Search NASA⌕ Search

SEARCH · Search NASA

Results for “machine learning operations”

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 271 records · Page 15

Opportunities in AI/ML for the Rubin LSST Dark Energy Science Collaboration

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will produce unprecedented volumes of heterogeneous astronomical data (images, catalogs, and alerts) that challenge traditional analysis pipelines. The LSST Dark Energy Science Collaboration (DESC) aims to derive robust constraints on dark energy and dark matter from these data, requiring methods that are statistically powerful, scalable, and operationally reliable. Artificial intelligence and machine learning (AI/ML) are already embedded across DESC science workflows, from photometric redshifts and transient classification to weak lensing inference and cosmological simulations. Yet their utility for precision cosmology hinges on trustworthy uncertainty quantification, robustness to covariate shift and model misspecification, and reproducible integration within scientific pipelines. This white paper surveys the current landscape of AI/ML across DESC's primary cosmological probes and cross-cutting analyses, revealing that the same core methodologies and fundamental challenges recur across disparate science cases. Since progress on these cross-cutting challenges would benefit multiple probes simultaneously, we identify key methodological research priorities, including Bayesian inference at scale, physics-informed methods, validation frameworks, and active learning for discovery. With an eye on emerging techniques, we also explore the potential of the latest foundation model methodologies and LLM-driven agentic AI systems to reshape DESC workflows, provided their deployment is coupled with rigorous evaluation and governance. Finally, we discuss critical software, computing, data infrastructure, and human capital requirements for the successful deployment of these new methodologies, and consider associated risks and opportunities for broader coordination with external actors.

Aubourg, Eric [APC, Paris] (ORCID:000000025592023X↗

Fast Solution in Sparse LDA for Binary Classification

An algorithm that performs sparse linear discriminant analysis (Sparse-LDA) finds near-optimal solutions in far less time than the prior art when specialized to binary classification (of 2 classes). Sparse-LDA is a type of feature- or variable- selection problem with numerous applications in statistics, machine learning, computer vision, computational finance, operations research, and bio-informatics. Because of its combinatorial nature, feature- or variable-selection problems are NP-hard or computationally intractable in cases involving more than 30 variables or features. Therefore, one typically seeks approximate solutions by means of greedy search algorithms. The prior Sparse-LDA algorithm was a greedy algorithm that considered the best variable or feature to add/ delete to/ from its subsets in order to maximally discriminate between multiple classes of data. The present algorithm is designed for the special but prevalent case of 2-class or binary classification (e.g. 1 vs. 0, functioning vs. malfunctioning, or change versus no change). The present algorithm provides near-optimal solutions on large real-world datasets having hundreds or even thousands of variables or features (e.g. selecting the fewest wavelength bands in a hyperspectral sensor to do terrain classification) and does so in typical computation times of minutes as compared to days or weeks as taken by the prior art. Sparse LDA requires solving generalized eigenvalue problems for a large number of variable subsets (represented by the submatrices of the input within-class and between-class covariance matrices). In the general (fullrank) case, the amount of computation scales at least cubically with the number of variables and thus the size of the problems that can be solved is limited accordingly. However, in binary classification, the principal eigenvalues can be found using a special analytic formula, without resorting to costly iterative techniques. The present algorithm exploits this analytic form along with the inherent sequential nature of greedy search itself. Together this enables the use of highly-efficient partitioned-matrix-inverse techniques that result in large speedups of computation in both the forward-selection and backward-elimination stages of greedy algorithms in general.

Moghaddam, Baback↗

AI/ML Support for LPSD Project

The slides presents a new approach to identify and characterize nuclear plant shutdown initiating events from operating experience data by using machine learning techniques. The slides were prepared for a presentation in an upcoming DOE/NRC MOU AI/ML meeting in September 2022.

99 GENERAL AND MISCELLANEOUS↗

Autotuning of Double-Dot Devices In Situ with Machine Learning

The current practice of manually tuning quantum dots (QDs) for qubit operation is a relatively time-consuming procedure that is inherently impractical for scaling up and applications. In this work, we report on the in situ implementation of a recently proposed autotuning protocol that combines machine learning (ML) with an optimization routine to navigate the parameter space. In particular, we show that a ML algorithm trained using exclusively simulated data to quantitatively classify the state of a double-QD device can be used to replace human heuristics in the tuning of gate voltages in real devices. Here, we demonstrate active feedback of a functional double-dot device operated at millikelvin temperatures and discuss success rates as a function of the initial conditions and the device performance. Modifications to the training network, fitness function, and optimizer are discussed as a path toward further improvement in the success rate when starting both near and far detuned from the target double-dot range.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Use of Sensors and Machine Learning for Signal Discovery in a Solvent Extraction Process – 23280

Reprocessing is an important step in the nuclear fuel cycle where usable nuclear materials are extracted from spent fuel for recycling. The extraction of materials for reuse simultaneously reduces not only the volume of nuclear waste, but its decay time to radioactivity levels similar to that of the originating uranium ore. As part of an initiative to steward research, development, and innovation into the nuclear fuel cycle, Idaho National Laboratory is designing and constructing a solvent extraction testbed named Beartooth. Beartooth will allow researchers to refine extraction processes, test innovative extraction processes, and give early career scientists opportunities to gain skills in performing separations chemistry that uses centrifugal contactors. In addition, the Beartooth testbed is being uniquely designed to enable novel technologies including machine learning capabilities for the characterization of chemical process operations in near real-time. To aid in the design of Beartooth, a team of researchers are installing a variety of atypical sensors into a system of contactors for signal discovery. The team will implement machine learning methods on sensor data to determine signal features with the goal of providing a process operator with a deeper understanding of the chemical process and equipment usage. This work will summarize sensors utilized and preliminary results from an infrared camera.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Learning plasma dynamics and robust rampdown trajectories with predict-first experiments at TCV

The rampdown phase of a tokamak pulse is difficult to simulate and often exacerbates multiple plasma instabilities. To reduce the risk of disrupting operations, we leverage advances in Scientific Machine Learning (SciML) to combine physics with data-driven models, developing a neural state-space model (NSSM) that predicts plasma dynamics during Tokamak à Configuration Variable (TCV) rampdowns. The NSSM efficiently learns dynamics from a modest dataset of 311 pulses with only five pulses in a reactor-relevant high-performance regime. The NSSM is parallelized across uncertainties, and reinforcement learning (RL) is applied to design trajectories that avoid instability limits. High-performance experiments at TCV show statistically significant improvements in relevant metrics. A predict-first experiment, increasing plasma current by 20% from baseline, demonstrates the NSSM’s ability to make small extrapolations. The developed approach paves the way for designing tokamak controls with robustness to considerable uncertainty and demonstrates the relevance of SciML for fusion experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Use of Sensors and Machine Learning for Signal Discovery in a Solvent Extraction Process

Reprocessing is an important step in the nuclear fuel cycle where usable nuclear materials are extracted from used fuel for recycling. The separation of materials for reuse simultaneously reduces not only the volume of nuclear waste, but its decay time to radioactivity levels similar to that of the originating uranium ore. As part of an initiative to steward research, development, and innovation into the nuclear fuel cycle, Idaho National Laboratory is designing and constructing a solvent extraction testbed named Beartooth. This testbed will allow researchers to refine separation processes, test innovative extraction processes, and give early career scientists opportunities to gain skills in performing separations chemistry utilizing centrifugal contactors. In addition, the Beartooth testbed is being uniquely designed to enable novel technologies including machine learning capabilities for the characterization of chemical process operations in near real-time. To aid in the design of Beartooth, a team of researchers are installing a variety of atypical sensors into a system of contactors for signal discovery. The team will implement machine learning methods on acquired sensor data to extract signal features. The goal is to provide a process operator with a deeper understanding of the chemical process and equipment usage. This work will summarize sensors utilized and preliminary results from an infrared camera.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Novel artificial neural network model for instantaneous power losses and operational efficiency mapping of MW-scale vanadium redox flow battery for improved technoeconomic analysis

A novel data-driven, machine-learning-based method for modeling the instantaneous power losses of a distribution-sited 2 MW/8MWh vanadium redox flow battery (VRFB), a grid-scale electrochemical storage technology, is introduced and compared against benchmark empirical modeling approaches, including symmetric and asymmetric models, as well as a recent convex hull modeling approach. The novel loss modeling method introduces several advantages over the benchmark models and over simplistic efficiency estimates, the most significant of which is that the model can accurately reflect the stepwise and non-linear parasitic losses associated with the duty cycles of mechanical auxiliary systems like pump motor drives and blower fans. Residuals of the models are compared; the proposed data driven model features significantly improved accuracy over the benchmark models. The model's coefficient of determination is also improved relative to that of the benchmark models. Furthermore, a novel method for visualization of operational efficiency of the grid-scale storage technology is introduced. To demonstrate the benefits of the novel data-driven method for modeling the VRFB, the benchmark models and the proposed models are embedded into an Open DSS distribution network model to study two applications of the grid-scale electrical storage system: load leveling for grid support and energy arbitrage. This article demonstrates that the accuracy of the instantaneous power loss model significantly impacts the understanding of the state of charge of the VRFB. In turn, the accuracy of the efficiency modeling of the VRFB impacts the understanding of the potential economic value and technical benefits to the distribution network operators. In conclusion, the presented power loss modeling approach is, therefore, highly relevant for utility-stakeholders, battery asset owners, system engineers, system designers, and financial planners interested in evaluating or optimizing the operation of grid-scale VRFBs.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Physics-informed machine learning of the correlation functions in bulk fluids

The Ornstein–Zernike (OZ) equation is the fundamental equation for pair correlation function computations in the modern integral equation theory for liquids. In this work, machine learning models, notably physics-informed neural networks and physics-informed neural operator networks, are explored to solve the OZ equation. The physics-informed machine learning models demonstrate great accuracy and high efficiency in solving the forward and inverse OZ problems of various bulk fluids. The results highlight the significant potential of physics-informed machine learning for applications in thermodynamic state theory.

97 MATHEMATICS AND COMPUTING↗

Conceptual Design of Integrated Energy Systems with Market Interaction Surrogate Models

Most integrated energy system (IES) optimization frameworks employ the price-taker approximation, which ignores important interactions with the market and can result in overestimated economic values. In this work, we propose a machine learning surrogate-assisted optimization framework to quantify IES/market interactions and thus go beyond price-taker. We use time series clustering to generate representative IES operation profiles for the optimization problem and use machine learning surrogate models to predict the IES/market interaction. We quantify the accuracy of the time series clustering and surrogate models in a case study to optimally retrofit a nuclear power plant with a polymer electrolyte membrane electrolyzer to co-produce electricity and hydrogen.

Chen, Xinhe↗

Report for 2.2.1 Task 5: Develop and Document a State-Based Alarm System for a Nuclear Power Plant Control Room Using Machine Learning. Light Water Reactor Sustainability Program report

In legacy control rooms the operators might have 20 to 30 alarms to deal with on 12 control boards, during a reactor trip. The new digital alarm system in the upgraded control room will display between 150 and 200 alarm points on a single screen. This overload of information prevents the operators from identifying any abnormal alarms quickly without having to scroll through the alarm list looking for the abnormal alarms. One way to tackle this problem could be to attempt to create a state-based alarm system which would identify plant states and would identify which alarms are expected in those states and suppress them from the alarm screen so that any unexpected alarms would be displayed and easily identified to the operators. This can be done today but is very labor intensive in that an operator or training instructor would be required to run scenarios and log the alarms to be suppressed. The hope is that Machine Learning (ML) can provide an alternative solution with less manual effort required.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Machine Learning 5G Attack Detection in Programmable Logic

Machine learning-assisted network security may significantly contribute to securing 5G components. However, machine learning network security inference speeds generally require tens to hundreds of milliseconds thereby introducing significant latency in 5G operations. The inference latency can be reduced by deploying the machine learning model to programmable logic in a field programmable gate array (FPGA) at the cost of a small loss in accuracy. In order to quantify this loss, as well as to establish baseline performance inference speeds for programmable logic implementations, this work explores an autoencoder and a ß-variational autoencoder deployed on two different FPGA evaluation boards and compares accuracy and performance against an NVIDIA A100 GPU implementation. A publicly available 5G dataset containing 10 types of attacks along with normal traffic is introduced as part of the evaluation.

97 MATHEMATICS AND COMPUTING↗

Anomaly Detection in Accelerator Facilities Using Machine Learning

Synchrotron light sources are user facilities and usually run about 5000 hours per year to support many beamlines operations in parallel. Reliability is a key parameter to evaluate machine performance. Even many facilities have achieved >95% beam reliability, there are still many hours of unscheduled downtime and every hour lost is a waste of operation costs along with a big impact on individual scheduled user experiments. Preventive maintenance on subsystems and quick recovery from machine trips are the basic strategies to achieve high reliability, which heavily depends on experts’ dedication. Recently, SLAC, APS, and NSLS-II collaborated to develop machine-learning-based approaches aiming to solve both situations, hardware failure prediction and machine failure diagnosis to find the root sources. In this paper, we report our facility operation status, development progress, and plans.

Accelerator Physics↗

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗

A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction

Airport configuration selection is a complex decision-making process that involves several operational and human factors. In this paper we propose a novel recursive multi-step machine learning (ML) approach to predict airport configuration. The multi-step approach guarantees stability of the predicted configuration by taking as input the configuration predicted at the previous time step. The features of the proposed model include weather data, future arrival and departure counts and current configuration. Due to the importance of arrival and departure counts in predicting the airport configuration, arrival counts are calculated using landing time predictions selected from physics-based landing time predictions available in FAA System Wide Information Management data feeds for each flight. The selection rules were developed and refined to select the most accurate time for different phases of flight. The proposed model predicts the airport configurations up to 6 hours ahead. In this paper we show the predictive performance of the proposed model for six major US airports, including Charlotte Douglas International Airport (CLT), Dallas/Fort Worth International Airport (DFW), John F. Kennedy International Airport (JFK), Newark Liberty International Airport (EWR), LaGuardia Airport (LGA) and Dallas Love Field Airport (DAL). We trained and evaluated models on 2019 and 2020 data in order to study the effect of the pandemic and how changes in traffic patterns affected the performance of the proposed model. Results are compared with a baseline assuming no airport configuration changes. In our results for DFW, we obtained a prediction accuracy of 89.3% for 3 hours ahead prediction, and 82.8% for 6 hours ahead when applied on 2019 data.

machine learning↗