Carnegie-Mellon University

NIH Research Projects · FY 2026 · 2026-04

PROJECT SUMMARY/ABSTRACT The goal of this project is to investigate the suitability of a smart garment (sensorized compression sleeve) as a muscle-activity monitoring and rehabilitation tracking device in patients following anterior cruciate ligament reconstruction (ACL-R). Every year, nearly 400,000 Americans rupture their ACL. Although ACL-R and physical therapy allow return to normal activity within a few months, long-term outcomes are discouraging, with 20-30% experiencing a reinjury in the following year and 50% developing post-traumatic osteoarthritis (PTOA) within one decade after the surgery. It is widely appreciated that failure to restore pre-injury biomechanics is one of the key contributing factors to poor long-term outcomes. While muscle strength is a primary target of physical therapy, studies document continued atrophy and weakness. It has therefore been hypothesized that the mechanism of muscle atrophy in ACL-R patients may be different from that of disuse, with at least a segment of this population not benefiting from continued muscle-strengthening exercises. Traditional tools used in the clinic to personalize therapy and clear patients for return to play are limited not only in terms of technical innovation, but also because they do not provide a realistic view into real-world muscle use and performance. Soft, flexible, sensorized garments could transform rehabilitation, but translational breakthroughs have been sparse, in part due to fundamental knowledge gaps on the relationship between novel sensing modalities and the physiological phenomena they are able to capture. We recently demonstrated that capacitive sensing, which can be easily embedded into e-textiles, can estimate muscle activity (muscle fiber length and bulging) and detect gait deviations. However, our prior work has been limited to healthy individuals. The proposed study will translate this technology to persons with ACL-R to determine if it can serve as a rehabilitation monitoring tool and investigate its mechanism of action.

NIH Research Projects · FY 2026 · 2026-04

ABSTRACT The Brain Image Library (BIL) serves as a central repository for advanced microscopy data, ensuring valuable neuroscience datasets are preserved and shared following FAIR principles. While BIL successfully handles traditional microscopy data, it currently lacks specialized tools for spatial transcriptomics datasets, which present unique computational challenges due to their high dimensionality and complex metadata structures. This project will develop essential computational infrastructure to make spatial transcriptomics data in BIL more accessible and analyzable for the broader neuroscience community. We will implement four key components: (1) standardization of data submission and storage using the community-supported Spatialdata format, enabling efficient handling of large-scale molecular data with coordinate system transformations and alignment capabilities, (2) development of an intuitive browser-based visualization system that allows researchers to explore gene expression patterns through interactive scatter plots and maps with density-based rendering, (3) creation of a natural language search interface leveraging Large Language Models for complex queries across molecular and anatomical parameters, and (4) integration of foundational models for automated cell type annotation and metadata generation. By standardizing data organization and providing intuitive exploration tools, this work will maximize the reuse potential of these valuable datasets and lay the groundwork for future cross-modality analyses that combine molecular and imaging data to generate new biological insights.

NSF Awards · FY 2026 · 2026-02

This award provides additional participant support for the first meeting of principal investigators in the NSF Artificial Intelligence, Formal Methods, and Mathematical Reasoning (AIMing) program, which will be held at the new NSF Institute for Computer-Aided Reasoning in Mathematics on April 2 and 3, 2026. The AIMing program seeks to support research at the interface of innovative computational and artificial intelligence (AI) technologies and new strategies/technologies in mathematical reasoning to automate knowledge discovery. Artificial intelligence for mathematics is a rapidly growing and extremely competitive field, especially in the US and China. The AI research community has come to realize that the only route to general intelligence is to combine the strengths of both neural and symbolic approaches to AI, specifically by developing AI systems capable of mathematical reasoning. AI that can reason with mathematical precision is crucial for developing safe, secure, and reliable hardware and software systems, making accurate financial and economic predictions, developing complex scientific and engineering models, and more. The meeting will foster collaboration and communication between the groups funded by the program, to support synergies to maximize the program's impact. This grant will support additional participants in the meeting, including students and collaborators of the PIs. The event website is available at https://icarm.io/event/nsf-aiming-pi-meeting/. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2026 · 2026-01

This project will support the continued access of U.S. researchers to the CLOUD consortium at CERN to study the chemistry and physics that drive new-particle formation and growth in Earth's atmosphere. CLOUD is a unique state-of-the-art 26 cubic meter stainless-steel chamber with precise control over temperature and relative humidity. There is no chamber facility within the USA with these capabilities. Because these particles influence cloud formation and how energy moves through the atmosphere, this research helps improve our understanding of weather behavior and atmospheric conditions. Experiments over the next three years using the CLOUD chamber will include a focus on developing molecular descriptions of new particle formation and advancing parameterizations for use in scientific models. These experiments will include mixtures relevant to the upper atmosphere, with sulfuric and nitric acids, ammonia and other bases, and iodine oxidation products. The CLOUD experiments often lead to advances in instrumentation and provide an ideal test bed for new instruments. Students and early career researchers will be integrated into an international collaborative network of students and senior researchers and have the opportunity to work with some of the most highly regarded scientists worldwide in the study of aerosols and atmospheric chemistry. 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 · 2025-12

Project Summary/Abstract Visually scanning the environment requires a series of rapid eye movements, known as saccades, interspersed with brief fixations, highlighting the remarkable speed of the brain’s oculomotor system. This raises fundamental questions about whether the brain plans each saccade sequentially or plans future movements before completing the current one. If the latter is true, this suggests that the brain must compute the amplitude and direction of the future movement while accounting for the first saccade before it is even initiated. This proposed research investigates the neural processing of the superior colliculus (SC) and frontal eye fields (FEF) – crucial nodes in the oculomotor circuit – during the generation of sequential saccades. The FEF, associated with cognitive processes like memory, and the SC, responsible for producing corollary discharge signals, communicate bidirectionally to compute saccade metrics. Using laminar multicontact probes, I will simultaneously record neural activity from the FEF and SC as rhesus macaques perform tasks designed in a 2x2 framework, involving single or two sequential saccades, in visually- or memory-guided domains. Advanced computational analyses will quantify the timing, directionality, and representation of neural communication between these regions, specifically testing the hypothesis that communication between the FEF and SC is enhanced during periods of increased cognitive processing (Aim 1). Furthermore, suprathreshold stimulation of the SC after prolonged fixation evokes a site-specific saccade, but stimulation immediately after a saccade does not produce this movement. The oculomotor system requires ~250 ms to reset after a saccade, aligning with typical intersaccadic intervals during visual scanning. I will apply electrical stimulation to the SC and FEF to test this phenomenon in tasks requiring sequential saccades. Specifically, this work will test the hypothesis that if sequential saccades are treated as a unified action, the resetting of the saccadic system should start with the second saccade (Aim 2). This proposed research will uncover the neural signatures associated with executing sequences of movements, providing fundamental insights into the neural mechanisms of sensorimotor integration. These findings may also inform treatments for conditions that disrupt eye movement control, such as neurodegenerative diseases and traumatic brain injury.

NSF Awards · FY 2025 · 2025-10

High frequency electromagnetic signals hold a significant promise for applications in communications, medical imaging, security, and more. Spintronics based technologies present an appealing route for high-frequency signal generation. For example, continuous rotational dynamics of the order parameter (magnetization or Néel vector) in magnetic systems driven by charge current, through the mechanisms of spin-orbit torque, can be used to realize conceptualized nano-oscillator devices for generating high frequency signals. Specifically, antiferromagnets based nano-oscillators, wherein the output signal frequency can range between gigahertz to terahertz range due to inherently ultra-fast rotational spin dynamics, has only recently been proposed and experimental demonstration of such a device remains critically missing. This research program aims at the first demonstration of prototype antiferromagnet based nano-oscillator devices by building a comprehensive understanding of spin dynamics in a new class of antiferromagnets and by providing unique strategies for electrically driving and detecting the broadband spin dynamic in antiferromagnets by exploiting unconventional form of spin current in topological semimetals. The project will impact society through the innovation to realize transformative high-frequency devices for next-generation technologies. This project will enable the training of next-generation researchers in the United States for workforce development. Through this project, a graduate student and undergraduate student(s) will be trained in experimental techniques, critical thinking, and solving complex research problems. Also, outreach events to kindle scientific interest and provide interactions and mentorship for middle school and high school students will be explored. Continuous rotational dynamics in antiferromagnets can enable next-generation device technologies, such as tunable and broadband sources and detectors of gigahertz to terahertz frequency signals. However, the understanding of dynamic phenomena in antiferromagnets, e.g., resonance modes, Néel vector precession, and magnon transport, which is essential to build antiferromagnet-based nano-oscillators, remains in its infancy. The discovery of magnetic order and unconventional topological spin current in two-dimensional systems provides a unique material platform to demonstrate antiferromagnet-based nano-oscillators. Relatively weak but highly tunable interlayer coupling in two-dimensional magnets, compared to traditional antiferromagnetic materials, provides a knob to control the magnetic state and response of the system, including accessing different dynamical regimes, and in turn, different frequency and response time ranges. Most importantly, to realize a prototype antiferromagnet-based nano-oscillators, one needs to experimentally demonstrate two key operations, i.e., charge current driven spin dynamics in antiferromagnets via spin-orbit torque mechanisms and subsequent broadband detection (electrical or other means) of spin dynamics in antiferromagnets in mesoscopic samples. This research program is aimed at demonstration of prototype antiferromagnet-based nano-oscillator devices and the specific scientific goals of this research program are threefold: (1) Characterization of magnetic order and dynamical response in mesoscopic two-dimensional antiferromagnetic systems; (2) Demonstration of antiferromagnetic spin dynamics driven by unconventional topological spin current; (3) Demonstration of antiferromagnet-based nano-oscillator device functionalities. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Nontechnical Description Gyroscopes are essential tools used in many technologies, such as navigation systems, drones, and camera stabilization. Thanks to advances in MicroElectroMechanical Systems (MEMS), mechanical gyroscopes have become much smaller and more efficient. However, for the highest levels of accuracy and stability, optical gyroscopes—like Ring Laser Gyroscopes (RLGs) and Fiber Optic Gyroscopes (FOGs)—are still the top choice. These devices use a physical principle called the Sagnac effect to detect rotation with incredible precision. Making these optical gyroscopes smaller could lead to major breakthroughs in low-power navigation systems for wearable devices and consumer electronics. One exciting direction is using integrated photonics—tiny optical circuits on a chip—to shrink these systems even further. A particularly innovative idea involves using something called exceptional points (EPs) in specially designed optical structures to boost sensitivity. This project explores a new type of photonic chip made from lithium tantalate thin films layered on silicon, known as the LTOI platform. The goal is to build a compact and stable optical gyroscope using this platform, controlled by electro-optic and acousto-optic modulators, and operating at these exceptional points. If successful, this new gyroscope could have a big impact on fields like navigation, aerospace, and virtual/augmented reality. Beyond that, the technology could also benefit areas like quantum computing, microwave communications, and advanced sensing—including gas detection, infrared imaging, and other motion-sensing devices. Technical Description This project explores the use of EPs in integrated coupled photonic cavities within the LTOI platform to enhance the scale factor (sensitivity) of a Sagnac gyroscope based on coupled optical racetrack resonators. The goal is to engineer methods to stabilize operation at EPs against environmental factors using high-performance acousto-optic and electro-optic modulators that can be directly integrated into the LTOI substrate. Vertical cavity surface-emitting lasers and photodetectors will be heterogeneously integrated on the same platform to create a very compact gyroscope. The ultimate technological objective is to develop a temperature-stable, miniaturized, and integrated Sagnac gyroscope with significantly improved size, weight, and power (SWaP) characteristics. We expect the proposed gyroscope to achieve an angle random walk of a few mdeg/√hr (comparable to the best state-of-the-art RLG) in a 100x reduced volume and 10x reduced power consumption compared to commercial RLGs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

There is a significant amount of scientific research that shows that intelligent tutoring systems (ITS) can help students learn better than through other forms of instruction and software. These systems give frequent, fine-grained guidance when students practice solving complex problems. ITS have proven to be useful in many domains, for example, STEM, business, and language learning. The primary goal of this project is to investigate whether and how Large Language Models (LLMs) can help instructors author ITS effectively and efficiently – a process that needs to be executed each time an ITS is created for a new content area. The project will provide insight into how LLMs can best assist in building ITS and in the efficiency gains that result. The project has the potential to help spread ITS widely into educational practice and thereby help many students learn better, from elementary school to college-level and even graduate school. To this end, the project will integrate existing LLMs with an existing set of authoring tools: the Cognitive Tutoring Authoring Tools (CTAT), which have long supported the efficient development of ITS, in two paradigms offering different capabilities. The project will assess whether and how LLMs can help with two time-consuming authoring tasks: generating a wide range of practice problems and developing rule-based cognitive models that capture the problem-solving knowledge that the ITS aims to help students learn. While assistance from an LLM is added, instructors remain in charge and contribute their specialized pedagogical content knowledge and skill in instructional design. The project will develop methods for authors to ensure that the generated content will be accessible and appealing to all students. The project will conduct evaluation studies to measure both the quality of the tutoring systems authored with LLM assistance and the gains in authoring efficiency that may result. This project is funded by the Research on Innovative Technologies for Enhanced Learning (RITEL) program that supports early-stage exploratory research in emerging technologies for teaching and learning. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Large Language Models (LLMs) show great promise for generating source code and automating programming tasks. But these models are error-prone and can produce code with subtle bugs. This poses a risk for deploying LLMs in industrial settings for software engineering tasks - the subtly erroneous code generated by LLMs can expose vulnerabilities that compromise system security. It has been shown that the weakness of LLMs for code generation primarily stems from not accounting for the semantic properties of programs when training, using, and evaluating these models. This project aims to improve LLMs’ ability to generate high-quality code by deeply integrating program analyses with all the stages in the life cycle of LLMs: training, code generation, and evaluation. This project develops novel quantitative program analyses techniques to provide feedback to LLMs during training and decoding. First, the project leverages symbolic execution and Bayesian program analyses to design meaningful metrics to evaluate LLM-generated code. This project then uses program scores to train a differentiable reward model that can assess the quality of partial or complete generated code. At training time, inspired by Reinforcement Learning with Human Feedback (RLHF), this project uses the reward model for fine-tuning LLMs to generate high-quality code. To improve code generation at decoding time, this project leverages the reward model and similarity-based program ranking techniques to constrain and prune the decoding tree. Finally, this project develops semantics-guided metrics and collects new benchmarks consisting of realistic coding tasks for training and evaluating code LLMs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Robots have come a long way in the past decade, but they still cannot reliably traverse or manipulate complex environments in the real world. This Faculty Early Career Development (CAREER) award supports research that looks to address two key reasons for this capability gap by developing 1) simulation tools that can efficiently and accurately capture relevant physics – including deformation, fluid interactions, and orbital dynamics – and 2) control methods that can reason about complex systems while simultaneously offering interpretability and safety guarantees. These advancements look to enable robots that can more efficiently, reliably, and safely interact with their environments and, therefore, bring us closer to a future in which robots are widely deployed to perform dangerous or tedious work. These robots could save lives by performing crucial tasks in dangerous environments instead of humans, and advance science by exploring places that are completely inaccessible to humans like deep oceans, space, and other planetary bodies. The education and outreach components of this award will also help inspire and recruit the next generation of scientists and engineers by directly engaging elementary and middle school students in future space missions. This project addresses two high-level technical goals. The first is to develop better modeling and simulation tools for situations in which sufficient data for reinforcement learning is too difficult or expensive to obtain on hardware. The focus, in particular, will be on multi-physics simulation for the emerging application areas of space and underwater robotics, where current simulation tools are lacking. The second addresses the need for computationally and data-efficient control methods that scale to high-dimensional inputs like vision and tactile sensors and exploit modern parallel computing hardware like graphics processing units (GPUs). To achieve this, the rich intersection between classical data-driven behavioral control – which offers interpretability and a rich set of system-theoretic analysis tools – and state-of-the-art diffusion policy methods from machine learning will be investigated. The education and outreach components of this project will help recruit and train the next generation of scientists and engineers by inspiring students to pursue careers in STEM, mentoring undergraduate student researchers as they enter the field of robotics, and training graduate students in cutting-edge optimization, dynamics, and control techniques. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

The objective of this NSF Civic Innovation Challenge (CIVIC) Stage 2 project is to explore and demonstrate an alternative model of rail transit development that revitalizes underutilized—or “legacy”—rail through partnerships. Rail transit in the United States faces several challenges that drive up costs and delay implementation: technical complexity requiring expensive signaling and inspection systems, complex regulatory hurdles, rigid funding structures that limit financial innovation, and reliance on costly feasibility studies to justify demand. These barriers reinforce high-cost, consultant-driven development models that discourage scalable, community-responsive transit. Rather than relying on billion-dollar rail projects, the project team asks: What if transit could start small, grow with demand, and be shaped by the people it serves? The project centers on how legacy rail can overcome barriers that inflate revitalization costs by combining modular rail technology, onboard sensing for safety and maintenance, new regulatory pathways, and grassroots planning. To achieve this, it brings together a partnership of eight organizations, led by Carnegie Mellon University, Pop-Up Metro, and the Delaware River Waterfront Corporation. Through this partnership, the project deploys and operates a full-scale metro along 1.61 miles of legacy rail along Philadelphia’s Delaware River. The demonstration showcases a scalable, cost-effective transit model, with replication in other communities supported through workshops for municipal leaders and transit agencies interested in revitalizing rail infrastructure. The outcome includes a public-facing blueprint outlining regulatory, technical, and financial strategies for replication, and a dedicated Pop-Up Metro division to guide implementation across the country’s thousands of miles of legacy rail. This project advances knowledge through an empirically rigorous, interdisciplinary approach to transform underutilized infrastructure into viable transit systems by combining engineering innovation, public engagement, and regulatory adaptation. It examines four key questions with significant knowledge contributions: (1) how onboard defect detection technology improves maintenance strategies and reduces costs; (2) how modular train units impact financial sustainability compared to fixed-infrastructure systems; (3) what regulatory pathways enable effective and scalable legacy rail reactivation; and (4) how grassroots engagement can replace traditional feasibility studies to inform needs, demand, and rail service justification. The research methodology combines rigorous technical validation, quantitative economic analysis, regulatory blueprint development, and mixed-methods community engagement analysis. A new paradigm is established for implementing transit by combining technical innovation with civic participation, creating empirically validated frameworks that can be applied across diverse urban and rural settings while deepening the fundamental understanding of infrastructure management. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Software testing is the predominant form of validating correctness for large real-world programs. Testing multi-threaded programs, where independent program tasks can execute concurrently and manipulate shared resources, remains challenging in practice since the sequence and ordering of interactions between multiple threads is hard to capture and replicate. This makes concurrency-related software bugs, called race conditions, very hard to detect and diagnose. If not identified proactively, concurrency bugs can manifest in production unexpectedly in disastrous ways, leading to potential human harm or loss of critical infrastructure. This project aims to develop principled techniques and tools for testing multi-threaded programs written in programming languages that depend on run-time management systems, such as, Java. Successful completion of this project will enable software engineers to perform reliable, efficient, and reproducible testing of large-scale concurrent applications to discover and eliminate race conditions. The project also includes synergistic educational activities, such as developing a debugging tool for novice programming classes and incorporating research findings into upper-division undergraduate computing courses. This project will develop a controlled concurrency testing solution for multi-threaded programs running alongside managed runtimes (e.g., Java Virtual Machine code), which will: (a) ensure a correct and efficient mechanism for deterministically controlling the scheduling of thread interactions; (b) provide support for systematically or randomly exploring thread schedules using state-of-the-art search strategies based on probabilistic or partial-order-based algorithms in order to uncover hard-to-find concurrency bugs; and (c) work for arbitrary programs off-the-shelf in a push-button fashion. A key pillar of this project is to focus on general-purpose applicability as a primary objective for research on concurrency testing platforms, alongside the traditional goals of performance and search-space optimization. The project will also enable the large-scale evaluation of modern search algorithms across a wide variety of real-world software that powers production-scale distributed systems and web applications. The project will develop open-source tooling for use by software engineers as well as for educational purposes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Cooperation of multiple devices to learn and make decisions based on their environment is especially valuable for Internet of Things (IoT) and other applications. Existing algorithms for cooperative learning often assume that all devices face the same set of decision choices, which is not often the case in networked system settings. This project, CoLeNe (Cooperative Learning in heterogeneous Networks), proposes to design and evaluate algorithms for multiple devices to cooperatively learn for decision making over a large set of choices in computer networks as the agents may face different sets of decisions. The developed algorithms allow devices to collaboratively explore the decision space and identify the best option from a large set faster. The project also applies these algorithms to the decision-making cases in networks and demonstrates their usefulness through the examples. In addition, the research effort is paired with educational and outreach initiatives that introduce students to the theory and practice of cooperative learning in networks. This proposal aims to develop cooperative online learning algorithms that are communication-efficient and robust to heterogeneity in computation, data, and privacy across agents. It consists of two research thrusts: (1) theoretical foundations and algorithms. The project will extend existing theoretical work on online learning to the settings of multiple heterogeneous agents with different sets of decision choices and privacy constraints on the information they can share with other agents. It develops algorithms for multiple devices to cooperatively learn to make optimal decisions. (2) Implementing two network applications. The developed cooperative online learning algorithms will be adapted to two applications in edge networks: distributed placement of computing applications across a network of devices, and optimization of wireless network configurations and transmission schedules. The learning framework developed through this project is expected to have broad applicability across IoT as well as other domains involving distributed, heterogeneous learning systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Inverse simulation is used to identify the needed inputs in order generate a desired outcome. It has found widespread usage across sciences and engineering. These include designing temperature-regulating components in aircraft, fabricating fluid systems, and detecting tumors in brain. Due to widespread utility of inverse simulations, there exists a large set of mature computational methodologies for inverse simulation tasks, predominantly based on taking continuous space and dividing into discrete sections, which is called a discretization approach. Unfortunately, the dependence on using discrete spaces in inverse simulations and the amount of time and computation it requires means that the existing methodologies are severely constrained. In an era of increasing demand for large-scale inverse simulation for a range of important applications (e.g., fabrication, medicine, robotics), there is a critical demand for methods that can overcome the existing bottlenecks. The project will address this challenge by developing a suite of Monte Carlo methods for inverse simulation that are highly scalable, parallelizable, output-sensitive, and significantly expand the applicable types of physical phenomena and representations. The project will achieve this goal through three inter-connected research thrusts: First, the project will research methods for Monte Carlo differentiable simulation, developing mathematical formulations and computational algorithms that can compute derivatives of solutions to partial differential equations (PDEs) with respect to arbitrary PDE parameters, without the need for discretization. Second, the project will research general and scalable material and geometry representations (e.g., volume representations, neural and point-based implicits) that lend themselves to inverse simulation problems. Third, the project will conduct evaluations through targeted applications, combining differentiable simulation, representations, and gradient-based optimization to solve inverse simulation problems in thermal design and electrical impedance tomography. This project will create new research areas both within computer graphics and at its intersection with other fields (mechanical engineering, medical imaging), and has the potential for transformative impact in all areas that use inverse simulation (e.g., fabrication, remote sensing, architecture, robotics, aviation, medicine), by developing new inverse simulation methods with greatly improved generality, robustness, and scalability. Finally, this project includes outreach activities synergistically with the above research activities. These activities will engage students in secondary education and undergraduate students, introduce them to computer graphics, and provide them with opportunities for research and hands-on development. These activities will promote greater participation in STEM, thus bringing long-term societal benefit. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

The goal of this project is to drastically simplify software development for intermittent systems by creating practical tools for modeling and analyzing the energy management and timing of intermittent software. Intermittent systems operate without a battery or tethered power supply and instead harvest energy from their environment, using motion, temperature gradients, and light. Eliminating the reliance on batteries improves robustness and reduces maintenance requirements, making intermittent systems suitable for deployment in harsh environments such as space or inside the body. Managing the interaction of energy collection and storage, intermittent operation, and the consequences of power failures are key challenges for the design of intermittent software. The project’s novelties are to address these challenges by using novel probabilistic modeling and analysis techniques to create serviceable timing and energy models for intermittent systems together with diagnosis tools that provide feedback during software development. The project's impacts are that energy and timing failures in intermittent systems can be diagnosed at development time and avoided without time-consuming trial-and-error experiments that run software on the target device. Moreover, the theoretical foundations of probabilistic programming for energy use in this domain enable a system designer to express, reason about, and validate energy consumption of mission critical embedded systems in domains like chip-scale satellites and in-body medical devices. The main technical innovation of the proposed work is to develop probabilistic programming as a tool for modeling and formally analyzing the energy use and execution time of intermittent systems. Probabilistic programming is an emerging technology that provides an efficient and flexible interface for Bayesian machine learning. Advantages of Bayesian learning include robustness and the ability to quantify uncertainty. For intermittent systems, probabilistic programs can serve as the link between a static analysis that is sound with respect to the formal semantics of probabilistic programs and a measurement-based energy model that can incorporate detailed domain knowledge about the hardware. The project focuses on the most challenging aspects of this approach and builds on the preliminary results and complementary strengths of the investigators in probabilistic programming, static analysis of probabilistic programs, intermittent systems, and the design and implementation of intermittent software. The main objectives are (1) developing and experimentally validating an effective methodology for creating probabilistic energy and timing models of different complexity and at different abstraction levels, (2) designing static analyses that automatically derive tail bounds on the computational costs as described by the models and to apply these bounds to provide global timing guarantees and (3) to apply and test the methodology in a cooperative scheduler and runtime system that uses probabilistic energy predictions to ensure fairness of energy consumption across tasks in an intermittent execution. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Non-technical Description: Ferroelectric wurtzites show great promise for enabling advanced communications technologies and for reducing computational energy consumption, both of which are key goals of the nation and the National Science Foundation. Their commercial deployment is hindered by limited understanding of the impacts of defect populations on properties, but current state-of-the-art computational techniques rely on unrealistic dilute-limit assumptions that ignore defect interactions with one another and/or with interfaces. This research aims to rigorously capture the interactions and effects of point defects such as heterovalent substitutions (e.g., oxygen replacing nitrogen) and extended defects (e.g., structural damage from bombardment during sputter growth) on properties in wurtzite nitrides. The team includes world experts in simulation, synthesis, characterization, and testing from the U.S. and Germany, and it includes partners from the Army Research Laboratory (ARL) and an industrial advisory board (IAB) who will build on relevant findings to accelerate scale-up and deployment as appropriate. The goal is to bridge the gap between calculations requiring simplifying assumptions and real films grown using commercial techniques to accelerate deployment of these and future DMREF-developed materials. Technical Description: To-date, when charged defects are simulated computationally (particularly within the electronic nitride space), they are assumed to be dilute and non-interacting, which is invalid for substitution levels in the several- to tens of atomic percent, such as those common in ferroelectric wurtzite alloys. This research will treat defects as components of complex alloys to capture disordered configurations as well as interactions of defects with one another and, eventually, with interfaces. Such calculations will be informed and validated by high resolution electron microscopy capable of measuring not only structural and chemical but also—via electron energy loss spectroscopy (EELS)—local bonding characteristics. The rigorous mechanistic understanding will enable predictive capabilities around interacting (non-dilute) point defects including heterovalent substitutions and will advance towards quantitative predictions of coercive fields, multiscale switching dynamics, and potential degradation processes important to the very devices that the ARL and industry partners on our team will be simultaneously advancing towards deployment. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Public sector peer-run behavioral health organizations (PROs) are essential providers supporting Americans facing tough times. These organizations offer holistic care that extends beyond mental and behavioral health to include services such as housing, employment support, and income assistance. However, many PROs operate with limited staffing and outdated infrastructure, making it difficult to meet rising service demands. While large healthcare systems increasingly use Artificial Intelligence (AI) to improve efficiency, peer-led organizations have had limited access to AI tools built specifically for their service models and day-to-day challenges. This project addresses that gap through a partnership with the Collaborative Support Programs of New Jersey (CSPNJ), a statewide, peer-run behavioral health organization. Together, we are developing PeerCoPilot, an AI-powered assistant designed to help peer providers deliver faster, more tailored, and more consistent support to service users, without replacing the human connection at the core of peer work. The project is designed to scale across similar peer-run organizations, creating a model for efficient technology integration in the safety-net behavioral health sector across the US. This project applies a community-centered AI design framework to develop and evaluate PeerCoPilot, an AI assistant for peer providers in high-need behavioral health settings. During our Stage 1 planning grant, we co-designed and tested the first component of PeerCoPilot – the Wellness Planner, a large language model (LLM)-powered tool that helps peer providers generate personalized wellness plans and locate relevant service resources. Initial testing showed the tool was easy to use and provided meaningful support, while also revealing key areas for improvement in information accuracy and follow-up management. In Stage 2, we will enhance PeerCoPilot by (1) improving the reliability of the Wellness Planner using techniques like retrieval-augmented generation and in-context learning, and (2) adding a new Wellness Check-in Management feature powered by LLMs and sequential decision-making methods (e.g., restless multi-armed bandits) to support timely, personalized follow-up. These features are aimed at increasing provider capacity and reducing missed opportunities for engagement. To support adoption and readiness, we will also co-develop onboarding guides, brief tutorials, and embedded support tools for peer providers. The project integrates expertise in AI, HCI, behavioral health, and social work, with the goal of building a scalable, efficient AI system that enhances the reach and quality of peer-led behavioral health services. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Radio spectrum is fundamental resource for wireless communication. As a result, the ability to allow radios to flexibly use different parts of the wireless spectrum is key to managing this scarce resource efficiently. Indeed, the original vision of a software defined radio is one in which a radio’s center frequency is tunable, as is the precise signal waveform that is to be transmitted. However, a pivotal element is typically not as flexible to program – the antenna. Indeed, most programmable radios today either require mechanical or electronic switching between antennas, or wide-band antennas whose performance is poor especially if form factor is constrained. This project develops Softenna, a first-of-its kind soft-robotic highly re-configurable antenna platform that dynamically adapts its RF properties, including center frequency, beam pattern, directionality and polarization. Softenna achieves this through a combination of mechanical and electronic re-configuration. Softenna combines innovation in wireless systems and soft robotics, enabling curriculum development at the university and K-12 levels that brings together both fields. Softenna is composed of multiple elements fabricated using stretchable and flexible materials and liquid metal. Softenna is designed to offer rich shape changes and a learning-based pipeline that studies which shape pattern is best suited to a given operating frequency, device location and environment. The system will be fully implemented on soft robotic antennas integrated with software defined radios operating in the sub-6 GHz frequency bands and evaluated these spectral bands. The project’s key contributions include: (1) The design of the Softenna platforms that enable rich variations in operating frequency; (2) Accompanying algorithms support learning and programming the behavior of these platforms to best suit any given environment. (3) A comprehensive implementation and study of the system through experimentation in varied testbeds, both indoors and outdoors. As part of educational and outreach efforts, the investigators will develop a workshop module where high-school students program wireless radios via Carnegie Mellon University’s Spark Saturday program and integrate findings in university-level courses in wireless systems and robotics. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

The human sense of touch is incredibly powerful and vital aspect of human experience. However, touch is still not well studied. The sense of touch allows one to move about, as well as identify, analyze properties of and manipulate objects. Touch complements other senses and is a critical part of a person’s ability to identify fine features, develop spatial awareness, learn, and form memories. Touch enables people to combine information across different senses and to form emotional experiences. However, most digital experiences have yet to include touch. This project will use a novel approach to enable images that combine sight and touch through detailed, dynamic, and rich haptic displays built from recently developed soft materials that significantly respond to changes in their settings. Including rich haptic experiences will transform human-computer interaction, much like the giant leap forward when computers began presenting graphics on screens. Enabling users to "feel" digital graphics has the potential to redefine experiences across a wide range of applications, including education, entertainment, and product design. Haptic interfaces offer access to visual information when users must look elsewhere. They also introduce interactions the combine touch with other senses to enhance tasks like operating machines from a distance, providing a person with the sense of being present elsewhere, and exploring complex scientific data or medical images. A key hurdle that has prevented people from living in this multimedia digital world is the lack of actuation technology specifically designed for tactile haptics, thereby limiting the palette of haptic interactions available to designers. Most work which incorporates tactile haptics in computation use vibration motors, which do not take advantage of the wide range of possible sensations and receptors in people’s skin. This project will make use of the enormous amount of innovation that has occurred in stimuli-responsive soft materials, and create high-resolution, dynamic and rich tactile digital graphical content. The investigators will develop novel tactile haptic technologies for digital illustrations through an iterative process in which haptic perception tests and functional evaluations will inform materials specifications based on three main research objectives: (1) Identify the potential of stimuli-responsive soft materials to create haptic pixels to render black and white graphical information to users through touch. (2) Incorporating dynamic material response into tactile pixels to expand the capabilities of rendered graphical content. (3) Augmenting the richness of graphical content rendered for users through an increase of the dynamic mechanical capabilities of tactile pixels. Through this proposed work, the investigators will bridge the gap between materials, haptics, and human-computer interaction research to create new tactile digital expressions. These advances will enable and accelerate new user interfaces and interactions through this underutilized information channel. 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 · 2025-09

PROJECT SUMMARY/ABSTRACT Congenital heart disease (CHD) affects over 12 million people globally and contributes to over 260,000 deaths each year. Among the most severe forms of CHD are single-ventricle heart defects, characterized by an underdeveloped or absent ventricle. Palliative repair for single-ventricle cardiac anomalies consists of a staged surgical approach culminating in the Fontan procedure, wherein the functioning ventricle is committed to supporting systemic circulation while venous return is routed passively to the lungs through a synthetic extracardiac graft without direct pump support. Although the Fontan procedure facilitates initial survival, patients experience significant long-term comorbidities due to the absence of a subpulmonary ventricular pump, ultimately leading to heart, liver, or other organ failure. One potential solution to this problem is to engineer a contractile cardiac conduit to replace the passive synthetic extracardiac graft, providing a second ventricular pump that can shift the univentricular Fontan circulation towards a more stable biventricular physiology. The Feinberg lab has recently demonstrated the use of Freeform Reversible Embedding of Suspended Hydrogels (FRESH) 3D bioprinting to produce cardiac pumps with synchronous contraction and pressure generation; however, these pumps have not achieved the contractile pressure generation necessary to serve as a substitute ventricle in the Fontan circulation. This limitation is likely due to the lack of a helical myocardial architecture, which has been shown in the adult ventricle to drive systolic ejection fraction through ventricular torsion and wall thickening. I hypothesize that cardiac conduits bioprinted with biomimetic helical architecture will demonstrate improved contractility and pressure generation similar to the native right ventricle (>15mmHg). In Aim 1, I will develop a novel multi-axis nonplanar printing platform that provides additional degrees of movement, allowing for precise deposition and alignment of material in biomimetic helical architectures. I will validate this technology by fabricating collagen conduits with varying circumferential and helical alignments and performing mechanical characterization to evaluate how fiber orientation affects biomechanical properties. In Aim 2, I will bioprint a contractile cardiac conduit with ventricular helical alignment and assess the construct for functional performance and contractility. The findings of this proposal will establish a scalable platform for engineering contractile cardiac tissues with native-like architecture and function, laying the foundation for future efforts to fabricate organ-scale cardiac constructs for organ transplantation.

NSF Awards · FY 2025 · 2025-09

This Smart and Connected Communities Integrative Research Grant (SCC-IRG) project supports research that aims to develop systems that help robots navigate public areas in a safe and socially acceptable manner. While robots have great potential to improve communities by providing services such as delivery, safety patrol, and sanitation, they are still limited in their ability to autonomously manage the complexities of real-world public walkways, which often feature sprawling obstacles, unmaintained sidewalks, and varying weather conditions. Robots may also find it challenging to navigate among people who are on their phones, not paying attention, using a cane or a walker, pushing a stroller, or walking their dog. This can lead to unsafe interactions and hinder the public acceptance necessary to realize the full potential of robotic services. However, robots cannot simply follow hard traffic rules and must learn to adapt and adjust their behavior based on the scenarios they encounter. This project will engage directly with community members to collect data and train new algorithms on how people want robots to behave in public. Intended outputs of this research are improved ways for robots to navigate in public and communicate with community members. By developing socially appropriate navigation and communication for these robots, this research seeks to create safer and more effective robotic systems. This work serves the national interest by advancing scientific progress in human-robot interaction, enhancing public welfare and safety on community walkways, and supporting the nation's leadership in developing and deploying civic robotic technology with direct citizen engagement, thereby enhancing the economic potential of public robots and improving community quality of life. Developing safe and socially appropriate behaviors for public-area robots requires intelligent methods for integrating human feedback from realistic scenarios into the robot's learning process. It also requires new design patterns that enable robots and people to communicate effectively when on the sidewalk. This project aims to enhance robot social navigation and communication on public walkways through continuous community engagement, including co-designing interaction scenarios, refining pedestrian models and sidewalk simulations, and training robot navigation and communication systems. The project will employ a phased development process that combines 3D simulation environments, laboratory studies, and field tests. A key technical approach involves using a novel machine-learning framework where community members will participate in pilot demonstrations in both virtual and physical settings to train the robot's navigation and communication systems. This method seeks to enable efficient training of socially aware behaviors and design patterns that can be transferred across different types of robots. The performance of the resulting robotic systems will be rigorously evaluated through laboratory studies and field tests within the community, using both wheeled and legged robots to test the generalizability of the learned behaviors. The Lawrenceville Corporation, a local community organization in Pittsburgh, and Carnegie Mellon's Metro21 Smart Cities Institute will help facilitate engagement with community members, grounding the project in real-world complexities and connecting it to the everyday experiences of residents. The project will also partner with the Urban Robotics Foundation, which will contribute its extensive knowledge of urban robot systems and support the research team in translating findings into policy and technical standards recommendations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

This I-Corps project is based on the development of a medical device designed to diagnose and treat muscle-wasting disorders. Muscle-wasting disorders affect millions of individuals globally, leading to decreased quality of life, loss of mobility, and increased healthcare costs. These disorders can result from aging, chronic diseases, traumatic injuries, or prolonged inactivity. Current solutions primarily focus on symptom management rather than addressing the underlying loss of muscle function and structure. This technology is based on electrical stimulation combined with diagnostic capabilities to effectively restore muscle health and functionality. Directly addressing muscle degeneration and promoting muscle regeneration may improve patient outcomes and quality of life. In addition, this device may reduce healthcare costs by providing an effective alternative to existing treatments, while enhancing patient independence. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of an advanced bioelectronic platform employing integrated therapeutic stimulation and diagnostic monitoring technologies. The core technology is based on a customizable, wearable bioelectronic interface capable of delivering targeted neuromuscular electrical stimulation to promote muscle regeneration and strength recovery. Unlike existing therapies, this technology uniquely incorporates real-time diagnostic sensors to continuously monitor muscle health, allowing tailored, patient-specific therapeutic interventions. The technology uses platinum nanoparticles and conductive polymer coatings to enhance electrode performance, biocompatibility, and signal fidelity, providing advantages over conventional electrode systems. The device may be used to quantitatively detect and reverse muscle atrophy, offering a cost-effective solution to patients and a biometrics companion to clinicians. Clinicians and patients benefit from improved therapeutic efficacy, personalized treatments, and reduced recovery times. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Fatigue refers to the mechanical failure of materials subjected to repeated cyclic loads. It occurs in all materials including metals and alloys, and is a significant limitation that affects all engineering structures and devices (both structural and functional). It is a common cause of failure (and accidents) in devices including computers, cars, bridges and airplanes, and thus a significant economic, societal and national defense challenge. Unfortunately many fundamental aspects of fatigue remain incompletely understood, even though decades of empirical knowledge have provided (conservative) material specific guidelines for the design of engineering components to avoid fatigue in specific materials. The advent of additive manufacturing and advanced alloys have pushed us beyond this empirical knowledge base. In particular, recent experimental observations suggest situations where additively manufactured metallic alloys have fatigue strength that greatly exceed their conventionally manufactured counterparts, and other situations where they greatly underperform. The investigators will develop a new data-driven approach predicated on the view that the spread of fatigue life can be attributed to particular details of defects and microstructure, and a fundamental understanding of this relationship can lead to new Structural Alloys for Fatigue Endurance (SAFE). The approach is to create an integrated database and knowledge map of material and processing parameters, microstructure, comprehensive mechanical characterization, post-failure analysis and computational experiments on the fatigue behavior of additively manufactured structural alloys. Innovations in methodology including efficient high throughput testing and serial sectioning combined with Electron Backscatter Diffraction (EBSD) to acquire three-dimensional images of microstructure with orientation around individual crack initiation sites, in operando synchrotron observation and accelerated approaches to simulation enable the creation of this database. New methodologies are developed to sample the knowledge map and add new experimental results and simulation until there is a significant area of knowledge where the control parameters predict the quantities of interest. This leads to a deep fundamental understanding of the exact mechanisms and features that initiate, inhibit and accelerate fatigue cracks, and subsequently to the inverse problem of designing new SAFE materials. The approach is developed with a focus on the widely used aerospace alloy Ti-6Al-4V. The team consists of experts in additive manufacturing, texture & characterization, fatigue, computational modeling and the application of machine learning to 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.

NSF Awards · FY 2025 · 2025-09

This I-Corps project is based on the development of an artificial intelligence (AI) framework for analyzing numeric and tabular data. Numeric data plays a crucial role in business operations and decision-making, whether financial data, supply chain data, consumer metrics, or patient health data. Currently, data professionals want easy-to-use tools for numeric and tabular data that have a simple workflow and integrate seamlessly into their existing processes, while providing the flexibility and power required for complex data analysis. There is a growing demand for AI-assisted advanced analytics and decision-making, however, advancements in AI have made limited progress with numeric and tabular data. This technology is designed to improve the efficiency and productivity of data professionals. The goal is to improve access to advanced analytical tools and bolster public engagement with AI while helping ease the impact of skilled labor shortages. The technology is industry-agnostic and may help to create new products, generate jobs, and enhance economic activity. The data analysis framework also may address systemic inefficiencies and improve practices related to analyzing numeric and tabular data. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of using an artificial intelligence (AI) technology for numeric data analysis. The rapid advancements in large language models (LLMs) have sparked enthusiasm for AI, but extending these advances to numeric and tabular data has had comparatively limited progress. The technology uses a new class of machine learning (ML) pipelines based on a novel hierarchical neuro-symbolic architecture. Classical as well as deep neural models require iterative rework in manual and ad hoc tuning of their features or model architectures. These challenges are addressed using a hierarchical architecture that is interpretable, can use domain knowledge, and is simplified. This solution may provide data professionals with an easy-to-use tool for numeric and tabular data while giving them the flexibility and power required for complex data analysis. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

This project uses advances in artificial intelligence, computer science, and statistics to develop the GeoRDyn toolkit to allow researchers to identify the spatial and relational complexities in more fine-grained data. This will permit new approaches to investigating longstanding questions in the field, such as if and how networks emerge and influence collective action and the conditions under which events diffuse across time and space. Events are often observed at different temporal and spatial scales, involving complex interactions among actors at the micro, macro, and meso levels. Methodological tools commonly used in the study of collective-action events are often less dynamic and/or do not account for complex dependencies. This project develops the GeoRDyn toolkit to help researchers dissect the spatial and relational complexities in events and draw rigorous causal inferences. The toolkit incorporates recent advances in artificial intelligence (AI), computer science, and statistics. The project develops new methods and software for studying complex and dynamic processes, such as diffusion, that can be used by researchers as well as those outside of the academic community who need to understand the causes and potential implications of events. 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.