SEARCH · Search NASA
Results for “Learning”
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.
Insights into the structure and dynamics of K + ions at the muscovite–water interface from machine learning potential simulations
The surfaces of many minerals are covered by naturally occurring cations that become partially hydrated and can be replaced by hydronium or other cations when the surface is exposed to water or an aqueous solution. These ion exchange processes are relevant to various chemical and transport phenomena, yet elucidating their microscopic details is challenging for both experiments and simulations. Here, in this work, we make a first step in this direction by investigating the behavior of the native K + ions at the interface between neat water and the muscovite mica (001) surface with ab-initio -based machine learning molecular dynamics and enhanced sampling simulations. Our results show that the desorption of the surface K + ions in pure ion-free water has a significant free energy barrier irrespective of their local surface arrangement. In contrast, facile K + diffusion between mica’s ditrigonal cavities characterized by different Al/Si orderings is observed. This behavior suggests that the K + ions may favor a dynamic disordered surface arrangement rather than complete desorption when exposed to deionized water.
Reinforcement Learning-Based Oscillation Dampening: Scaling Up Single-Agent Reinforcement Learning Algorithms to a 100-Autonomous-Vehicle Highway Field Operational Test
In this article, we explore the technical details of the reinforcement learning (RL) algorithms that were deployed in the largest field test of automated vehicles designed to smooth traffic flow in history as of 2023, uncovering the challenges and breakthroughs that come with developing RL controllers for automated vehicles. We delve into the fundamental concepts behind RL algorithms and their application in the context of self-driving cars, discussing the developmental process from simulation to deployment in detail, from designing simulators to reward function shaping. We present the results in both simulation and deployment, discussing the flow-smoothing benefits of the RL controller. From understanding the basics of Markov decision processes to exploring advanced techniques such as deep RL, our article offers a comprehensive overview and deep dive of the theoretical foundations and practical implementations driving this rapidly evolving field. We also showcase real-world case studies and alternative research projects that highlight the impact of RL controllers in revolutionizing autonomous driving. From tackling complex urban environments to dealing with unpredictable traffic scenarios, these intelligent controllers are pushing the boundaries of what automated vehicles can achieve. Furthermore, we examine the safety considerations and hardware-focused technical details surrounding deployment of RL controllers into automated vehicles. As these algorithms learn and evolve through interactions with the environment, ensuring their behavior aligns with safety standards becomes crucial. Here, we explore the methodologies and frameworks being developed to address these challenges, emphasizing the importance of building reliable control systems for automated vehicles.
Lessons Learned from the Energy to Communities (E2C) Peer-Learning Cohort on Engaging with Electric Utilities for Successful Local Partnerships
NLR designed and led a six-month peer-learning cohort from July through December 2025 on “Engaging With Electric Utilities for Successful Local Partnerships” as part of the U.S. Department of Energy’s Energy to Communities (E2C) program. Representatives from 15 local and regional governments from across the United States participated in monthly workshops covering best practices for engaging with their utilities to advance their own energy goals. This document shares key takeaways, lessons learned, and resources from the cohort that may be useful to local governments, regional governments, or Tribes.
Low Activity Waste Glass Optimization with Property Models from Machine Learning, Part 2: Experimental Validation and Active Learning
The United States Department of Energy is responsible for managing legacy nuclear waste stored in underground tanks at the Hanford Site. To treat the waste, it is planned as the current baseline to separately vitrify low-activity waste (LAW) and high-level waste fractions. Previously, machine learning (ML) based glass property models (e.g., chemical durability, viscosity, electrical conductivity and SO3 solubility) were developed with prediction uncertainties. A waste glass optimization approach was then established to enable the capability of using these ML models in LAW glass formulation. In this study, the previous ML models were first experimentally validated, and the results were incorporated back into the database to update the ML models. The updated models and formulations showed increased waste loading while reducing the failure rate, demonstrating improved predictive accuracy, reduced uncertainties, and the effectiveness of active learning in guiding high-dimensional, nonlinear LAW glass design. This represents the first experimental validation of ML based LAW glass formulation, with practical benefits such as higher waste loading, shorter mission duration, and lower operational risk.
Using Active Learning to Rapidly Develop Machine Learned Diffusion Coefficients of CO 2 Conversion Reagents in Metal–Organic Frameworks
Here, we used a combined molecular dynamics/active learning (AL) approach to create machine learning models that can predict the diffusion coefficient of epichlorohydrin and chloropropene carbonate, the reactant and product of a common CO 2 cycloaddition reaction, in metal–organic frameworks (MOFs). Nanoporous MOFs are effective catalysts for the cycloaddition of CO 2 to epoxides. The diffusion rates within nanoporous catalysts can control the rate of reaction as the reactants and products must diffuse to the active sites within the MOF and then out of the nanoporous material for reusability. However, the diffusion process is routinely ignored when searching for new materials in catalytic applications. Here we verified improvement during the AL process by consistently tracking metrics on the same groups of MOFs to ensure consistency. Metal identity was found to have little impact on diffusion rates, while structural features like pore limiting diameter act as a threshold where a minimum value is needed for high diffusion rates. We identified the MOFs with the highest epichlorohydrin and chloropropene carbonate diffusion coefficients which can be used for further studies of reaction energetics.
Hierarchical transfer learning: an agile and equitable strategy for machine-learning interatomic models
Machine-learned interatomic models are growing in popularity due to their ability to afford near quantum-accurate predictions for complex phenomena with orders-of-magnitude greater computational efficiency. However, these models struggle when applied to systems of many element types due to the approximately exponential increase in number of parameters that must be determined. To mitigate this challenge, we present a new hierarchical transfer learning approach that allows the fitting problem to be decomposed into smaller independent and reusable parameter blocks that enable development of explicitly chemically extensible ML-IAM. Application of this strategy is demonstrated for C and N mixtures under conditions ranging from nominally ambient to ~10,000 K and 200 GPa for compositions from 0 to 100% N. Ultimately, this strategy makes model generation for chemically complex systems more tractable and efficient, facilitates comprehensive model validation, and makes ML-IAM development for problems of this nature more accessible to users with limited access to extreme computing infrastructure.
Automated ICRF heating surrogate modeling via machine learning
This work introduces automated machine learning workflows that address critical bottlenecks in surrogate model development for Ion Cyclotron Range of Frequencies (ICRF) heating applications. The automated framework includes data analysis tools that transform raw datasets into actionable insights in seconds, replacing weeks of manual exploratory effort and ensuring consistent, reproducible dataset characterization. By integrating advanced hyperparameter optimization (HPO) methods including Bayesian optimization via BoTorch and Tree-structured Parzen Estimators (TPE), the framework significantly reduces model development time from weeks to hours, decreasing computational cost and required expertise, while enabling high-accuracy surrogate models. Compared to traditional hyperparameter scanning (HPS) techniques such as methodical, randomized, and grid searches, HPO methods achieve superior convergence and predictive performance, even when compared to already well-tuned reference models. On NSTX High Harmonic Fast Wave (HHFW) heating datasets, both Random Forest Regressor (RFR) and neural network surrogates demonstrate improved accuracy, achieving R 2 values beyond 0.97 and 0.98, respectively. The results show that while HPO gains are modest for robust architectures like RFR, they become essential for more sensitive models such as neural networks, highlighting the trade-offs across optimization strategies. Through automated workflows that eliminate manual hyperparameter tuning and require minimal ML expertise, this work enables widespread adoption of high-fidelity surrogate models across the fusion community for real-time plasma control, uncertainty quantification, rapid experimental scenario development, and integrated system optimization.
Privacy Preserving Federated Learning for Advanced Scientific Ecosystems
We present a framework to provide privacy preserving (PP) federating learning (FL) across multiple computational and experimental facilities. This work joins the compute capabilities of National Energy Research Scientific Computing Center (NERSC) and Oak Ridge National Laboratory Research Cloud (ORC) with simulated experimental data, such as those produced at the SLAC National Accelerator Laboratory and Spallation Neutron Source (SNS). We describe the software infrastructure developed to provide privacy for computational and experimental networks. We developed algorithmic privacy across the federated system by embedding database security, computation, and communication into the federation architecture, utilizing scientific tools developed by the experimental community.
DP-TwoLevel: two-stage gradient subspace learning for differentially private federated learning
Federated learning (FL) enables collaborative model training across distributed data sources without sharing raw data, but faces fundamental challenges in communication efficiency and privacy. Differentially private (DP) training mitigates information leakage but introduces noise that degrades model performance, especially in high-dimensional settings. We propose DP-TwoLevel, a hierarchical gradient projection method that improves utility under fixed DP constraints by exploiting low-dimensional structure in model updates. Our approach learns a two-level PCA-based representation of gradients and applies DP noise in a reduced-dimensional subspace, thereby lowering the effective noise magnitude while preserving dominant signal components. We evaluate the method across three datasets (MNIST, Fashion-MNIST, CIFAR-10) and three privacy regimes (ϵ∈0.5, 1.0, 2.0). Across nine experimental settings, DP-TwoLevel consistently outperforms DP-FedAvg, achieving an average accuracy improvement of 9.44%, with larger gains observed in lower ϵ(higher-noise) regimes (up to +22.31%). We further analyze scalability across models ranging from 100K to 1.49M parameters and identify a variance-based success criterion: performance remains strong when the projection preserves more than 75% of gradient variance, degrades in a marginal regime (65–75%), and fails below this threshold. Our results demonstrate that structure-aware dimensionality reduction can significantly improve the privacy–utility tradeoff in FL without modifying formal privacy guarantees. We also provide empirical evidence of scaling limitations for global projections and motivate per-layer extensions for larger models.
SRF cavity instability detection with machine learning at CEBAF
During the operation of the Continuous Electron Beam Accelerator Facility (CEBAF), one or more unstable superconducting radio-frequency (SRF) cavities often cause beam loss trips while the unstable cavities themselves do not necessarily trip off. The present RF controls for the legacy cavities report at only 1 Hz, which is too slow to detectfast transient instabilities during these trip events. These challenges make the identification of an unstable cavity out of the hundreds installed at CEBAF a difficult and time-consuming task. To tackle these issues, a fast data acquisition system (DAQ) for the legacy SRF cavities has been developed, which records the sample at 5 kHz. An unsupervised learning framework has been developed to identify anomalous SRF cavity behavior. We will discuss the present status of the DAQ system and our framework, along with recent successes in detecting anomalous cavity behavior. Overall, our method offers a practical solution for identifying unstable SRF cavities, contributing to increased beam availability and machine reliability.
Developing predictive models for µ opioid receptor binding using machine learning and deep learning techniques
Opioids exert their analgesic effect by binding to the µ opioid receptor (MOR), which initiates a downstream signaling pathway, eventually inhibiting pain transmission in the spinal cord. However, current opioids are addictive, often leading to overdose contributing to the opioid crisis in the United States. Therefore, understanding the structure-activity relationship between MOR and its ligands is essential for predicting MOR binding of chemicals, which could assist in the development of non-addictive or less-addictive opioid analgesics. This study aimed to develop machine learning and deep learning models for predicting MOR binding activity of chemicals. Chemicals with MOR binding activity data were first curated from public databases and the literature. Molecular descriptors of the curated chemicals were calculated using software Mold2. The chemicals were then split into training and external validation datasets. Random forest, k-nearest neighbors, support vector machine, multi-layer perceptron, and long short-term memory models were developed and evaluated using 5-fold cross-validations and external validations, resulting in Matthews correlation coefficients of 0.528–0.654 and 0.408, respectively. Furthermore, prediction confidence and applicability domain analyses highlighted their importance to the models’ applicability. Our results suggest that the developed models could be useful for identifying MOR binders, potentially aiding in the development of non-addictive or less-addictive drugs targeting MOR.
Lessons Learned from the Clean Energy to Communities (C2C) Peer-Learning Cohort Enhancing Resilience at Critical Facilities through Solar, Storage, and Microgrids
The National Renewable Energy Laboratory (NREL), with support from World Resources Institute (WRI), designed and led a six-month peer-learning cohort from January through July 2024 on "Enhancing Resilience at Critical Facilities through Solar, Storage, and Microgrids" as part of the U.S Department of Energy's Clean Energy to Communities (nrel.gov/c2c) (C2C) program. Representatives from 15 municipalities, municipal utilities, colleges, and Tribes from across the United States participated in monthly workshops covering best practices for planning and deploying local resilience projects. This document shares key takeaways, lessons learned, and resources from the cohort including considerations for planning, proposing, designing, procuring, and funding microgrid projects for resilience.
Lessons Learned from the Clean Energy to Communities (C2C) Peer-Learning Cohort Evaluating and Prioritizing Municipal Buildings for Energy Efficiency and Decarbonization Investment
The National Renewable Energy Laboratory (NREL), with support from World Resources Institute (WRI), designed and led a six-month peer-learning cohort from January through July 2024 on "Enhancing Resilience at Critical Facilities through Solar, Storage, and Microgrids" as part of the U.S Department of Energy's Clean Energy to Communities (nrel.gov/c2c) (C2C) program. Representatives from 15 municipalities, municipal utilities, colleges, and Tribes from across the United States participated in monthly workshops covering best practices for planning and deploying local resilience projects. This document shares key takeaways, lessons learned, and resources from the cohort.
Lessons Learned from the Clean Energy to Communities (C2C) Peer-Learning Cohort Incorporating Community Priorities into Electric Vehicle Plans and Projects
The National Renewable Energy Laboratory (NREL), with support from World Resources Institute (WRI), designed and led a six-month peer-learning cohort from January through July 2024 on "Enhancing Resilience at Critical Facilities through Solar, Storage, and Microgrids" as part of the U.S Department of Energy's Clean Energy to Communities (nrel.gov/c2c) (C2C) program. Representatives from 15 municipalities, municipal utilities, colleges, and Tribes from across the United States participated in monthly workshops covering best practices for planning and deploying local resilience projects. This document shares key takeaways, lessons learned, and resources from the cohort including key steps that can be taken to help entities design effective community engagement processes around clean mobility.
Machine Learning Accelerated First-Principles Study of the Hydrodeoxygenation of Propanoic Acid
The complex reaction network of catalytic biomass conversions often involves hundreds of surface intermediates and thousands of reaction steps, greatly hindering the rational design of metal catalysts for these conversions. Here, we present a framework of machine learning (ML)-accelerated first-principles studies for the hydrodeoxygenation (HDO) of propanoic acid over transition metal surfaces. The microkinetic model (MKM) is initially parametrized by ML-predicted energies and iteratively improved by identifying the rate-determining species and steps (RDS), computing their energies by density functional theory (DFT), and reparameterizing the MKM until all the RDS are computed by DFT. The Gaussian process (GP) model performs significantly better than the linear ridge regression model for predicting both the adsorption free energies and transition state free energies. Parameterized with energies from the GP model, only 5–20% of the full reaction network has to be computed by DFT for the MKM to possess DFT-level accuracy for the TOF and dominant reaction pathway. While the linear ridge regression model performs worse than the GP model, its performance is greatly improved when only transition states are predicted by the regression model and adsorption energies are computed by DFT. Overall, we find that a high accuracy in adsorption free energies is more important for a reliable MKM than a high accuracy in TS free energies. Lastly, based on the GP model with GOH and GCHCHCO as catalyst descriptors, we build two-dimensional volcano plots in activity and selectivity that can help design promising alloy catalysts for HDO reactions of organic acids.
Identifying divertor detachment using a machine learning model trained on divertor camera images from DIII-D
Explore the source record for details and available documents.
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.