Search NASA⌕ Search

SEARCH · Search NASA

Results for “Scalable Model”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 559 records · Page 31
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Accelerating Multivariate Functional Approximation Computation with Domain Decomposition Techniques⋆

Modeling large datasets through Multivariate Functional Approximations (MFA) provide an elegant way to handle many visualization and scientific analysis workflows. The process necessitates scalable data partitioning methods to compute MFA representations efficiently without compromising the accuracy or continuity of the reconstructed solution. We propose a domain -decomposed method for computing the MFA with B -spline bases, which reduces the total work per task and uses a restricted Additive Schwarz (RAS) method to converge the control point data degrees -of -freedom along subdomain boundaries. We provide an in-depth analysis of the parallel approach with domain decomposition solvers, aiming to minimize local subdomain error residuals and recover high -order continuity at subdomain interfaces with appropriate choices of knot overlaps. The communication cost, determined by the overlap regions in the RAS implementation, is optimized to recover the numerical error profile of the single subdomain case. Our proposed method stands in contrast to previous methods, which typically only recover either C 0 or at best C 1 continuity for arbitrary B -spline degree expansions, or those that require post -processing to blend discontinuities in the reconstructed data. We demonstrate the effectiveness of our approach using analytical and real -world datasets in 1D, 2D, and 3D through both strong and weak scaling studies. The performance results indicate that the overall cost of computing the approximation is directly proportional to the underlying nearest -neighbor communication implementation, and is only weakly dependent on the overlap region size that determines the size of the messages. This finding underscores the efficiency and scalability of our proposed method, making it a promising solution for handling large datasets in scientific workflows.

additive Schwarz solvers↗
DOE OSTI2025

Remote Americium Detection Using an Optical Sensor: A D-Optimal Strategy for Efficient PLS-Based Modeling

A fiber-optic visible–near-infrared absorption spectroscopy system in a glove box was demonstrated for remote quantification of Am(III) (0–500 µM) and HNO 3 (0.1–9 M) using partial least squares regression (PLSR) models. The sensor platform, featuring a simple plug-and-play spectrophotometer, can enable noninvasive, real-time monitoring of actinide process solutions. To establish a flexible PLSR model calibration strategy, a D-optimal design developed using Nd(III) in previous studies was successfully extended to an actinide system with Am(III) to effectively minimize sample set size while maintaining robust prediction performance. The results suggest strong spectral similarities between Nd(III) and Am(III) and validate Nd(III) as an effective optical surrogate for trivalent actinide species. This work also supports the generalizability of a D-optimal training set selection approach for two-factor systems. The PLS1 models for Am(III) and HNO 3 outperformed a PLS2 model and maintained reasonable performance in the presence of interfering U(VI). The resulting sensor system and multivariate approach provides a flexible and scalable solution for process monitoring, control, and safety in diverse nuclear applications.

actinide↗
NASA NTRS1991

The Jade File System

File systems have long been the most important and most widely used form of shared permanent storage. File systems in traditional time-sharing systems, such as Unix, support a coherent sharing model for multiple users. Distributed file systems implement this sharing model in local area networks. However, most distributed file systems fail to scale from local area networks to an internet. Four characteristics of scalability were recognized: size, wide area, autonomy, and heterogeneity. Owing to size and wide area, techniques such as broadcasting, central control, and central resources, which are widely adopted by local area network file systems, are not adequate for an internet file system. An internet file system must also support the notion of autonomy because an internet is made up by a collection of independent organizations. Finally, heterogeneity is the nature of an internet file system, not only because of its size, but also because of the autonomy of the organizations in an internet. The Jade File System, which provides a uniform way to name and access files in the internet environment, is presented. Jade is a logical system that integrates a heterogeneous collection of existing file systems, where heterogeneous means that the underlying file systems support different file access protocols. Because of autonomy, Jade is designed under the restriction that the underlying file systems may not be modified. In order to avoid the complexity of maintaining an internet-wide, global name space, Jade permits each user to define a private name space. In Jade's design, we pay careful attention to avoiding unnecessary network messages between clients and file servers in order to achieve acceptable performance. Jade's name space supports two novel features: (1) it allows multiple file systems to be mounted under one direction; and (2) it permits one logical name space to mount other logical name spaces. A prototype of Jade was implemented to examine and validate its design. The prototype consists of interfaces to the Unix File System, the Sun Network File System, and the File Transfer Protocol.

Rao, Herman Chung-Hwa↗
NASA NTRS2016

End-to-End Assessment of a Large Aperture Segmented Ultraviolet Optical Infrared (UVOIR) Telescope Architecture

Key challenges of a future large aperture, segmented Ultraviolet Optical Infrared (UVOIR) Telescope capable of performing a spectroscopic survey of hundreds of Exoplanets will be sufficient stability to achieve 10^-10 contrast measurements and sufficient throughput and sensitivity for high yield Exo-Earth spectroscopic detection. Our team has collectively assessed an optimized end to end architecture including a high throughput coronagraph capable of working with a segmented telescope, a cost-effective and heritage based stable segmented telescope, a control architecture that minimizes the amount of new technologies, and an Exo-Earth yield assessment to evaluate potential performance. These efforts are combined through integrated modeling, coronagraph evaluations, and Exo-Earth yield calculations to assess the potential performance of the selected architecture. In addition, we discusses the scalability of this architecture to larger apertures and the technological tall poles to enabling it.

Feinberg, Lee↗
NASA NTRS2001

The Computational Complexity, Parallel Scalability, and Performance of Atmospheric Data Assimilation Algorithms

The computational complexity of algorithms for Four Dimensional Data Assimilation (4DDA) at NASA's Data Assimilation Office (DAO) is discussed. In 4DDA, observations are assimilated with the output of a dynamical model to generate best-estimates of the states of the system. It is thus a mapping problem, whereby scattered observations are converted into regular accurate maps of wind, temperature, moisture and other variables. The DAO is developing and using 4DDA algorithms that provide these datasets, or analyses, in support of Earth System Science research. Two large-scale algorithms are discussed. The first approach, the Goddard Earth Observing System Data Assimilation System (GEOS DAS), uses an atmospheric general circulation model (GCM) and an observation-space based analysis system, the Physical-space Statistical Analysis System (PSAS). GEOS DAS is very similar to global meteorological weather forecasting data assimilation systems, but is used at NASA for climate research. Systems of this size typically run at between 1 and 20 gigaflop/s. The second approach, the Kalman filter, uses a more consistent algorithm to determine the forecast error covariance matrix than does GEOS DAS. For atmospheric assimilation, the gridded dynamical fields typically have More than 10(exp 6) variables, therefore the full error covariance matrix may be in excess of a teraword. For the Kalman filter this problem can easily scale to petaflop/s proportions. We discuss the computational complexity of GEOS DAS and our implementation of the Kalman filter. We also discuss and quantify some of the technical issues and limitations in developing efficient, in terms of wall clock time, and scalable parallel implementations of the algorithms.

Lyster, Peter M.↗
DOE OSTI2024

A Perspective on Scalable AI on High-Performance Computing and Leadership Class Supercomputing Facilities [Industrial and Governmental Activities]

Many scientific applications that support the mission of the US Department of Energy (US-DoE) require modeling complex engineering and/or physical systems. Here, examples of such complex systems arise from: (a) materials science to develop new compounds with exceptional mechanical and thermodynamical properties (e.g., resistance to mechanical stresses and high temperatures), (b) structural and nuclear engineering to model the temporal evolution of the structural damage of concrete shields exposed to continuous neutron and gamma radiations emitted by the nuclear reactor core, (c) urban sciences (e.g., transportation and smart buildings), and (d) power grid systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Higher Efficiency, Demand Flexible Refrigerator with On-Demand Micro-Vibrational De-icing Technology

Refrigerator technology has advanced significantly over the last couple of decades. Today’s refrigerators use only about 25% of the energy that was required to power models built in 1975. Even as they continually improve efficiency to meet standards, refrigerators have increased in size by almost 20%, added energy-consuming features such as through-the-door ice, and provide more benefits than ever before. However, a few challenges and technology gaps are preventing further improvement of the demand responsiveness and efficiency of the refrigerators. One of the major technology gaps in existing refrigerators is their outdated de-icing process. When the evaporator generates frost, an old-fashioned resistive heating element melts the ice. Most refrigerators have a timed defrost cycle, rather than an active system that could monitor the state of the frost. In these systems, not only is the precious electricity used at its least efficient form of conversion (direct conversion of electricity to heat), but also all the latent heat associated with the ice is wasted during the melting process. On top of that, the refrigerator needs to work harder to pull the temperature down after defrosting, and, last but not least, the food quality is severely impacted by the temperature swings during the defrost cycle. According to a study, the EU alone wastes 89 million tons of food in the supply chain every year. Any temperature swing during defrosting (about 6F according to Emerson for low-temperature cases) can negatively impact the shelf life of meat and other products for multiple days. All these issues can happen during the peak demand time of the electric grid. Unlike the conventional systems, the proposed novel advanced micro-vibrational deicing process uses no heat for defrosting. Instead, it uses the micro vibrations generated by a piezoelectric or vibration-generating module to mechanically break ice from the heat exchanger almost instantaneously. The project titled “Higher Efficiency, Demand Flexible Refrigerator with On-Demand Micro-Vibrational De-icing Technology, performed by Ultrasonic Technology Solutions, LLC (UTS) of Knoxville, TN, in collaboration with Emerson (now Copeland), represents the final phase of a multi-year effort funded under the U.S. Department of Energy’s Building Technologies Office (BTO) BENEFIT FOA 2020. Initiated on October 1, 2021, and completed after a nine-month no-cost extension ending September 30, 2025, this project aimed to develop and validate a novel micro-vibrational mechanical defrosting system, achieving more than 25% improvement in defrosting energy efficiency over conventional baseline defrosting technologies. Over sixteen quarters, the project advanced from fundamental ice-mechanical characterization and prototype development to full-scale system integration and validation. Initial efforts established project management infrastructure and characterized ice adhesion properties, followed by the design and fabrication of early aluminum-based prototypes for resonance frequency testing. Subsequent quarters saw rapid technical progression, including the identification of optimal piezoelectric and motor-based vibration mechanisms, the demonstration of effective de-icing over 6x6-inch aluminum surfaces. The team achieved its Go/No-Go milestone by exceeding the 25% energy-efficiency improvement target—reaching up to 3,340% under optimized conditions—and later confirmed that motor-driven systems offered superior performance and energy efficiency compared to piezoelectric alternatives. Continued refinement led to the development of amplifier systems on printed circuit boards, improved control and instrumentation hardware, and integration into full-scale heat exchanger (HX) prototypes at both UTS and Copeland facilities. Multiple vibration-mounting studies and frost-growth experiments guided mechanical optimization and noise-mitigation strategies, achieving a 17.5 dB reduction in sound pressure level and verifying robust mechanical performance. Advanced analyses, including modal and harmonic simulations, established a quantitative understanding of vibrational behavior and de-icing efficiency across >1000 cm² systems. The final project phase successfully demonstrated scalable integration within reach-in and chest freezer prototypes, confirmed >25% efficiency improvements in large-area systems, and completed a comprehensive business model and scale-up strategy identifying electric defrost systems as the primary beachhead market. The culmination of this DOE-supported effort establishes micro-vibrational defrosting as a viable, high-efficiency, low-noise, and demand-flexible de-icing technology, paving the way for commercial deployment and broader application in next-generation refrigeration systems.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

mzPeak: Designing a Scalable, Interoperable, and Future-Ready Mass Spectrometry Data Format

Advances in mass spectrometry (MS) instrumentation, such as higher resolution, faster scan speeds, and improved sensitivity, have significantly increased the volume and complexity of data. The growing adoption of imaging and ion mobility further amplifies these challenges across MS-based omics fields, including proteomics, metabolomics, and lipidomics. While these technologies unlock new possibilities, they also present significant challenges in data management, storage, and accessibility. Existing open formats, such as the XML-based community standards mzML and imzML, struggle to meet the demands of modern MS workflows due to their large file sizes, slow data access, and limited metadata support. Vendor-specific formats, while optimized for proprietary instruments, lack interoperability, comprehensive metadata support and long-term archival reliability. This white paper lays the groundwork for mzPeak, a next-generation community data format designed to address these challenges and support high-throughput, multi-dimensional MS workflows. By adopting a hybrid model that combines efficient binary storage for numerical data and both human and machine-readable metadata storage, mzPeak will reduce file sizes, accelerate data access, and offer a scalable, adaptable solution for evolving MS technologies. For researchers, mzPeak will enable enhanced interoperability across platforms, seamless support for complex workflows including ion mobility and MS imaging, and faster data access compared to existing community formats such as mzML. Its design will ensure data is managed in compliance with regulatory standards, essential for applications such as precision medicine and chemical safety, where long-term data integrity and accessibility are critical. For vendors, mzPeak provides a streamlined, open alternative to proprietary formats, reducing the burden of regulatory compliance while aligning with the industry's push for transparency and standardization. By offering a high-performance, interoperable solution, mzPeak positions vendors to meet customer demands for sustainable data management tools which will be able to handle emerging and future data types and workflows. mzPeak aspires to become the cornerstone of MS data management, empowering researchers, vendors, and developers to innovate and collaborate more effectively.

data formats↗
DOE OSTI2024

chatHPC: Empowering HPC users with large language models

The ever-growing number of pre-trained large language models (LLMs) across scientific domains presents a challenge for application developers. While these models offer vast potential, fine-tuning them with custom data, aligning them for specific tasks, and evaluating their performance remain crucial steps for effective utilization. However, applying these techniques to models with tens of billions of parameters can take days or even weeks on modern workstations, making the cumulative cost of model comparison and evaluation a significant barrier to LLM-based application development. To address this challenge, we introduce an end-to-end pipeline specifically designed for building conversational and programmable AI agents on high performance computing (HPC) platforms. Our comprehensive pipeline encompasses: model pre-training, fine-tuning, web and API service deployment, along with crucial evaluations for lexical coherence, semantic accuracy, hallucination detection, and privacy considerations. Here, we demonstrate our pipeline through the development of chatHPC, a chatbot for HPC question answering and script generation. Leveraging our scalable pipeline, we achieve end-to-end LLM alignment in under an hour on the Frontier supercomputer. We propose a novel self-improved, self-instruction method for instruction set generation, investigate scaling and fine-tuning strategies, and conduct a systematic evaluation of model performance. The established practices within chatHPC will serve as a valuable guidance for future LLM-based application development on HPC platforms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

A generative machine learning model for designing metal hydrides applied to hydrogen storage

Developing new metal hydrides is a critical step toward efficient hydrogen storage in carbon-neutral energy systems. However, existing materials databases, such as the Materials Project, contain a limited number of well-characterized hydrides, which constrains the discovery of optimal candidates. This work presents a framework that integrates causal discovery with a lightweight generative machine learning model to generate novel metal hydride candidates that may not exist in current databases. Using a dataset of 450 samples (270 training, 90 validation, and 90 testing), the model generates 1000 candidates. After ranking and filtering, six previously unreported chemical formulas and crystal structures are identified, four of which are validated by density functional theory simulations and show strong potential for future experimental investigation. Overall, the proposed framework provides a scalable and time-efficient approach for expanding hydrogen storage datasets and accelerating materials discovery.

generative model↗
DOE OSTI2024

Exploring the cost and emissions impacts, feasibility and scalability of battery electric ships

The United States’ greenhouse gas (GHG) emissions reduction goals, along with targets set by the International Maritime Organization, create an opportunity for battery electric shipping. In this study, we model life-cycle costs and GHG emissions from shipping electrification, leveraging ship activity datasets from across the United States in 2021. We estimate that retrofitting 6,323 domestic ships under 1,000 gross tonnage to battery electric vessels would reduce US domestic shipping GHG emissions by up to 73% by 2035 from 2022 levels. By 2035, electrifying up to 85% of these ships could become cost effective versus internal combustion engine ships if they cover 99% of annual trips and charge from a deeply decarbonized grid. We find that charging demands from electrifying these ships could be concentrated at just 20 of 150 major ports nationwide. This study demonstrates that retrofitting to battery electric vessels has economic potential and could significantly accelerate GHG emission reductions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

NanoPSD: A software for automatic detection of Nano-Particle Shape Distribution in electron microscopy images

Accurate quantification of the size and morphology of nanoparticles from electron microscopy (EM) images is essential to understand growth mechanisms, surface reactivity, and functional behavior in nanoscale materials. Manual analysis remains slow, subjective, and difficult to reproduce in large datasets. We introduce NanoPSD (Nano-Particle Shape Distribution), an open-source and fully automated framework for quantitative particle detection and morphology analysis from EM images. NanoPSD integrates adaptive contrast enhancement, polarity-agnostic scale-bar detection, Optical Character Recognition (OCR)-based calibration, and classical segmentation via Otsu thresholding with morphological refinement. Particle contours are used to extract geometric descriptors, including equivalent circular diameter, aspect ratio, circularity, and solidity, enabling automated classification into spherical, rod-like, and aggregate morphologies. The framework supports both single-image and batch processing, generating publication-quality visualizations, LaTeX-ready tables, and structured comma-separated values (CSV) datasets. As a demonstration, we applied NanoPSD to plasma-synthesized nanoparticle samples diagnosed via transmission electron microscopy (TEM). The code produced statistically robust size and morphology distributions spanning a few to tens of nanometers with minimal user supervision. The pipeline demonstrates high reproducibility and scalability, processing large image collections with consistent calibration and output formatting. Its modular design enables seamless integration of future deep-learning-based segmentation models, providing a pathway toward intelligent, data-driven electron microscopy analysis.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Exploring the keV-scale physics potential of CUORE

We present the analysis techniques developed to explore the keV-scale energy region of the Cryogenic Underground Observatory for Rare Events (CUORE) experiment, based on more than 2 metric ton yr of data collected over five years. By prioritizing a stricter selection over a larger exposure, we are able to optimize data selection for thresholds at 10 keV and 3 keV with 691 kg yr and 11 kg yr of data, respectively. We study how the performance varies among the 988-detector array with different detector characteristics and data-taking conditions. We achieve an average baseline resolution of 2.54 ±0.14 keV FWHM and 1.18 ±0.02 keV FWHM for the data selection at 10 keV and 3 keV, respectively. The analysis methods employed reduce the overall background by about an order of magnitude, reaching 2.06±0.05 counts/(keV kg days) and 16±2 counts/(keV kg days) at the thresholds of 10 keV and 3 keV. We evaluate for the first time the near-threshold reconstruction efficiencies of the CUORE experiment, and find these to be 50 ±2% and 26 ±4% at 10 keV and 3 keV, respectively. This analysis provides crucial insights into rare decay studies, new physics searches, and keV-scale background modeling with CUORE. We demonstrate that ton-scale cryogenic calorimeters can operate across a wide energy range, from keV to MeV, establishing their scalability as versatile detectors for rare event and dark matter physics. These findings also inform the optimization of future large mass cryogenic calorimeters to enhance the sensitivity to low-energy phenomena.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2006

Validation of a Scalable Solar Sailcraft

The NASA In-Space Propulsion (ISP) program sponsored intensive solar sail technology and systems design, development, and hardware demonstration activities over the past 3 years. Efforts to validate a scalable solar sail system by functional demonstration in relevant environments, together with test-analysis correlation activities on a scalable solar sail system have recently been successfully completed. A review of the program, with descriptions of the design, results of testing, and analytical model validations of component and assembly functional, strength, stiffness, shape, and dynamic behavior are discussed. The scaled performance of the validated system is projected to demonstrate the applicability to flight demonstration and important NASA road-map missions.

Murphy, D. M.↗
NASA NTRS2015

Advanced Mirror Technology Development (AMTD) Project: 3.0 Year Status

Advanced Mirror Technology Development (AMTD) is a funded NASA Strategic Astrophysics Technology project. Begun in 2011, we are in Phase 2 of a multi-year effort. Our objective is to mature towards TRL6 critical technologies needed to produce 4-m or larger flight-qualified UVOIR mirrors by 2018 so that a viable astronomy mission can be considered by the 2020 Decadal Review. The developed technology must enable missions capable of both general astrophysics and ultra-high contrast observations of exoplanets. Just as JWST's architecture was driven by launch vehicle, a future UVOIR mission's architecture (monolithic, segmented or interferometric) will depend on capacities of future launch vehicles (and budget). Since we cannot predict the future, we must prepare for all potential futures. Therefore, we are pursuing multiple technology paths. AMTD uses a science-driven systems engineering approach. We mature technologies required to enable the highest priority science AND result in a high-performance low-cost low-risk system. One of our key accomplishments is that we have derived engineering specifications for advanced normal-incidence monolithic and segmented mirror systems needed to enable both general astrophysics and ultra-high contrast observations of exoplanets missions as a function of potential launch vehicle and its inherent mass and volume constraints. Another key accomplishment is that we have matured our technology by building and testing hardware. To demonstrate stacked core technology, we built a 400 mm thick mirror. Currently, to demonstrate lateral scalability, we are manufacturing a 1.5 meter mirror. To assist in architecture trade studies, the Engineering team develops Structural, Thermal and Optical Performance (STOP) models of candidate mirror assembly systems including substrates, structures, and mechanisms. These models are validated by test of full- and subscale components in relevant thermo-vacuum environments. Specific analyses include: maximum mirror substrate size, first fundamental mode frequency (i.e., stiffness) and mass required to fabricate without quilting, survive launch, and achieve stable pointing and maximum thermal time constant.

Stahl, H. Philip↗
DOE OSTI2025

Topology-Aware Reinforcement Learning for Voltage Control: Centralized and Decentralized Strategies

Volt-VAR control (VVC) methods based on deep reinforcement learning (DRL) can effectively control distribution grid voltage and minimize power loss by implementing corrective and preventive control measures on the reactive power output of inverter-based distributed energy resources (DERs). However, model-free DRL-based VVC approaches usually cannot capture the important topological feature of the power system since they use a fully-connected network (FCN) to deliver the action. Therefore, this paper proposes a graph convolutional network (GCN)-based DRL approach that can employ the topological information of the network to take better control action for regulating the voltage. Our implementation allows for both centralized and decentralized configurations, utilizing a single agent and multiple agents respectively. Although the centralized GCN-based DRL approach has its advantages of minimizing voltage fluctuation and power loss, it is not suitable for large scale power systems due to its challenges in terms of scalability, computation speed and potential single points of failure. Therefore, these problems can be resolved using the decentralized GCN-based DRL approach. Moreover, to ensure the safe operation of the model, our proposed approach incorporates an exponential barrier function while formulating the reward function for each agent. To validate performance of the proposed approaches, the proposed model is tested on modified IEEE test systems and the performances are measured in terms on voltage fluctuation reduction, minimization of power loss and computational speed. Finally, the results show that the proposed topology-aware approach outperforms the FCN-based DRL approach in terms of reducing voltage fluctuation and minimizing power loss of the network. Moreover, it is shown that the decentralized GCN-based DRL has faster computational speed than other approaches.

42 ENGINEERING↗
DOE OSTI2025

Bayesian Optimized Deep Ensemble for Uncertainty Quantification of Deep Neural Networks: a System Safety Case Study on Sodium Fast Reactor Thermal Stratification Modeling

Deep neural networks (DNNs) are increasingly important to scientific computing and engineering system simulations. Accurate uncertainty quantification (UQ) for DNNs is critical in safety-sensitive engineering domains. Traditional Deep Ensemble (DE) methods, while easy to implement, frequently suffer from poorly calibrated uncertainty estimates and limited predictive accuracy due to reliance on fixed architectures with varied weight initializations. To address these issues, we introduce a workflow that combines Bayesian Optimization (BO) and DE. The workflow is modular, scalable, and integrates parallel BO initialized with Sobol sequences to individually optimize the hyperparameters of each ensemble member. This method enhances ensemble diversity, improves predictive accuracy, and provides reliable uncertainty estimates. We evaluate the proposed BODE approach in a sodium fast reactor thermal stratification modeling case study, where we used a densely connected convolutional neural network to predict turbulent viscosity during the reactor transient with consideration of data noise. We benchmark its performance against several optimization approaches, including baseline deep ensemble, evolutionary algorithm-optimized ensemble, ensemble formed via random search combined with greedy selection, and a BO ensemble using random initialization. Here, our results demonstrate superior performance of the developed BODE approach. In noise-free scenarios, BODE notably reduces incorrect aleatoric uncertainty and significantly enhances predictive accuracy. Under conditions of 5% and 10% Gaussian noise, BODE adaptively quantifies uncertainty proportional to data noise, achieving up to an 80% reduction in root mean square error compared to baseline methods and producing well-calibrated prediction intervals.

Bayesian optimization↗
DOE OSTI2026

Robust Rapid Cellular Metabolite Sensing Using Benchtop NMR and SABRE-Hyperpolarized [1- 13 C]Pyruvate

Hyperpolarized NMR has emerged as a powerful analytical technique to significantly enhance targeted NMR signals, improving the sensitivity for investigations of unique chemical and biological dynamics. Here, we demonstrate the use of a hyperpolarization strategy based on Signal Amplification By Reversible Exchange (SABRE) to generate highly reproducible doses of a hyperpolarized [1- 13 C]pyruvate probe for benchtop characterization of yeast metabolism. This method allows rapid, scalable, and benchtop preparation of biocompatible hyperpolarized solutions suitable for live-cell experiments. We show that this production can be dove-tailed into a modular, compact workflow to characterize real-time metabolism in cell cultures, using Saccharomyces cerevisiae (Baker’s yeast) as a model organism. With high temporal resolution, we show that this method can resolve the conversion of hyperpolarized [1- 13 C]pyruvate into oxidative decarboxylation products CO 2 and bicarbonate. This conversion exhibits sustained and detectable metabolic activity for over 300 s after introduction of the agent to the cells. We model the metabolite kinetics to show decarboxylation activity and derive estimates of the pH over time from the CO 2 and bicarbonate (carbonic acid buffer system) equilibrium to probe changes in the cellular environment during active metabolism. These results highlight the utility of benchtop SABRE-hyperpolarized [1- 13 C]pyruvate as a scalable, specific probe for metabolic phenotyping of living cells using compact, low-cost instrumentation well-suited for future high-throughput applications across microbial engineering, drug response profiling, and dynamic metabolic screening.

fungi↗