Search NASA⌕ Search

SEARCH · Search NASA

Results for “compressed training”

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 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Simple Wins: Driving Effective Implementation of Energy Efficiency Projects

Small- and medium-sized manufacturers (SMMs) in the United States, making up over 98% of manufacturing firms, account for 48.3% of the sector’s energy consumption. Despite access to energy assessments and technical resources, such as those from the US Department of Energy’s Better Plants program and Industrial Training and Assessment Centers, implementation rates for identified energy-saving measures remain lower than expectations, averaging 47%. Some barriers were identified by other researchers as limited workforce capacity, inadequate training, and time constraints hinder progress. This article studied the development of customized checklists together with the energy assessment report as a practical solution to improve implementation rates for energy efficiency projects among SMMs. The framework for the checklists was grounded in five principles: short, precise, actionable, relevant, and codeveloped (SPARC). A case study of a 500,000-squarefoot SMM demonstrates how tailored checklists targeting significant energy users—such as heating, ventilating and air conditioning (HVAC), lighting, and compressed air systems for this facility—can simplify the process for staff to do facility walkthroughs to monitor best practices and track the progress of energy efficiency project implementation.

Guo, Wei [Oak Ridge National Laboratory (ORNL), Oa↗
DOE OSTI2025

Space-Time Finite Element Tensor Network Approach for the Time-Dependent Convection–Diffusion–Reaction Equation with Variable Coefficients

In this paper, we present a new space-time Galerkin-like method, where we treat the discretization of spatial and temporal domains simultaneously. This method utilizes a mixed formulation of the tensor-train (TT) and quantized tensor-train (QTT) (please see Section Tensor-Train Decomposition), designed for the finite element discretization (Q1-FEM) of the time-dependent convection–diffusion–reaction (CDR) equation. We reformulate the assembly process of the finite element discretized CDR to enhance its compatibility with tensor operations and introduce a low-rank tensor structure for the finite element operators. Recognizing the banded structure inherent in the finite element framework’s discrete operators, we further exploit the QTT format of the CDR to achieve greater speed and compression. Additionally, we present a comprehensive approach for integrating variable coefficients of CDR into the global discrete operators within the TT/QTT framework. The effectiveness of the proposed method, in terms of memory efficiency and computational complexity, is demonstrated through a series of numerical experiments, including a semi-linear example.

convection–diffusion–reaction equation↗
NASA NTRS2015

Feasibility Assessment of an EVA Glove Sensing Platform to Evaluate Potential Hand Injury Risk Factors

Injuries to the hands are common among astronauts who train for extravehicular activity (EVA). When the gloves are pressurized, they restrict movement and create pressure points during tasks, sometimes resulting in pain, muscle fatigue, abrasions, and occasionally more severe injuries such as onycholysis. A brief review of the Lifetime Surveillance of Astronaut Health's injury database reveals that 58% of total astronaut hand and arm injuries from NBL training between 1993 and 2010 occurred either to the fingernail, MCP, or fingertip. The purpose of this study was to assess the potential of using small sensors to measure force acting on the fingers and hand within pressurized gloves and other variables such as blood perfusion, skin temperature, humidity, fingernail strain, skin moisture, among others. Tasks were performed gloved and ungloved in a pressurizable glove box. The test demonstrated that fingernails saw greater transverse strain levels for tension or compression than for longitudinal strain, even during axial fingertip loading. Blood perfusion peaked and dropped as the finger deformed during finger presses, indicating an initial dispersion and decrease of blood perfusion levels. Force sensitive resistors to force plate comparisons showed similar force curve patterns as fingers were depressed, indicating suitable functionality for future testing. Strategies for proper placement and protection of these sensors for ideal data collection and longevity through the test session were developed and will be implemented going forward for future testing.

Reid, Christopher R.↗
DOE OSTI2026

Economic Benefits of Alternative Fuels in Rural Alabama

This fact sheet provides an overview of potential alternative fuel vehicle operational cost savings, opportunities to generate revenue for the state and income for individuals, and several examples of alternative fuels supporting jobs in Alabama. Alternative transportation fuels could potentially provide economic benefits to rural Alabama by lowering operating costs for vehicle owners and creating new revenue streams and job opportunities for fuel providers, technicians, and dealerships. The use of electricity, compressed natural gas (CNG), liquefied natural gas (LNG), liquefied petroleum gas (LPG or propane) could result in reduced fuel costs and maintenance expenses for both public and private fleets, as well as individual consumers. Fleet operators often invest in private fueling infrastructure, while publicly accessible stations could attract traffic and commerce to local businesses and communities. Technicians capable of retrofitting alternative fuels and maintaining the vehicles and fueling infrastructure could develop business opportunities as well, but they require special training and often certifications. These cost savings and increased revenues may be recirculated through the economy and impact more than just the beneficiaries listed above.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2022

Real-time semantic segmentation on FPGAs for autonomous vehicles with hls4ml

In this paper, we investigate how field programmable gate arrays can serve as hardware accelerators for real-time semantic segmentation tasks relevant for autonomous driving. Considering compressed versions of the ENet convolutional neural network architecture, we demonstrate a fully-on-chip deployment with a latency of 4.9 ms per image, using less than 30% of the available resources on a Xilinx ZCU102 evaluation board. The latency is reduced to 3 ms per image when increasing the batch size to ten, corresponding to the use case where the autonomous vehicle receives inputs from multiple cameras simultaneously. We show, through aggressive filter reduction and heterogeneous quantization-aware training, and an optimized implementation of convolutional layers, that the power consumption and resource utilization can be significantly reduced while maintaining accuracy on the Cityscapes dataset.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2025

End-to-end deep learning pipeline for real-time Bragg peak segmentation: from training to large-scale deployment

X-ray crystallography reconstruction, which transforms discrete X-ray diffraction patterns into three-dimensional molecular structures, relies critically on accurate Bragg peak finding for structure determination. As X-ray free electron laser (XFEL) facilities advance toward MHz data rates (1 million images per second), traditional peak finding algorithms that require manual parameter tuning or exhaustive grid searches across multiple experiments become increasingly impractical. While deep learning approaches offer promising solutions, their deployment in high-throughput environments presents significant challenges in automated dataset labeling, model scalability, edge deployment efficiency, and distributed inference capabilities. We present an end-to-end deep learning pipeline with three key components: (1) a data engine that combines traditional algorithms with our peak matching algorithm to generate high-quality training data at scale, (2) a modular architecture that scales from a few million to hundreds of million parameters, enabling us to train large expert-level models offline while deploying smaller, distilled models at the edge, and (3) a decoupled producer-consumer architecture that separates specialized data source layer from model inference, enabling flexible deployment across diverse computing environments. Using this integrated approach, our pipeline achieves accuracy comparable to traditional methods tuned by human experts while eliminating the need for experiment-specific parameter tuning. Although current throughput requires optimization for MHz facilities, our system's scalable architecture and demonstrated model compression capabilities provide a foundation for future high-throughput XFEL deployments.

Wang, Cong↗
NASA NTRS1987

Vapor Compression Distillation Subsystem (VCDS) Component Enhancement, Testing and Expert Fault Diagnostics Development, Volume 2

A wide variety of Space Station functions will be managed via computerized controls. Many of these functions are at the same time very complex and very critical to the operation of the Space Station. The Environmental Control and Life Support System is one group of very complex and critical subsystems which directly affects the ability of the crew to perform their mission. Failure of the Environmental Control and Life Support Subsystems are to be avoided and, in the event of failure, repair must be effected as rapidly as possible. Due to the complex and diverse nature of the subsystems, it is not possible to train the Space Station crew to be experts in the operation of all of the subsystems. By applying the concepts of computer-based expert systems, it may be possible to provide the necessary expertise for these subsystems in dedicated controllers. In this way, an expert system could avoid failures and extend the operating time of the subsystems even in the event of failure of some components, and could reduce the time to repair by being able to pinpoint the cause of a failure when one cannot be avoided.

Mallinak, E. S.↗
DOE OSTI2026

Multigroup Thermal Radiation Transport with Tensor Trains

We investigate the application of tensor-train (TT) algorithms to multigroup thermal radiation transport (i.e., photon radiation transport). The TT framework enables simulations at discretizations that might otherwise be computationally infeasible on conventional hardware. We show that solutions to certain multigroup problems possess an intrinsic low-rank structure, which the TT representation leverages effectively. This enables us to solve problems where the discretized solution size exceeds a trillion parameters on a single node. The solver is evaluated on a range of test problems with varying levels of complexity, consistently achieving compression factors greater than 100× and speedups exceeding 2×. We also investigate alternative TT topologies by analyzing the low-rank structure of the merged spatio-spectral core to assess the potential for greater compression. This analysis suggests that compression gains could increase by factors as large as 7. Our results indicate that the low-rank structure of the merged spatio-spectral core captures the spatio-spectral complexity of the solution, largely driven by the opacity structure of the medium. Beyond identifying opportunities for improved compression, this analysis highlights the types of errors that may arise in angle-integrated quantities when exploiting this low-rank structure.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS2008

Crew Activity Analyzer

The crew activity analyzer (CAA) is a system of electronic hardware and software for automatically identifying patterns of group activity among crew members working together in an office, cockpit, workshop, laboratory, or other enclosed space. The CAA synchronously records multiple streams of data from digital video cameras, wireless microphones, and position sensors, then plays back and processes the data to identify activity patterns specified by human analysts. The processing greatly reduces the amount of time that the analysts must spend in examining large amounts of data, enabling the analysts to concentrate on subsets of data that represent activities of interest. The CAA has potential for use in a variety of governmental and commercial applications, including planning for crews for future long space flights, designing facilities wherein humans must work in proximity for long times, improving crew training and measuring crew performance in military settings, human-factors and safety assessment, development of team procedures, and behavioral and ethnographic research. The data-acquisition hardware of the CAA (see figure) includes two video cameras: an overhead one aimed upward at a paraboloidal mirror on the ceiling and one mounted on a wall aimed in a downward slant toward the crew area. As many as four wireless microphones can be worn by crew members. The audio signals received from the microphones are digitized, then compressed in preparation for storage. Approximate locations of as many as four crew members are measured by use of a Cricket indoor location system. [The Cricket indoor location system includes ultrasonic/radio beacon and listener units. A Cricket beacon (in this case, worn by a crew member) simultaneously transmits a pulse of ultrasound and a radio signal that contains identifying information. Each Cricket listener unit measures the difference between the times of reception of the ultrasound and radio signals from an identified beacon. Assuming essentially instantaneous propagation of the radio signal, the distance between that beacon and the listener unit is estimated from this time difference and the speed of sound in air.] In this system, six Cricket listener units are mounted in various positions on the ceiling, and as many as four Cricket beacons are attached to crew members. The three-dimensional position of each Cricket beacon can be estimated from the time-difference readings of that beacon from at least three Cricket listener units

Murray, James↗
DOE OSTI2020

Process Heating Assessments Using DOE’s Manufacturing Energy Assessment Software for Utility Reduction (MEASUR) Tool Suite

Process heating is the most energy-intensive manufacturing process for most sectors of industry. To quantify energy savings from various energy conservation measures, the Department of Energy (DOE) sponsored the development of the Process Heating Assessment and Survey Tool (PHAST) and similar tools for other industrial systems in the early 2000s. It has been used extensively in the Save Energy Now Program’s Energy Savings Assessments and the Better Plants Program’s In-Plant Trainings. Since the initial development of the legacy tools, both computer operating systems and software development have evolved significantly. Thus, DOE has invested in the modernization of PHAST and other legacy software tools to create the Manufacturing Energy Assessment Software for Utility Reduction (MEASUR) tool suite. MEASUR offers a collection of software tools that can aid manufacturing facilities in improving the efficiency of energy systems and equipment (specifically pumps, fans, steam, and process heating) and in conducting “Energy Treasure Hunts”. Eventually, the tool will also add compressed air and process cooling systems. The Process Heating Assessment (PHA) module of MEASUR is an upgrade of the PHAST tool. PHA provides the means to model fuel-fired, steam-based, and electric process heating systems, covering process heating for most industrial plants in manufacturing sector. It also includes several key upgrades, including the ability to consider multi-component charge loads and account for several different areas of energy losses. The new tool includes a comprehensive flue gas calculator to quantify available heat and heat loss for various gaseous, liquid, and solid fuels and new heat loss calculators. It generates a report and a dynamic Sankey diagram to show the energy consumption in various areas of energy use. MEASUR has significantly improved the user experience by adopting a modern software design. This paper details the structure and workflow of PHA and presents a real-world case study to demonstrate energy savings quantification and MEASUR’s outstanding reporting capabilities.

Nimbalkar, Sachin U.↗
DOE OSTI2024

Lion Cub: Minimizing Communication Overhead in Distributed Lion

Communication overhead is a key challenge in distributed deep learning, especially on slower Ethernet intercon nects, and given current hardware trends, communication is likely to become a major bottleneck. While gradient compression techniques have been explored for SGD and Adam, the Lion optimizer has the distinct advantage that its update vectors are the output of a sign operation, enabling straightforward quantization. However, simply compressing updates for communication and using techniques like majority voting fails to lead to end-to-end speedups due to inefficient communication algorithms and reduced convergence. We analyze three factors critical to distributed learning with Lion: optimizing communication methods, identifying effective quantization methods, and assessing the necessity of momentum synchronization. Our findings show that quantization techniques adapted to Lion and selective momentum synchronization can significantly reduce communication costs while maintaining convergence. We combine these into Lion Cub, which enables up to 5x speedups in end-to-end training compared to Lion. This highlights Lion’s potential as a communication-efficient solution for distributed training.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Review of Wave Energy Converter Power Take-Off Systems, Testing Practices, and Evaluation Metrics: Preprint

While the field of wave energy has been the subject of numerical simulation, scale model testing, and precommercial project testing for decades, wave energy technologies remain in the early stages of development and must continuing proving themselves as a promising modern renewable energy field. One of the difficulties that wave energy systems have been struggling to overcome is the design of highly efficient energy conversion systems that can convert the mechanical power, derived from the oscillation of wave activated bodies, into another useful product. Often the power take-off (PTO) is defined as the single unit responsible for converting mechanical power into another usable form such as electricity, pressurized fluid, compressed air, and others. The PTO, and the entire power conversion chain (PCC), is of great importance as it affects not only how efficient wave power is converted into electricity, but also contributes to the mass, size, structural dynamics, and levelized cost of energy (LCOE) of the wave energy converter (WEC). Unlike wind and solar, there is no industrial standard device, or devices, for wave energy conversion and this diversity is transferred to the PTO system. The majority of current WEC PTO systems incorporate a mechanical or hydraulic drive train, power generator, and an electrical control system. The challenge of WEC PTO designs is designing a mechanical-to-electrical component that can efficiently convert irregular, bi-directional, low frequency and low alternating velocity wave motions. While gross average power levels can be predicted in advance, the variable wave elevation input has to be converted into smooth electrical output and hence usually necessitates some type of energy storage system, such as battery storage, accumulator super capacitors, etc., or other means of compensation such as an array of devices. One of the primary challenges for wave energy converter systems is the fluctuating nature of wave resources, which require WEC components to be designed to handle loads (i.e. torques, forces, and powers) that are many times greater than the average load. This approach requires a much greater PTO capacity than the average power output and indicates a higher cost. In addition, supporting mechanical coupling and or gearing can be added to the PCC to help alleviate the difficulties with transmission and control of fluctuating large loads with low frequencies (indicative of wave forcing) into smaller loads at higher frequencies (optimum for conventional electrical machine design) can quickly increase the complexity of the PCC which could result in a greater number of failure modes and increased maintenance costs. All of the previous points demonstrate how the PTO influences WEC dynamics, reliability, performance and cost which are critical design factors. This paper further explores these topics by providing a review of the state-of-the-art PTO systems currently under development, how these novel PTO systems are tested and derisked prior to precommercial deployment, and the evaluation metrics historically used to differentiate between PTO designs and how they can be improved to support control co-design focused development of wave energy systems.

laboratory testing↗
DOE OSTI2023

A semi-agnostic ansatz with variable structure for variational quantum algorithms

Quantum machine learning—and specifically Variational Quantum Algorithms (VQAs)—offers a powerful, flexible paradigm for programming near-term quantum computers, with applications in chemistry, metrology, materials science, data science, and mathematics. Here, one trains an ansatz, in the form of a parameterized quantum circuit, to accomplish a task of interest. However, challenges have recently emerged suggesting that deep ansatzes are difficult to train, due to flat training landscapes caused by randomness or by hardware noise. This motivates our work, where we present a variable structure approach to build ansatzes for VQAs. Our approach, called VAns (Variable Ansatz), applies a set of rules to both grow and (crucially) remove quantum gates in an informed manner during the optimization. Consequently, VAns is ideally suited to mitigate trainability and noise-related issues by keeping the ansatz shallow. We employ VAns in the variational quantum eigensolver for condensed matter and quantum chemistry applications, in the quantum autoencoder for data compression and in unitary compilation problems showing successful results in all cases.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

HPC-FAIR: A Framework Managing Data and AI Models for Analyzing and Optimizing Scientific Applications

The increasing reliance on machine learning (ML) to analyze and optimize large-scale scientific applications on supercomputers faces a significant bottleneck: the lack of readily available, high-quality training datasets and the difficulty in reusing existing AI models. This project was motivated by the urgent need to address the “FAIR” principles (Findability, Accessibility, Interoperability, Reusability) for both training datasets and AI models in the high-performance computing (HPC) domain. The project developed HPC-FAIR, a high-performance computing data management framework designed to centralize HPC-related datasets and AI models within a unified hub. To ensure interoperability, the framework established a standardized representation and vocabulary (ontology) for both data and models. HPC-FAIR also implemented automated workflows to streamline data processing, model access, and benchmarking. Additionally, the project focused on optimizing data harnessing efficiency through advanced techniques like deep reuse and compression-based analytics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Solving high-dimensional inverse problems using amortized likelihood-free inference with noisy and incomplete data

Here, we present a likelihood-free probabilistic inversion method based on normalizing flows for high-dimensional inverse problems. The proposed method is composed of two complementary networks: a summary network for data compression and an inference network for parameter estimation. The summary network encodes raw observations into a fixed-size vector of summary features, while the inference network generates samples of the approximate posterior distribution of the model parameters based on these summary features. The posterior samples are produced in a deep generative fashion by sampling from a latent Gaussian distribution and passing these samples through an invertible transformation. We construct this invertible transformation by sequentially alternating conditional invertible neural network and conditional neural spline flow layers. The summary and inference networks are trained simultaneously. We apply the proposed method to an inversion problem in groundwater hydrology to estimate the posterior distribution of the log-conductivity field conditioned on spatially sparse time-series observations of the system’s hydraulic head responses. The conductivity field is represented with 706 degrees of freedom in the considered problem. Comparison with the likelihood-based iterative ensemble smoother PEST-IES method demonstrates that the proposed method accurately estimates the parameter posterior distribution and the observations’ predictive posterior distribution at a fraction of the inference time of PEST-IES.

conditional invertible neural network↗
DOE OSTI2020

Calculation of the detonation state of HN 3 with quantum accuracy

HN 3 is a unique liquid energetic material that exhibits ultrafast detonation chemistry and a transition to metallic states during detonation. We combine the Chebyshev interaction model for efficient simulation (ChIMES) many-body reactive force field and the extended-Lagrangian multiscale shock technique molecular dynamics method to calculate the detonation properties of HN 3 with the accuracy of Kohn–Sham density-functional theory. ChIMES is based on a Chebyshev polynomial expansion and can accurately reproduce density-functional theory molecular dynamics (DFT-MD) simulations for a wide range of unreactive and decomposition conditions of liquid HN 3 . We show that addition of random displacement configurations and the energies of gas-phase equilibrium products in the training set allows ChIMES to efficiently explore the complex potential energy surface. Schemes for selecting force field parameters and the inclusion of stress tensor and energy data in the training set are examined. Structural and dynamical properties and chemistry predictions for the resulting models are benchmarked against DFT-MD. We demonstrate that the inclusion of explicit four-body energy terms is necessary to capture the potential energy surface across a wide range of conditions. Our results generally retain the accuracy of DFT-MD while yielding a high degree of computational efficiency, allowing simulations to approach orders of magnitude larger time and spatial scales. The techniques and recipes for MD model creation we present allow for direct simulation of nanosecond shock compression experiments and calculation of the detonation properties of materials with the accuracy of Kohn–Sham density-functional theory.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

A mesoscale 3D model of irradiated concrete informed via a 2.5 U-Net semantic segmentation

The concrete biological shield in light-water reactors is exposed to neutron and gamma irradiation, which deteriorates the concrete’s mechanical properties in the long term. To assess the irradiation-induced damage, predictive mechanical models are developed and used in parallel with the characterization of irradiated concrete samples. Realistic 3D simulation domains can drastically improve a model’s prediction. In this work, we utilized x-ray computed tomography (XCT) data of a concrete specimen to reconstruct its 3D microstructure. The XCT data shows low contrast between the concrete’s aggregates and cement paste, resulting in poor image segmentation when using traditional unsupervised techniques. To address this issue, we developed and trained a 2.5D U-Net model on only 24 pre-labeled XCT layers to segment 651 layers of the XCT data. The overall F1-score of the model is approximately 96%. Then, we created a 3D finite element (FE) mesh based on the stack of segmented images. The FE model contains radiation-induced expansion, damage, and creep. The constitutive equations are adapted to each phase (aggregates and cement paste). Here, we simulated the effects of neutron irradiation in the concrete specimen as well as the specimen’s mechanical response to uniaxial compression. Finally, model validation was performed using experimental data on similar concrete specimens in the literature.

2.5D U-Net↗
DOE OSTI2025

MOOSE Thermal-Hydraulics Module - MOOSE workshop

The MOOSE Thermal Hydraulics Module (THM) is designed to facilitate the development of thermal hydraulic system models. It provides the capability to assemble networks of coupled components such as pipes, junctions, valves, turbomachinery, and heat exchangers. Its library of components supports a single-phase, compressible flow model based on a variable-area formulation of the Euler equations of gas dynamics and discretized using a finite volume scheme. THM offers a flexible system for specifying closures such as friction factors or heat transfer coefficients, allowing the user to choose from built-in correlations or define their own in the input file. A control logic system can be used to control input parameters, necessary for implementing transient scenarios and mirroring real control systems in thermal hydraulic systems. THM can be coupled with other MOOSE-based applications for multiphysics calculations. This training will give an introduction to the capabilities of THM and provide some examples of its usage and validation.

97 - MATHEMATICS AND COMPUTING↗