Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ML T&E”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

FAIR AI models in high energy physics

Abstract The findable, accessible, interoperable, and reusable (FAIR) data principles provide a framework for examining, evaluating, and improving how data is shared to facilitate scientific discovery. Generalizing these principles to research software and other digital products is an active area of research. Machine learning models—algorithms that have been trained on data without being explicitly programmed—and more generally, artificial intelligence (AI) models, are an important target for this because of the ever-increasing pace with which AI is transforming scientific domains, such as experimental high energy physics (HEP). In this paper, we propose a practical definition of FAIR principles for AI models in HEP and describe a template for the application of these principles. We demonstrate the template’s use with an example AI model applied to HEP, in which a graph neural network is used to identify Higgs bosons decaying to two bottom quarks. We report on the robustness of this FAIR AI model, its portability across hardware architectures and software frameworks, and its interpretability.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Platform for Automated Anomaly Detection in the Mercury Process System at the Target System in the Spallation Neutron Source

The Spallation Neutron Source (SNS) at Oak Ridge National Laboratory accelerates proton beams, which are directed toward a mercury target to generate the world’s most intense neutron beams via spallation. The target system consists of several interconnected subsystems and accounts for a major share of the facility’s overall downtime. Early detection of anomalies in the target system response can thus provide the possibility of taking corrective actions to reduce downtime. Accelerator facilities have largely focused on the beam side for data-driven fault prognostics. On the target side, SNS relies on operational shift technicians (OSTs), who respond to alarms and manually flag anomalies onto the System Tracking and Reliability (STAR) platform. This paper presents one of the first studies of using machine learning (ML) to automate anomaly detection in the target system. The study focused on the mercury process system as the first use case and employed reconstruction-based anomaly detection on minutely sampled time series signals. The pipeline was integrated into the STAR platform to autonomously rank and flag anomalies every week. The STAR platform provides a user interface for the OSTs to evaluate the flagged anomalies, thereby incorporating human feedback.

Anomaly detection↗
DOE OSTI2025

Reagent-free Hyperspectral Diagnosis of SARS-CoV-2 Infection in Saliva Samples

Rapid, reagent-free pathogen-agnostic diagnostics that can be performed at the point of need are vital for preparedness against future outbreaks. Yet, many current strategies are pathogen-specific and require several reagents. We present hyperspectral sensing, using light to non-invasively measure the composition of several molecules to form a spectral signature, to overcome these barriers. To generate these spectral signatures, we present the ProSpectral TM V1, a novel, miniaturized hyperspectral platform with high spectral resolution with two mini-spectrometers. Furthermore, we developed state-of-the-art ML pipelines for near real-time analysis of spectral signatures in saliva samples. We found that we could accurately identify SARS-CoV-2 infection status in double-blinded saliva samples and demonstrate 100% accuracy on a hold out test dataset. To our knowledge, this establishes the fastest hyperspectral diagnostic platform and in a small form factor, and executable with liquid samples, without ligands or reagents, all while maintaining PCR level specificity and sensitivity.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1993

Important influence of respiration on human R-R interval power spectra is largely ignored

Frequency-domain analyses of R-R intervals are used widely to estimate levels of autonomic neural traffic to the human heart. Because respiration modulates autonomic activity, we determined for nine healthy subjects the influence of breathing frequency and tidal volume on R-R interval power spectra (fast-Fourier transform method). We also surveyed published literature to determine current practices in this burgeoning field of scientific inquiry. Supine subjects breathed at rates of 6, 7.5, 10, 15, 17.1, 20, and 24 breaths/min and with nominal tidal volumes of 1,000 and 1,500 ml. R-R interval power at respiratory and low (0.06-0.14 Hz) frequencies declined significantly as breathing frequency increased. R-R interval power at respiratory frequencies was significantly greater at a tidal volume of 1,500 than 1,000 ml. Neither breathing frequency nor tidal volume influenced average R-R intervals significantly. Our review of studies reporting human R-R interval power spectra showed that 51% of the studies controlled respiratory rate, 11% controlled tidal volume, and 11% controlled both respiratory rate and tidal volume. The major implications of our analyses are that breathing parameters strongly influence low-frequency as well as respiratory frequency R-R interval power spectra and that this influence is largely ignored in published research.

NASA Discipline Cardiopulmonary↗
NASA NTRSDate not supplied

A Flexible Multibody Approach to Space Launch System Liftoff Pad Separation with Umbilical Disconnect

A flexible multibody dynamic framework is applied to the Space Launch System (SLS) liftoff Coupled Loads Analysis (CLA), enabling computational efficiencies and systematic inclusion of components and various nonlinearities. The objective of this work is to simulate the SLS liftoff transient pad separation event inclusive of umbilical disconnects. The vehicle stacking preloads and cryogenically induced preloads, due to the fueling of the Core Stage (CS), are integrated via specialized deformed coupling procedure to ensure full capture of the liftoff transient event. The Vehicle Stabilizer System (VSS) hydraulic struts are nonlinear and provide amplitude dependent damping, significantly contributing to the liftoff transient event. Ground Extensible Columns (EC) provide additional stiffness and preloads on the Mobile Launcher (ML), also significantly contributing to the liftoff transient. An advanced Henkel-Mar algorithm, with separation and re-contact, thus tracking the potential post separation re-contact between the booster aft skirt and ML, the umbilicals and vehicle, and ground EC separation. For this system level nonlinear dynamic simulation, the flexible multibody framework provides a systematic approach for the addition of component specific algorithms to enforce nonlinearities. This multibody framework is similar to an object-oriented programming language integrating classes of associated subroutines. As such, the complexities associated with the nonlinear CLA are greatly reduced, facilitating simplified bookkeeping as well as accelerating the execution of the nonlinear time-domain simulations.

Joel W Sills Jr.↗
NASA NTRS1994

Hypervolemia During Drinking of ISO and Hypertonic Fluids by Men at Rest and Exercise

To test the hypothesis that drink composition is more important than osmolality (osm) for maintaining and increasing plasma vol. (PV, hypervolemia) at rest and exercise (21.8C Tdb, 50% rh), 6 men (22-39 vr) underwent six treatments while sitting for 90 min (0.39 l/min) followed by 70 min of sitting ergometer exercise (2.08 l/min, 70%vO2 peak). Resting, intermittent drinking (10 ml/kg, 768 ml) beverages were: P1 (20 mEq Na+, 365 mosmol/kg), P2 (40 mEq Na+. 791 mosm), P2G (40 mEq Na+, 80 ml glycerol, 1.392 mosm). AA (157 mEq Na+, 253 mosm), and 01 and 02 (nothing). The exercise drink (10 ml/kg, 768 ml) was P1 for all experiments, except 02 (nothing). Resting % delta PV (Hb-Hct) increased (P<0.05) by 3-6% only with P2 and AA, respectively, Exercise % delta PV was + 1 to + 3% (NS) with AA, -6% to 0% (KS) with P1, P2. P2G, and 01,and 8% to -5% (P<0.05) with 02- Thus, AA with 157 mEq Na+ and the lowest osmolality maintained PV at rest and exercise, while the other beverages with low Na+ and higher osmolality (365 to 1.382 mosm) did not.

Greenleaf, J. E.↗
DOE OSTI2020

The joint automated repository for various integrated simulations (JARVIS) for data-driven materials design

The Joint Automated Repository for Various Integrated Simulations (JARVIS) is an integrated infrastructure to accelerate materials discovery and design using density functional theory (DFT), classical force-fields (FF), and machine learning (ML) techniques. JARVIS is motivated by the Materials Genome Initiative (MGI) principles of developing open-access databases and tools to reduce the cost and development time of materials discovery, optimization, and deployment. The major features of JARVIS are: JARVIS-DFT, JARVIS-FF, JARVIS-ML, and JARVIS-tools. To date, JARVIS consists of ≈40,000 materials and ≈1 million calculated properties in JARVIS-DFT, ≈500 materials and ≈110 force-fields in JARVIS-FF, and ≈25 ML models for material-property predictions in JARVIS-ML, all of which are continuously expanding. JARVIS-tools provides scripts and workflows for running and analyzing various simulations. We compare our computational data to experiments or high-fidelity computational methods wherever applicable to evaluate error/uncertainty in predictions. In addition to the existing workflows, the infrastructure can support a wide variety of other technologically important applications as part of the data-driven materials design paradigm. The JARVIS datasets and tools are publicly available at the website: https://jarvis.nist.gov .

36 MATERIALS SCIENCE↗
NASA NTRS2000

Low LBNP tolerance in men is associated with attenuated activation of the renin-angiotensin system

Plasma vasoactive hormone concentrations [epinephrine (p(Epi)), norepinephrine (p(NE)), ANG II (p(ANG II)), vasopressin (p(VP)), endothelin-1 (p(ET-1))] and plasma renin activity (p(RA)) were measured periodically and compared during lower body negative pressure (LBNP) to test the hypothesis that responsiveness of the renin-angiotensin system, the latter being one of the most powerful vasoconstrictors in the body, is of major importance for LBNP tolerance. Healthy men on a controlled diet (2,822 cal/day, 2 mmol. kg(-1). day(-1) Na(+)) were exposed to 30 min of LBNP from -15 to -50 mmHg. LBNP was uneventful for seven men [25 +/- 2 yr, high-tolerance (HiTol) group], but eight men (26 +/- 3 yr) reached presyncope after 11 +/- 1 min [P < 0.001, low-tolerance (LoTol) group]. Mean arterial pressure (MAP) did not change measurably, but central venous pressure and left atrial diameter decreased similarly in both groups (5-6 mmHg, by approximately 30%, P < 0.05). Control (0 mmHg LBNP) hormone concentrations were similar between groups, however, p(RA) differed between them (LoTol 0.6 +/- 0.1, HiTol 1.2 +/- 0.1 ng ANG I. ml(-1). h(-1), P < 0.05). LBNP increased (P < 0. 05) p(RA) and p(ANG II), respectively, more in the HiTol group (9.9 +/- 2.2 ng ANG I. ml(-1). h(-1) and 58 +/- 12 pg/ml) than in LoTol subjects (4.3 +/- 0.9 ng ANG I. ml(-1). h(-1) and 28 +/- 6 pg/ml). In contrast, the increase in p(VP) was higher (P < 0.05) in the LoTol than in the HiTol group. The increases (P < 0.05) for p(NE) were nonsignificant between groups, and p(ET-1) remained unchanged. Thus there may be a causal relationship between attenuated activation of p(RA) and p(ANG II) and presyncope, with p(VP) being a possible cofactor. Measurement of resting p(RA) may be of predictive value for those with lower hypotensive tolerance.

NASA Center ARC↗
DOE OSTI2025

Harnessing Machine Learning for Agnostic Biodetection

The United States’ current list-based approach to biodefense is limited because it considers only known biological agents. Alternatively, developing and adopting a system based on agent-agnostic signatures would enable detection and characterization of both known and novel agents, thereby engendering greater adaptability in the face of an evolving threat landscape. Machine learning (ML) could aid in such a transition, as it can recognize and encode highly complex patterns from multiple input data modalities and has already demonstrated success in many healthcare and defense applications. Functionalizing ML for environmental biodetection requires understanding current technical capabilities. In this article, we provide a systematic review of existing ML platforms and discuss anticipated development efforts needed to achieve effective ML-enabled, agnostic biodetection.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Integrated RMP-based ELM-crash-control process for plasma performance enhancement during ELM crash suppression in KSTAR

The integrated Resonant Magnetic Perturbation (RMP)-based Edge-Localized Mode (ELM)-crash-control process aims to enhance the plasma performance during the RMP-driven ELM crash suppression, where the RMP induces an unwanted confinement degradation. In this study, the normalized beta ($\beta_\textrm{N}$) is introduced as a metric for plasma performance. The integrated process incorporates the latest achievements in the RMP technique to enhance $\beta_\textrm{N}$ efficiently. The integrated process triggers the n = 1 Edge-localized RMP (ERMP) at the L–H transition timing using the real-time Machine Learning (ML) classifier. The pre-emptive RMP onset can reduce the required external heating power for achieving the same $\beta_\textrm{N}$ by over 10% compared to the conventional onset. During the RMP phase, the adaptive feedback RMP ELM controller, demonstrating its performance in previous experiments, plays a crucial role in maximizing $\beta_\textrm{N}$ during the suppression phase and sustaining the $\beta_\textrm{N}$-enhanced suppression state by optimizing the RMP strength. The integrated process achieves $\beta_\textrm{N}$ up to ~2.65 during the suppression phase, which is ~10% higher than the previous KSTAR record but ~6% lower than the target of the K-DEMO first phase ($\beta_\textrm{N}$ = 2.8), and maintains the suppression phase above the lower limit of target $\beta_\textrm{N}$ (= 2.4) for ~4 s (~60$\tau_\textrm{E}$). In addition to $\beta_\textrm{N}$ enhancement, the integrated process demonstrates quicker restoration of the suppression phase and recovery of $\beta_\textrm{N}$ compared to the adaptive control with the n = 1 Conventional RMP (CRMP). The post-analysis of the experiment shows the localized effect of the ERMP spectrum in radial and the close relationship between the evolution of $\beta_\textrm{N}$ and the electron temperature.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Results and lessons learned from accelerating radio frequency modeling using machine learning [slides]

The “advanced tokamak” reactor concept is a leading candidate for a steady state fusion pilot plant. An advanced tokamak (AT) sustains a majority of the required plasma current with effects resulting from maintenance of the peaked pressure at the device center. This current is augmented by auxiliary current drive sources. These auxiliary actuators may consist of neutral particle beams and/or radio frequency (RF) systems such as lower hybrid current drive (LHCD) and high harmonic fast wave (HHFW) current drive using radio and microwaves from antennas. The primary focus of this work is to develop models of RF current profile control suitable for use in integrated modeling frameworks and for real-time control in experiments. Direct physics models of RF current drive can be computationally intensive. In order to achieve predictive times appropriate for the thousands of calls needed in real-time control of experiments and for use in integrated models, we will apply modern machine learning (ML) techniques to accelerate these models and interpolate their results. To generate the fast and accurate models for use in control level algorithms and integrated modeling we need to replace present models with high dimensional interpolation of their results. We will perform additional simulations across a broader parameter range for EAST and other tokamaks in different physics regimes (Alcator C-Mod, DIII-D, WEST, CFETR, ARC, ITER) and combine them into a larger database for training and testing of the ML models. Further testing of the control level models with experimental current profile data from EAST and C-Mod tokamaks will provide additional confirmation of the control level model before integration in a tokamak control system or integrated modeling suite. ML will be used to optimize the selection of training data consisting of RF current driven at different values of density profile, temperature profile, plasma current, and wavenumber. ML will also be used to facilitate classification of current drive from these input data. The output of this effort will be a validated classifier capable of determining the current drive profiles for HHFW CD and LHCD on a mille-second timescale. This will provide a breakthrough capability enabling real-time control of RF driven current profiles in experiments including ITER ICRF and use integrated modeling frameworks requiring thousands of current profile calculations in discharge simulations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1977

Insulin and glucose responses during bed rest with isotonic and isometric exercise

The effects of daily intensive isotonic (68% maximum oxygen uptake) and isometric (21% maximum extension force) leg exercise on plasma insulin and glucose responses to an oral glucose tolerance test (OGTT) during 14-day bed-rest (BR) periods were investigated in seven young healthy men. The OGTT was given during ambulatory control and on day 10 of the no-exercise, isotonic, and isometric exercise BR periods during the 15-wk study. The subjects were placed on a controlled diet starting 10 days before each BR period. During BR, basal plasma glucose concentration remained unchanged with no exercise, but increased (P less 0.05) to 87-89 mg/100 ml with both exercise regimens on day 2, and then fell slightly below control levels on day 13. The fall in glucose content during BR was independent of the exercise regimen and was an adjustment for the loss of plasma volume. The intensity of the responses of insulin and glucose to the OGTT was inversely proportional to the total daily energy expenditure during BR. It was estimated that at least 1020 kcal/day must be provided by supplemental exercise to restore the hyperinsulinemia to control levels.

Dolkas, C. B.↗
NASA NTRS1995

Cell shape, cytoskeletal mechanics, and cell cycle control in angiogenesis

Capillary endothelial cells can be switched between growth and differentiation by altering cell-extracellular matrix interactions and thereby, modulating cell shape. Studies were carried out to determine when cell shape exerts its growth-regulatory influence during cell cycle progression and to explore the role of cytoskeletal structure and mechanics in this control mechanism. When G0-synchronized cells were cultured in basic fibroblast growth factor (FGF)-containing defined medium on dishes coated with increasing densities of fibronectin or a synthetic integrin ligand (RGD-containing peptide), cell spreading, nuclear extension, and DNA synthesis all increased in parallel. To determine the minimum time cells must be adherent and spread on extracellular matrix (ECM) to gain entry into S phase, cells were removed with trypsin or induced to retract using cytochalasin D at different times after plating. Both approaches revealed that cells must remain extended for approximately 12-15 h and hence, most of G1, in order to enter S phase. After this restriction point was passed, normally 'anchorage-dependent' endothelial cells turned on DNA synthesis even when round and in suspension. The importance of actin-containing microfilaments in shape-dependent growth control was confirmed by culturing cells in the presence of cytochalasin D (25-1000 ng ml-1): dose-dependent inhibition of cell spreading, nuclear extension, and DNA synthesis resulted. In contrast, induction of microtubule disassembly using nocodazole had little effect on cell or nuclear spreading and only partially inhibited DNA synthesis. Interestingly, combination of nocodazole with a suboptimal dose of cytochalasin D (100 ng ml-1) resulted in potent inhibition of both spreading and growth, suggesting that microtubules are redundant structural elements which can provide critical load-bearing functions when microfilaments are partially compromised. Similar synergism between nocodazole and cytochalasin D was observed when cytoskeletal stiffness was measured directly in living cells using magnetic twisting cytometry. These results emphasize the importance of matrix-dependent changes in cell and nuclear shape as well as higher order structural interactions between different cytoskeletal filament systems for control of capillary cell growth during angiogenesis.

Non-NASA Center↗
NASA NTRS2016

Development of a Protocol to Test Proprioceptive Utilization as a Predictor for Sensorimotor Adaptability

Astronauts returning from space flight show significant inter-subject variations in their abilities to readapt to a gravitational environment because of their innate sensory weighting. The ability to predict the manner and degree to which each individual astronaut will be affected would improve the effectiveness of countermeasure training programs designed to enhance sensorimotor adaptability. We hypothesize participant's ability to utilize individual sensory information (vision, proprioception and vestibular) influences adaptation in sensorimotor performance after space flight. The goal of this study is to develop a reliable protocol to test proprioceptive utilization in a functional postural control task. Subjects "stand" in a supine position while strapped to a backpack frame holding a friction-free device using air-bearings that allow the subject to move freely in the frontal plane, similar to when in upright standing. The frame is attached to a pneumatic cylinder, which can provide different levels of a gravity-like force that the subject must balance against to remain "upright". The supine posture with eyes closed ensures reduced vestibular and visual contribution to postural control suggesting somatosensory and/or non-otolith vestibular inputs will provide relevant information for maintaining balance control in this task. This setup is called the gravity bed. Fourteen healthy subjects carried out three trials each with eyes open alternated with eyes closed, "standing" on their dominant leg in the gravity bed environment while loaded with 60 percent of their body weight. Subjects were instructed to: "use your sense of sway about the ankle and pressure changes under the foot to maintain balance." Maximum length of a trial was 45 seconds. A force plate underneath the foot recorded forces and moments during the trial and an inertial measurement unit (IMU) attached on the backpack's frame near the center of mass of the subject recorded upper body postural responses. Series of linear and non-linear analyses were carried out on several force plate and IMU data including stabilogram diffusion analysis on the center of pressure (COP) to find a subset of parameters that were sensitive to detect differences in postural performance between eyes open and closed conditions. Results revealed that seven parameters (root mean square (RMS) of medio-lateral (ML) COP, range of ML COP, RMS of roll moment, range of trunk roll, minimum time-to-boundary (TTB), integrated TTB, and critical mean square planar displacement (delta r (sup 2) (sub c)) were significantly different between eyes open and closed conditions. We will present data to show the efficacy of using performance in single leg stance with eyes closed on the gravity bed to assess individuals' ability to utilize proprioceptive information in a functional postural control task to predict re-adaptation for sensorimotor and functional performance.

Goel, R.↗
DOE OSTI2020

Gas-Phase Complexes of Americium and Lanthanides with a Bis-triazinyl Pyridine: Reactivity and Bonding of Archetypes for F-Element Separations

Bis-triazinyl pyridines (BTPs) exhibit solution selectivity for trivalent americium over lanthanides (Ln), the origins of which remain uncertain. Here in this paper, electrospray ionization was used to generate gas-phase complexes [ML 3 ] 3+ , where M = La, Lu, or Am and L is EtBTP 2,6-bis(5,6-diethyl-1,2,4-triazin-3-yl)-pyridine. Collision-induced dissociation (CID) of [ML 3 ] 3+ in the presence of H 2 O yielded a protonated ligand [L(H)] + and hydroxide [ML 2 (OH)] 2+ or hydrate [ML(L–H)(H 2 O)] 2+ , where (L–H) - is a deprotonated ligand. Although solution affinities indicate stronger binding of BTPs toward Am 3+ versus Ln 3+ , the observed CID process is contrastingly more facile for M = Am versus Ln. To understand the disparity, density functional theory was employed to compute potential energy surfaces for two possible CID processes, for M = La and Am. In accordance with the CID results, both the rate determining transition state barrier and the net energy are lower for [AmL 3 ] 3+ versus [LaL 3 ] 3+ and for both product isomers, [ML 2 (OH)] 2+ and [ML(L–H)(H 2 O)] 2+ . More facile removal of a ligand from [AmL 3 ] 3+ by CID does not necessarily contradict stronger Am 3+ –L binding, as inferred from solution behavior. In particular, the formation of new bonds in the products can distort kinetics and thermodynamics expected for simple bond cleavage reactions. In addition to correctly predicting the seemingly anomalous CID behavior, the computational results indicate greater participation of Am 5f versus La 4f orbitals in metal–ligand bonding.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Online LIBS–ML Framework for Dynamic Characterization of Heterogeneous Waste-Derived Gasification Feedstocks

LIBS−ML framework for real time feedstock characterization during continuous conveyor transport Heterogeneous waste derived feedstocks (e.g., waste coal, biomass and blends) introduce rapid variability in heating value and ash chemistry that affect gasifier operation, yet conventional laboratory characterization techniques are too slow to support proactive control. To address this gap, this study reports on an online, in situ, dynamic characterization framework that couple’s laser-induced breakdown spectroscopy (LIBS) with leakage safe machine learning (ML) regression to deliver real time, decision quality predictions of gasifier relevant properties. A controlled sample matrix spanning two different waste coals, two different biomasses, and engineered blends under two particle size conditions were constructed and benchmarked using standardized laboratory analyses for proximate/ultimate properties and ash composition. LIBS spectra were acquired dynamically as material flowed on a conveyor belt, using high energy 1064 nm laser ablation and shot averaging to improve repeatability and precision. Supervised regression models (multi layer perceptron (MLP) /artificial neural network (ANN), random forest (RF), and support vector regression (SVR)) and an optimized weighted ensemble were trained on emission line feature sets using nested cross validation with Bayesian hyperparameter tuning and validated against an independent hold out set. The proposed LIBS−ML workflow achieves near laboratory predictive fidelity across parametric targets (including higher heating value (HHV), ash content, fixed carbon, sulfur, major ash forming oxides, and initial deformation temperature (IDT)), with the weighted ensemble providing a robust default predictor under dynamic measurement conditions. These results demonstrate a practical pathway for real time feedstock characterization that can enable feedforward adjustments and more resilient gasifier operation for variable quality waste derived fuels.

Biomass↗
DOE OSTI2026

Hierarchical Testing of a Hybrid Machine Learning‐Physics Global Atmosphere Model

Machine learning (ML)-based models have demonstrated high skill and computational efficiency, often outperforming conventional physics-based models in weather and subseasonal predictions. While prior studies have assessed their fidelity in capturing synoptic-scale atmospheric dynamics, their performance across timescales and under out-of-distribution forcing, such as +3K or +4K uniform-warming forcings, and the sources of biases remain elusive, to establish the model's reliability for Earth science. Here, we design three sets of experiments targeting synoptic-scale phenomena, interannual variability, and out-of-distribution uniform-warming forcings. We evaluate the Neural General Circulation Model (NeuralGCM), a hybrid model integrating a dynamical core with ML-based component, against observations and physics-based Earth system models (ESMs). At the synoptic scale, NeuralGCM captures the evolution and propagation of extratropical cyclones with performance comparable to ESMs. At the interannual scale, when forced by El Niño-Southern Oscillation sea surface temperature (SST) anomalies, NeuralGCM successfully reproduces associated teleconnection patterns but exhibits deficiencies in capturing nonlinear response. Under out-of-distribution uniform-warming forcings, NeuralGCM simulates similar responses in global-average temperature and precipitation and reproduces large-scale tropospheric circulation features similar to those in ESMs. Notable weaknesses include overestimating the tracks and spatial extent of extratropical cyclones, biases in the teleconnected wave train triggered by tropical SST anomalies, and differences in upper-level warming and stratospheric circulation responses to SST warming compared to physics-based ESMs. The causes of these weaknesses were explored. Despite the noted weaknesses, NeuralGCM reproduces responses across experiments reasonably and performs comparably to ESMs. By integrating a dynamical core with ML, NeuralGCM shows potential for developing ML-based ESMs.

global warming↗
DOE OSTI2025

Machine Learning for Automated Weld Quality Monitoring and Control

Resistance Spot Welding (RSW) is a critical process in the automotive industry, valued for its cost-effectiveness, short cycle time, and robustness. However, achieving consistent high-quality joints remains challenging due to the complex interplay of various factors, like materials, processes, and manufacturing uncertainties, etc. Under the collaborative project between Oak Ridge National Laboratory (ORNL) and General Motors (GM), we have developed a robust and expansible machine learning (ML) framework aimed at enhancing quality control in RSW. By harnessing the power of machine learning, we have developed the ability to ensure every aspect of the welding process, from the initial process design stage to the final weld joint quality. The framework operates by analyzing a variety of data streams, including in-line process signals, process parameters, materials, and postprocessed weld joint data. Through this analysis, the models have been trained to detect deviations from optimal quality standards, leveraging their ability to identify signature data patterns and anomalies within in-line signals and construct complex correlations between these signals and weld quality parameters. Meanwhile, the machine learning framework is designed to adapt to a variety of materials, including high strength steels and aluminum alloys, etc. Its flexible architecture facilitates the incorporation of diverse data sources and features, enabling precise modeling and prediction across a broad range of material properties and weld quality variables. The expansible ML frameworks represent a promising transformation in weld quality monitoring and control, empowering industry to achieve high levels of efficiency, consistency, and reliability in manufacturing.

42 ENGINEERING↗