionals in cutting-edge data science techniques. MDS-Rely focuses on advancing materials reliability through data science and machine learning. Machine learning enables researchers to build predictive models for how material properties depend on processing or how they degrade during usage from data. These models can then be used to predict material behavior, or to optimize material properties to mitigate degradation and to increase their reliability. The center's research thrusts include developing solutions for materials degradation, creating robust study protocols, and applying machine learning to predict material performance. CMU's unique contributions include expertise in machine learning, and in developing sustainable materials. CMU brings advanced manufacturing and characterization facilities to support interdisciplinary research between machine learning practitioners, materials researchers and industry participants. By fostering collaboration between academia, industry, and government, MDS-Rely aims to enhance material reliability, support workforce development, and drive technological advancements. The center's strategic goals have evolved to include a stronger focus on sustainability and the application of machine learning to new materials challenges. This approach ensures that MDS-Rely remains at the forefront of materials research, and furthers NSF’s mission by contributing to better engineered products and technology through improving the reliability and performance of critical materials. 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 · 2025-04

Project Summary This proposal aims to develop an integrated ultrahigh-throughput 3D volumetric super-resolution imaging system by combining a multiscale fluorescence mesoscope with Magnify Expansion Microscopy. Current imaging systems are largely limited to 2D tissue sections at single-cell resolution, which fails to capture the subcellular interactions critical in cancer research. Our innovative system will allow for rapid, large-area 3D imaging with exceptional detail, enabling the study of interactions between cancer and immune cells in situ. We will employ a novel specimen immobilization approach, advanced fluorescence staining techniques, and a fully automated workflow to enhance reproducibility and throughput. By using a genetically engineered mouse model of lung cancer, we will visualize subcellular changes in immune synapses and identify distinct spatial clustering during tumorigenesis and immune invasion. This project will advance the application of spatial biology in cancer research, offering new insights into immune evasion mechanisms and improving immunotherapy strategies. Our technology promises to revolutionize our understanding of cellular interactions in oncology, providing high- precision, large-scale 3D super-resolution imaging capabilities.

NSF Awards · FY 2025 · 2025-03

People interact with materials all the time. Typically materials are used for specific purposes, for example, wood to hold up rigid floors or foam for cushioning. These materials do not change their properties as people interact with them. While a rigid floor is great for walking, cushioning properties would be desirable to prevent injuries if a person falls. This project will develop materials that can change their properties to adapt to people and their actions. Generally, materials consist of microscale structures. For example, looking very closely at foam materials that are often used for packaging reveals that they consist of many tiny randomly oriented struts. This project will generate strategically oriented struts to perform a desired deformation when people interact with it. In the floor example, structures that soften upon a large impact, such as a person falling, will be generated. Similarly, material structures to provide ergonomic support can be computed, such as a chair cushion that becomes thicker in the back when people sit down to promote an upright posture. Such materials can also be tailored to people's bodies, for example, as assistive braces for rehabilitation that can adapt their stiffness with every step. The scientific aims of this project are to understand when and how users will interact with such adaptive materials and how to design and fabricate them. Materials governed by their microstructures (also known as mechanical metamaterials) with designed bulk behavior have been investigated in engineering disciplines yet a deeper understanding of their human-centered qualities is missing, for example, their capabilities for interaction, user needs, or utility. To approach such interactive passive materials systematically, this project will address three classes of challenges. The first challenge is understanding how users will interact with the materials. The second challenge is developing advanced computational design tools that optimize new structures and allow them to be shared, evaluated, and reused. The final challenge is the long-term integration and in-depth user evaluations of applications. The project will first create a robust analysis of application domains and user evaluations of initial prototypes, resulting in a new design space that characterizes the material-changing interfaces. Based on these findings, libraries of structures and computational design tools will be built to foster a growing repository of designs and associated performance evaluations, enhancing collaborative opportunities and innovation in the field. With this foundational framework, design tool, and evaluation protocols, application areas will be investigated in-depth. 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 · 2025-03

This project develops and tests a new tool that measures how much a person thinks about what they should do based on what they expect other people to do. This is an important topic to study because people may behave differently depending on expectations and such thinking can affect important outcomes like pricing. For example, when buying a house Person A may think a price should be based on its current value, but Person B might expect that other buyers will be optimistic about future growth. Thus, Person B may pay more now, thinking they can sell at a higher price later. When many people think like Person B, this can drive up home values beyond their actual worth creating a housing bubble. Another example is salary expectations. When starting a new job Person A might just ask for what they think is a fair salary, but Person B may expect their employer to give a low counter offer after a salary request, and thus ask for more up front, ending up better off. These different strategies can lead to unfair differences in pay. Many economists have shown this to be the case but there is a lack of research on understanding characteristics of people who engage in greater thinking about others’ behaviors. This project develops a tool to measure strategic anticipation assessing how much strategy a person uses in how they anticipate, or expect, other people to do in response to their own behavior. If scientists can measure people's strategic anticipation, they can design experiments to test how much it matters in many different settings, such as buying houses and negotiating with other people to get the best salaries. A better understanding of these processes can help people to make better decisions and get fairer outcomes. This project develops and validates a novel measure of strategic anticipation - the ability to engage in recursive thinking about others' mental states and potential actions. Drawing from economic theories of level-k reasoning and psychological research on theory of mind, the Strategic Anticipation Measure (SAM) provides researchers with a standardized tool to assess individual differences in strategic iterative thinking. The research involves replicating and validating the SAM in a nationally representative sample, ensuring psychometric rigor through measures of discriminant validity and test-retest reliability. Additionally, the project applies the SAM across various social contexts—including voluntary disclosure in test-optional college admissions, self-sacrifice in close relationships, strategic decision-making in resource competition, and responses to shared knowledge in bystander intervention. By establishing the SAM as a reliable measure, the project aims to open new avenues for theoretical and empirical research, advancing our understanding of strategic anticipation in multiple domains of human behavior. 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 · 2025-03

This Faculty Early Career Development (CAREER) grant will support research that attempts to expand the understanding of how robots can make safe decisions. Ensuring robot safety is a complex challenge, as robots must understand how their actions could lead to harmful outcomes. Most robot safety efforts focus primarily on preventing collisions. However, safety concerns in real-world environments, such as homes, city streets, and hospitals, are far more nuanced. For example, robots should avoid entering areas marked with caution tape, slow down when transporting hot cups of coffee to prevent spills, and seek clarification when uncertain about a task. This research aims to develop new methods that enable robots to better understand and respond to these nuanced safety challenges. The project intends to advance scientific knowledge and contribute to national well-being by making robots more trustworthy and effective in a wide range of settings, from personal homes to public and professional spaces. Additionally, the project will support robotics education through the following initiatives: i) Developing a “Robotics Red-Teaming Challenge,” where college students will design and stress-test robot safety algorithms; ii) Engaging K-12 students in hands-on experiences with robot programming and troubleshooting; iii) Collaborating with government and industry stakeholders to establish safety benchmarks and promote the broader adoption of safe robotics practices. This project will develop an algorithmic framework to align a robot’s understanding of safety with that of human stakeholders. The approach builds on foundational methods in safe control, which characterize safety constraints as arbitrary sets in state space. While this mathematical model is theoretically highly expressive, its practical application has been limited to controlled, collision-avoidance scenarios. This award supports fundamental research to generalize this framework to more nuanced safety specifications by embedding learned patterns about the real world, inferred from robot and human data. Research conducted under this grant will strive to enable robots to: i) Automatically assess the trustworthiness of their learned models of human behavior and react accordingly; ii) Update their internal safety representations using online human feedback, such as language; iii) Generalize their safety policies to account for uncertainty and learned latent state spaces. These advancements will be evaluated through hardware experiments, where robot arms and mobile manipulators will be deployed in unstructured environments, such as kitchen and atrium settings. In addition to these research objectives, the project includes a robust education and outreach plan to teach the next generation a nuanced yet practical perspective on robot safety. This will emphasize the often subtle and unanticipated consequences of robot interactions in real-world contexts. 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 · 2025-03

For many years, cryptography has enabled a secure communication infrastructure by relying on the difficulty or "hardness" of solving special mathematical problems. These computational problems, which are hard on average, form the foundational building block of modern cryptography. Yet, despite decades of research studying and leveraging these problems, only a few sources of hardness are commonly used in cryptography. This project is dedicated to a holistic study of new sources of hardness, including: (a) a fundamental exploration of the hardness of new cryptographic assumptions, (b) the design of new techniques to leverage these hard problems for achieving new feasibility results and more efficient cryptographic mechanisms—such as efficient encrypted computation, including multi-party computation and homomorphic encryption — and (c) addressing the severe shortage of assumptions underpinning post-quantum cryptography. The project aims to systematically study the theory and applications of code-based and multivariate cryptography, as well as new assumptions based on point lattices. Additionally, the project aims to advance the use of natural computational problems that arise in the domain of statistical inference for cryptography and the underlying mechanisms (such as low-degree tests and sum-of-squares lower bounds) to rigorously analyze any new assumptions. The investigator devotes time to the Prison Math Project that aims to improve employment prospects and reduce recidivism of individuals upon release from correctional facilities. Additional educational efforts of the investigator should also serve to further bridge the fields of cryptography, complexity theory and statistics. 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 · 2025-02

Software engineering (SE) is a rich field focused on improving the effectiveness of software development through a variety of tools and techniques. Research in software engineering brings core aspects of computer science, such as programming languages and artificial intelligence, to bear on software concerns such as privacy, security, scalability, and adaptability. This project supports an ongoing Research Experiences for Undergraduates (REU) site at Carnegie Mellon University that provides students with training in interdisciplinary software engineering research, including understanding the research literature, formulating and refining research questions, developing novel solutions to software engineering problems, and applying scientific evaluation methods. The project will support the training of 30 undergraduate students over 3 years with research experiences in interdisciplinary software engineering at Carnegie Mellon University. These students will be recruited primarily from institutions where students have limited research opportunities, with a special emphasis on providing opportunities to first- and second-year students. Participants will perform research with faculty mentors, participate in community activities, and communicate their work both through traditional publication channels and student research competitions and via department-wide presentations at the end of the summer. A primary goal for the program is to improve the pipeline from college to graduate school and from graduate school into faculty positions; simultaneously, the site actively trains graduate students to act as effective mentors of research assistants, a critical skill for future faculty members. The project will build on the experience of many of the faculty mentors who have worked with undergraduate students in their individual research. While built on computer science fundamentals, software engineering is simultaneously an engineering discipline. That is, both the practice of SE and SE research problems revolve around technical solutions that successfully resolve conflicting constraints. As such, trade-offs between costs and benefits are an integral part of evaluating the effectiveness of methods and tools. This REU site takes this further by tackling research problems that are at the intersection of software engineering and another field, including privacy, security, mobility, and human psychology and other sciences. A subset of these problems includes machine learning in production; automatically debugging robotic systems; formal verification of high-level multiparty cryptographic protocols; and sustainable open-source communities. 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 · 2025-02

Abstract In order for wearable robotic exoskeletons to assist the American public throughout daily life, researchers need to develop a control framework that satisfies real-world use. Over 10% of adults have difficulty walking, which hinders their ability to perform daily activities, maintain independence, and have a satisfactory quality of life. To address this issue, wearable exoskeletons have the potential to augment the walking ability of a diverse array of community members throughout their daily lives. That is, if researchers can establish an exoskeleton control framework that is 1) easy to use and 2) adequately assists the walking needs of users throughout daily life. However, current exoskeleton controllers can only assist a few stereotypical movements or require hours of arduous expert tuning using specialized equipment. Thus, there is a critical need for an exoskeleton control framework that rapidly and easily tunes to the diversity of user movement patterns during real-world ambulation. Until such exoskeleton controls exist, many community members, especially those with distinct movement patterns and limited resources, will continue to lack the mobility to achieve independent community ambulation. Our long-term goal is to develop an exoskeleton control framework that is easy to use and can quickly tune to any user and effectively assist their daily ambulation. Here, we will progress towards our goal by developing and evaluating a hip exoskeleton control framework that leverages artificial intelligence to rapidly tune to both young and older adult movement patterns in minutes. We expect such exoskeleton tuning to improve user ability to navigate an outdoor course with hills, stairs, and turns better than current ‘state-of-the-art’ exoskeleton controllers. Further, we will mechanistically explain how each exoskeleton control framework affects user walking performance by pairing the outdoor testing with indoor lab tests that involve detailed physiological measurements. The first aim of this research focuses on developing ‘Trailblazer Exoskeleton Control’ - a versatile hip exoskeleton control framework that leverages artificial intelligence to interpret the movements of new users in minutes via robot integrated wearable sensors, thereby enabling a non-specific and task-agnostic control strategy. The second aim’s objective is to evaluate the ability of young and older adults using Trailblazer Exoskeleton Control and three alternative conditions to navigate an outdoor walking course. Our engineering innovation using artificial intelligence to develop a new and easy to use exoskeleton control framework will set the stage of wearable robotic exoskeletons to assist the movement patterns of community members across the lifespan.