Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “training data”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Data-driven learning of nonlocal models: from high-fidelity simulations to constitutive laws

We show that machine learning can improve the accuracy of simulations of stress waves in one-dimensional composite materials. We propose a data-driven technique to learn nonlocal constitutive laws for stress wave propagation models. The method is an optimization-based technique in which the nonlocal kernel function is approximated via Bernstein polynomials. The kernel, including both its functional form and parameters, is derived so that when used in a nonlocal solver, it generates solutions that closely match high-fidelity data. The optimal kernel therefore acts as a homogenized nonlocal continuum model that accurately reproduces wave motion in a smaller-scale, more detailed model that can include multiple materials. We apply this technique to wave propagation within a heterogeneous bar with a periodic microstructure. Several one-dimensional numerical tests illustrate the accuracy of our algorithm. The optimal kernel is demonstrated to reproduce high-fidelity data for a composite material in applications that are substantially different from the problems used as training data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Developing Data-Driven Synthetic Infrastructure Models for Resilience Analysis

Research on infrastructure resilience has produced promising methods to simulate and optimize complex networks to improve performance. However, restrictions on sharing infrastructure models and the steep cost of developing and maintaining infrastructure models presents a roadblock to adoption. To overcome this limitation, this research focuses on methods to create data-driven infrastructure models that will help improve infrastructure resilience and security. The analysis couples incomplete utility data, geospatial data, machine learning, and synthetic network generation methods to rapidly develop and update infrastructure models. The methods are validated using realistic utility models and site-specific data, with a focus on Puerto Rico due to its unique infrastructure challenges and available data. This research highlights promising opportunities for the use of synthetic network generation and machine learning to create infrastructure models when very little data is available. Results demonstrate that hybrid methods, which combine sparse utility data with synthetic models, can enhance model accuracy, and machine learning can predict model attributes using training data from other models. However, the complexity of infrastructure systems means that even minor changes in network connectivity can significantly impact simulation results. Resilience analysis using synthetic infrastructure models shows that while some system behaviors are preserved, the magnitude of disruptions may not be accurately represented, indicating the need for more research and validation before using synthetic models for critical infrastructure investment decisions. The framework outlined in this report represents a significant advance to infrastructure model development and could be applied to additional domains and sites. Future research will continue to streamline and validate methods to help reduce roadblocks to resilience analysis.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1980

Influence of soils on Landsat spectral signatures of corn

Landsat data have been investigated extensively to determine crop types and acreage. However, confounding site factors have been found to reduce accuracy. Soils data in a small, contiguous area in southeast South Dakota were used to stratify Landsat data. A June 5 and July 29 CCT were used in a statistical analysis of corn training data. Significant soil parameters causing differences in study area soils were slope and parent material. Implication of the results is that, in this region, stratification of CCT data along parent material boundaries would improve corn classification accuracy. Research expanding on the interaction of soils and crops is both in progress and scheduled for additional studies in east central South Dakota.

Dalsted, K. J.↗
NASA NTRS1978

Transfer of training on manual control systems differing in short period frequency and damping characteristics

Each of four groups of 16 subjects was trained on one of four compensatory tracking tasks that differed with regard to short period natural frequency and damping characteristics. After completion of the training sessions, the members of each group either transferred to a task on which they had not been trained or continued with their original task. Analysis of the training data indicated that relative task difficulty was largely determined by system damping which, however, had little effect on the amount of transfer during the transfer trials. The effect of system frequency was essentially reversed, and a marked interaction between training and transfer frequencies was observed in the transfer data. Similar results were obtained both with relative error scores and transinformation scores. Positive transfer was exhibited by most of the groups when they transferred to tasks on which they had not been trained.

Lincoln, R. S.↗
DOE OSTI2024

Continuous integration data-driven platform of industrial-scale subsurface storage for real-time analytics

This project helped address the growing need for efficient and scalable models to support geological carbon and energy storage, which are crucial for achieving net-zero emissions. Traditionally accurate high-fidelity numerical models have been used to simulate relevant storage processes under a handful of processes, however such models are computationally demanding, making uncertainty quantification impractical. Consequently, we first developed a machine learning framework, based on Graph Neural Operators (GNOs), to improving the accuracy of model predictions for a fixed computational budget. We then developed an Ensemble of Improved Neural Operators (ENO), which uses bagging and Monte Carlo dropout techniques, to further improve prediction accuracy. Lastly, we developed the way to explain progressive transfer learning methods to reduce the amount of training data and computational cost of training (i.e., reduce trainable parameters) when using our models for multiple storage sites. Our numerical investigation, which used real-world case studies, demonstrated that our framework can significantly improve the safety and efficiency of geological storage operations, with potential applications in other domains such as geothermal reservoirs and climate modeling.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Rapid Adaptation of Chemical Named Entity Recognition Using Few-Shot Learning and LLM Distillation

Named entity recognition (NER) has been widely used in chemical text mining for the automatic identification and extraction of chemical entities. However, existing chemical NER systems primarily focus on scenarios with abundant training data, requiring significant human effort on annotations. This poses challenges for applications in the chemical field, such as catalysis, where many advancements have traditionally relied on trial-and-error investigations and incremental adjustment of variables. This hinders catalysis science and technology progress in addressing emerging energy and environmental crises. In this work, we propose a few-shot NER model that can quickly adapt to extract new types of chemical entities by using only a limited number of annotated examples. Our model employs a metric-learning approach to transfer entity similarity knowledge from high-resource chemical domains (with abundant annotations) to enable effective entity recognition in low-resource specialized domains (limited annotation). We validate the effectiveness of our model on a few-shot chemical NER benchmark built based on six existing chemical NER data sets. Experiments show that the proposed few-shot NER model can achieve reasonable performance with only 5 examples per entity type and shows consistent improvement as the number of examples increases. Furthermore, we demonstrate how the proposed model can be trained with large language model (LLM) annotated data, opening a new pathway for rapid adaptation of NER systems. Furthermore, our approach leverages the knowledge broadness of large language models for chemistry while distilling this knowledge into a lightweight model suitable for efficient and in-house use.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Weakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences

Biomedical repositories such as the UK Biobank provide increasing access to prospectively collected cardiac imaging, however these data are unlabeled, which creates barriers to their use in supervised machine learning. We develop a weakly supervised deep learning model for classification of aortic valve malformations using up to 4,000 unlabeled cardiac MRI sequences. Instead of requiring highly curated training data, weak supervision relies on noisy heuristics defined by domain experts to programmatically generate large-scale, imperfect training labels. For aortic valve classification, models trained with imperfect labels substantially outperform a supervised model trained on hand-labeled MRIs. In an orthogonal validation experiment using health outcomes data, our model identifies individuals with a 1.8-fold increase in risk of a major adverse cardiac event. This work formalizes a deep learning baseline for aortic valve classification and outlines a general strategy for using weak supervision to train machine learning models using unlabeled medical images at scale.

60 APPLIED LIFE SCIENCES↗
NASA NTRS2017

Tracking Historical NASA EVA Training: Lifetime Surveillance of Astronaut Health (LSAH) Development of the EVA Suit Exposure Tracker (EVA SET)

During a spacewalk, designated as extravehicular activity (EVA), an astronaut ventures from the protective environment of the spacecraft into the vacuum of space. EVAs are among the most challenging tasks during a mission, as they are complex and place the astronaut in a highly stressful environment dependent on the spacesuit for survival. Due to the complexity of EVA, NASA has conducted various training programs on Earth to mimic the environment of space and to practice maneuvers in a more controlled and forgiving environment. However, rewards offset the risks of EVA, as some of the greatest accomplishments in the space program were accomplished during EVA, such as the Apollo moonwalks and the Hubble Space Telescope repair missions. Water has become the environment of choice for EVA training on Earth, using neutral buoyancy as a substitute for microgravity. During EVA training, an astronaut wears a modified version of the spacesuit adapted for working in water. This high fidelity suit allows the astronaut to move in the water while performing tasks on full-sized mockups of space vehicles, telescopes, and satellites. During the early Gemini missions, several EVA objectives were much more difficult than planned and required additional time. Later missions demonstrated that "complex (EVA) tasks were feasible when restraints maintained body position and underwater simulation training ensured a high success probability".1,2 EVA training has evolved from controlling body positioning to perform basic tasks to complex maintenance of the Hubble Space Telescope and construction of the International Space Station (ISS). Today, preparation is centered at special facilities built specifically for EVA training, such as the Neutral Buoyancy Laboratory (NBL) at NASA's Johnson Space Center ([JSC], Houston) and the Hydrolab at the Gagarin Cosmonaut Training Centre ([GCTC], Star City, outside Moscow). Underwater training for an EVA is also considered hazardous duty for NASA astronauts. This activity places astronauts at risk for decompression sickness and barotrauma as well as various musculoskeletal disorders from working in the spacesuit. The medical, operational and research communities over the years have requested access to EVA training data to better understand the risks. As a result of these requests, epidemiologists within the Lifetime Surveillance of Astronaut Health (LSAH) team have compiled records from numerous EVA training venues to quantify the exposure to EVA training. The EVA Suit Exposure Tracker (EVA SET) dataset is a compilation of ground-based training activities using the extravehicular mobility unit (EMU) in neutrally buoyant pools to enhance EVA performance on orbit. These data can be used by the current ISS program and future exploration missions by informing physicians, researchers, and operational personnel on the risks of EVA training in order that future suit and mission designs incorporate greater safety. The purpose of this technical report is to document briefly the various facilities where NASA astronauts have performed EVA training while describing in detail the EVA training records used to generate the EVA SET dataset.

Laughlin, Mitzi S.↗
DOE OSTI2023

Decentralized digital twins of complex dynamical systems

Abstract In this article, we introduce a decentralized digital twin (DDT) modeling framework and its potential applications in computational science and engineering. The DDT methodology is based on the idea of federated learning, a subfield of machine learning that promotes knowledge exchange without disclosing actual data. Clients can learn an aggregated model cooperatively using this method while maintaining complete client-specific training data. We use a variety of dynamical systems, which are frequently used as prototypes for simulating complex transport processes in spatiotemporal systems, to show the viability of the DDT framework. Our findings suggest that constructing highly accurate decentralized digital twins in complex nonlinear spatiotemporal systems may be made possible by federated machine learning.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Automatic microseismic event picking via unsupervised machine learning

SUMMARY Effective and efficient arrival picking plays an important role in microseismic and earthquake data processing and imaging. Widely used short-term-average long-term-average ratio (STA/LTA) based arrival picking algorithms suffer from the sensitivity to moderate-to-strong random ambient noise. To make the state-of-the-art arrival picking approaches effective, microseismic data need to be first pre-processed, for example, removing sufficient amount of noise, and second analysed by arrival pickers. To conquer the noise issue in arrival picking for weak microseismic or earthquake event, I leverage the machine learning techniques to help recognizing seismic waveforms in microseismic or earthquake data. Because of the dependency of supervised machine learning algorithm on large volume of well-designed training data, I utilize an unsupervised machine learning algorithm to help cluster the time samples into two groups, that is, waveform points and non-waveform points. The fuzzy clustering algorithm has been demonstrated to be effective for such purpose. A group of synthetic, real microseismic and earthquake data sets with different levels of complexity show that the proposed method is much more robust than the state-of-the-art STA/LTA method in picking microseismic events, even in the case of moderately strong background noise.

Chen, Yangkang↗
DOE OSTI2025

A Weakly Supervised Machine Learning Procedure for Magnet Quench Diagnostics

Voltage taps remain the standard and reliable diagnostic tool for detecting quenches in superconducting magnets. However, they identify a quench only at the time of voltage rise and do not provide information on earlier physical precursors. In this work, we investigate whether acoustic emission data can reveal precursor activity that occurs before conventional voltage detection using machine learning techniques. We introduce an event selection method and a weakly supervised machine learning procedure to learn data-driven criteria for identifying potential acoustic precursors to quenches. Two Convolutional Neural Network (CNN) architectures are trained: one on acoustic sensor events from our selection procedure and one on the Fast Fourier Transforms (FFTs) of these events. Both networks are trained iteratively using confidence-weighted loss functions to associate certain subsets of training data with a precursor label. We evaluate the performance of these models by examining the time distribution of events classified as potential precursors relative to the quench onset. Results indicate that the proposed approach can possibly distinguish acoustic emission events occurring closer to the quench from earlier acoustic activity during ramping, suggesting the potential for flagging quench precursors in acoustic data.

Khan, Maira [Fermilab] (ORCID:0009000891602387)↗
DOE OSTI2025

Development and transferability of neural-network models for plasma-surface interactions

Plasma-surface interactions are increasingly critical to modern technologies; yet, accurate molecular dynamics simulations remain limited by the capabilities of interatomic potentials. Deep Potentials (DPs) promise to revolutionize the field by providing a systematic method for producing accurate interatomic potentials. The primary challenge of DP development is selecting a dataset, which efficiently spans the set of atomic environments one expects to encounter in the subsequent molecular dynamics simulations. The computational cost of density functional theory calculations, which are the typical basis for DP development, makes it impossible to directly verify the quality of a given DP. To address this challenge, we explore the development of a deep-learned interatomic potential, “DeepREBO,” trained to reproduce the behavior of the REBO2 empirical potential, enabling direct validation of training methodology and transferability. Using an active learning framework, we begin with a minimal dataset and iteratively expand it to train a Deep Potential-Smooth Edition model that faithfully reproduces REBO2 results for 25 eV hydrogen bombardment of diamond (001), a particularly challenging case. We show that small, carefully curated datasets can outperform large, unguided ones, with effective models requiring fewer than 15 000 snapshots. Subsequent transferability tests demonstrate that while DeepREBO generalizes well to diamond (111) surfaces, performance degrades for amorphous carbon or higher-energy impacts, highlighting the need for use-case-specific training data. We also evaluate methods to improve short-range repulsion. This study outlines best practices for training robust deep potentials and underscores the importance of dataset design for predictive plasma simulations.

Ab-initio molecular dynamics↗
DOE OSTI2022

Preliminary Results for Using Uncertainty and Out-of-distribution Detection to Identify Unreliable Predictions.

As machine learning (ML) models are deployed into an ever-diversifying set of application spaces, ranging from self-driving cars to cybersecurity to climate modeling, the need to carefully evaluate model credibility becomes increasingly important. Uncertainty quantification (UQ) provides important information about the ability of a learned model to make sound predictions, often with respect to individual test cases. However, most UQ methods for ML are themselves data-driven and therefore susceptible to the same knowledge gaps as the models themselves. Specifically, UQ helps to identify points near decision boundaries where the models fit the data poorly, yet predictions can score as certain for points that are under-represented by the training data and thus out-of-distribution (OOD). One method for evaluating the quality of both ML models and their associated uncertainty estimates is out-of-distribution detection (OODD). We combine OODD with UQ to provide insights into the reliability of the individual predictions made by an ML model.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2007

Integrated Cloud-Aerosol-Radiation Product using CERES, MODIS, CALIPSO and CloudSat Data

This paper documents the development of the first integrated data set of global vertical profiles of clouds, aerosols, and radiation using the combined NASA A-Train data from the Aqua Clouds and Earth's Radiant Energy System (CERES) and Moderate Resolution Imaging Spectroradiometer (MODIS), Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), and CloudSat. As part of this effort, cloud data from the CALIPSO lidar and the CloudSat radar are merged with the integrated column cloud properties from the CERES-MODIS analyses. The active and passive datasets are compared to determine commonalities and differences in order to facilitate the development of a 3- dimensional cloud and aerosol dataset that will then be integrated into the CERES broadband radiance footprint. Preliminary results from the comparisons for April 2007 reveal that the CERES-MODIS global cloud amounts are, on average, 0.14 less and 0.15 greater than those from CALIPSO and CloudSat, respectively. These new data will provide unprecedented ability to test and improve global cloud and aerosol models, to investigate aerosol direct and indirect radiative forcing, and to validate the accuracy of global aerosol, cloud, and radiation data sets especially in polar regions and for multi-layered cloud conditions.

Sun-Mack, Sunny↗
NASA NTRS2012

Towards Reliable Evaluation of Anomaly-Based Intrusion Detection Performance

This report describes the results of research into the effects of environment-induced noise on the evaluation process for anomaly detectors in the cyber security domain. This research was conducted during a 10-week summer internship program from the 19th of August, 2012 to the 23rd of August, 2012 at the Jet Propulsion Laboratory in Pasadena, California. The research performed lies within the larger context of the Los Angeles Department of Water and Power (LADWP) Smart Grid cyber security project, a Department of Energy (DoE) funded effort involving the Jet Propulsion Laboratory, California Institute of Technology and the University of Southern California/ Information Sciences Institute. The results of the present effort constitute an important contribution towards building more rigorous evaluation paradigms for anomaly-based intrusion detectors in complex cyber physical systems such as the Smart Grid. Anomaly detection is a key strategy for cyber intrusion detection and operates by identifying deviations from profiles of nominal behavior and are thus conceptually appealing for detecting "novel" attacks. Evaluating the performance of such a detector requires assessing: (a) how well it captures the model of nominal behavior, and (b) how well it detects attacks (deviations from normality). Current evaluation methods produce results that give insufficient insight into the operation of a detector, inevitably resulting in a significantly poor characterization of a detectors performance. In this work, we first describe a preliminary taxonomy of key evaluation constructs that are necessary for establishing rigor in the evaluation regime of an anomaly detector. We then focus on clarifying the impact of the operational environment on the manifestation of attacks in monitored data. We show how dynamic and evolving environments can introduce high variability into the data stream perturbing detector performance. Prior research has focused on understanding the impact of this variability in training data for anomaly detectors, but has ignored variability in the attack signal that will necessarily affect the evaluation results for such detectors. We posit that current evaluation strategies implicitly assume that attacks always manifest in a stable manner; we show that this assumption is wrong. We describe a simple experiment to demonstrate the effects of environmental noise on the manifestation of attacks in data and introduce the notion of attack manifestation stability. Finally, we argue that conclusions about detector performance will be unreliable and incomplete if the stability of attack manifestation is not accounted for in the evaluation strategy.

cyber defense↗
NASA NTRSDate not supplied

Development of TEMPO Products and Tools to Support Air Quality Management Decisions

The TEMPO mission has been observing air pollutants every hour during the daytime across its Field of Regard (FoR) covering greater North America since First Light on August 2, 2023. The highly anticipated public release of TEMPO data occurred on May 20, 2024, consisting of level 2 and level 3 trace gas data products of nitrogen dioxide, formaldehyde, and ozone. Our project at the NASA SPoRT Center is developing value-added products and tools to support the TEMPO mission and Early Adopters program with special attention on the air quality management community. The initial focus of this project is evaluating the TEMPO products over stakeholder target areas using Pandora and surface monitor observations. Methods for oversampling TEMPO data to 1 km resolution are being applied over the target areas to resolve fine-scale emission sources and pollutant gradients. Machine learning techniques using TEMPO, surface monitor, and model data to estimate surface-level nitrogen dioxide concentrations are being developed over the target areas. Our SPoRT viewer has been updated to include visualizations of the TEMPO products and an ArcGIS dashboard is being designed for enabling air quality management stakeholders to efficiently analyze TEMPO data. Training materials including user guides are being developed to ensure the effective and sustained use of TEMPO data in air quality management applications. The major outcome of this project is to support the inclusion of TEMPO data in exceptional event demonstrations by active engagement with stakeholders and ultimately enable more informed air quality management decisions in the future. This talk will provide an update on our project activities and showcase use cases of TEMPO data for monitoring different emission sources including wildland fire smoke.

air quality↗
NASA NTRSDate not supplied

Transitioning TEMPO Data for Air Quality Management Applications at the NASA SPoRT Center

The TEMPO mission has been observing air pollutants every hour during the daytime across its Field of Regard (FoR) covering greater North America since First Light on August 2, 2023. The highly anticipated public release of TEMPO data occurred on May 20, 2024, consisting of level 2 and level 3 trace gas data products of nitrogen dioxide, formaldehyde, and ozone. The NASA SPoRT Center is developing value-added products and tools to support the TEMPO mission and Early Adopters program with special attention on stakeholder from air agencies. One component of our work is focused on evaluating the TEMPO products over stakeholder target areas using Pandora and surface monitor observations to characterize the uncertainties and develop best practices for processing and analyzing TEMPO data. Methods for oversampling TEMPO data to 1 km resolution are being applied over the target areas to resolve fine-scale emission sources and pollutant gradients. Machine learning techniques using TEMPO, surface monitor, and model data to estimate surface-level nitrogen dioxide concentrations are being developed over the target areas. A SPoRT viewer for TEMPO has been launched for providing visualizations of the TEMPO products and an ArcGIS dashboard is being designed for enabling air agency stakeholders to efficiently analyze TEMPO data. Training materials including user guides are being developed to ensure the effective and sustained use of TEMPO data at our stakeholder agencies. One major goal of our TEMPO initiatives at SPoRT is to better enable the inclusion of TEMPO data in air quality management applications such as exceptional event demonstrations through our close engagement with stakeholders. This talk will provide an update on our SPoRT activities and showcase use cases of TEMPO data for monitoring different emission sources including wildland fire smoke.

Air Quality↗
DOE OSTI2021

Explicit physics-informed neural networks for nonlinear closure: The case of transport in tissues

In upscaling methods, closures for nonlinear problems present a well-known challenge. While a number of theoretical methods have been proposed for handling such closures, nonlinearities still remain a significant obstacle for many problems. In this work, we use a combination of formal upscaling and data-driven machine learning for explicitly closing a nonlinear transport and reaction process in multiscale tissues. The classical effectiveness factor model is used to formulate the macroscale reaction kinetics. We train a multilayer perceptron network using training data generated by direct numerical simulations over microscale examples. Once trained, the network is used in an algorithm for numerically solving the upscaled (coarse-grained) differential equation describing mass transport and reaction in two example tissues. The network is described as being explicit in the sense that the network is trained using macroscale concentrations and gradients of concentration as components of the feature space rather than incorporating them as part of a constraint in the optimization process. Network training and solutions to the macroscale transport equations were computed for two different tissues. The two tissue types (brain and liver) exhibit markedly different geometrical complexity and spatial scale (cell size and sample size). The upscaled solutions for the average concentration are compared with numerical solutions derived from the microscale concentration fields by a posteriori averaging. There are three outcomes of this work of particular note. 1) Our overall approach results in an upscaled nonlinear PDE. The PDE is closed using a neural network, and our approach results in the definition of the classical effectiveness factor for effecting closure. 2) We identify particular source terms for the closure problem that are important for representing the structure of the closure. These source terms involve macroscale concentrations and their gradients. We adopt these source terms to use as explicit features in the learning algorithm. We find the trained networks that include the macroscale source terms generate models that are able to predict the correction factor with increased fidelity over those that do not. 3) We find that the trained network exhibits good generalizability, and it is able to predict the effectiveness factor with high fidelity for realistically-structured tissues despite the significantly different scale and geometrical complexity of the two example tissue types. This latter result emphasizes our purposeful connection between conventional averaging methods with the use of machine learning for closure; this contrasts with some machine learning methods for upscaling where the exact form of the macroscale equation remains unknown.

97 MATHEMATICS AND COMPUTING↗