Search NASA⌕ Search

SEARCH · Search NASA

Results for “operator 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.

At least 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Application of deep learning methods for beam size control during user operation at the Advanced Light Source

Past research at the Advanced Light Source (ALS) provided a proof-of-principle demonstration that deep learning methods could be effectively employed to compensate for the significant perturbations to the transverse electron beam size induced by user-controlled adjustments of the insertion devices. However, incorporating these methods into the ALS’ daily operations has faced notable challenges. The complexity of the system’s operational requirements and the significant upkeep demands has restricted their sustained application during user operation. Here, we introduce the development of a more robust neural network (NN)-based algorithm that utilizes a novel online fine-tuning approach and its systematic integration into the day-to-day machine operations. Our analysis emphasizes the process of NN model selection, demonstrates the superior performance of the NN-based method over traditional feedback methods, and examines the effectiveness and resilience of the new algorithm during user-operation scenarios. Published by the American Physical Society 2024

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

When ChatGPT Meets Vulnerability Management: The Good, the Bad, and the Ugly

Vulnerability management is a very challenging and time-consuming task. For many organizations, security operators need to learn about the properties of vulnerabilities to prioritize and mitigate them. Due to the lack of automated tools for vulnerability assessment, operators usually manually search for and read related information from sources online. Recent advances in large language models, like ChatGPT, open up an opportunity for time savings and may prompt operators to use these models as vulnerability information sources. In this work, we evaluate the ability of ChatGPT and several of its siblings to accurately answer user questions about vulnerability properties as well as to provide information for how to mitigate a vulnerability. We also explore their summarization capabilities when multiple vulnerability advisory documents are provided. We find that the models perform poorly on information retrieval tasks, but they perform quite well on summarization.

McClanahan, Kylie↗
DOE OSTI2024

Updates and Lessons Learned from NuMI Beamline at Fermilab

The Neutrinos at the Main Injector (NuMI) beamline at Fermilab generates an intense muon neutrino beam for the NOvA (NuMI Off-axis 𝜈𝑒 Appearance) long baseline neutrino experiment. Over the years, the NuMI beamline has been pivotal in advancing neutrino physics, providing invaluable data and insights. This presentation offers updates and a comprehensive review of the lessons learned from the operation, maintenance, and monitoring of the NuMI beamline. Key topics include the optimization of beam performance, challenges in maintaining beamline stability, and proposed Machine Learning implementations to enhance monitoring. The talk aims to share best practices and provide a roadmap for future beamline projects, including the Long-Baseline Neutrino Facility (LBNF).

Wickremasinghe, Athula↗
DOE OSTI2025

Physics-informed latent neural operator for real-time predictions of time-dependent parametric PDEs

Deep operator network (DeepONet) has shown significant promise as surrogate models for systems governed by partial differential equations (PDEs), enabling accurate mappings between infinite-dimensional function spaces. However, when applied to systems with high-dimensional input-output mappings arising from large numbers of spatial and temporal collocation points, these models often require heavily overparameterized networks, leading to long training times. Latent DeepONet addresses some of these challenges by introducing a two-step approach: first learning a reduced latent space using a separate model, followed by operator learning within this latent space. While efficient, this method is inherently data-driven and lacks mechanisms for incorporating physical laws, limiting its robustness and generalizability in data-scarce settings. Here, in this work, we propose PI-Latent-NO, a physics-informed latent neural operator framework that integrates governing physics directly into the learning process. Our architecture features two coupled DeepONets trained end-to-end: a Latent-DeepONet that learns a low-dimensional representation of the solution, and a Reconstruction-DeepONet that maps this latent representation back to the physical space. By embedding PDE constraints into the training via automatic differentiation, our method eliminates the need for labeled training data and ensures physics-consistent predictions. The proposed framework is both memory and compute-efficient, exhibiting near-constant scaling with problem size and demonstrating significant speedups over traditional physics-informed operator models. We validate our approach on a range of parametric PDEs, showcasing its accuracy, scalability, and suitability for real-time prediction in complex physical systems.

Latent representations↗
DOE OSTI2025

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.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

A Reinforcement Learning Approach to Augment Conventional PID Control in Nuclear Power Plant Transient Operation

The ability of nuclear reactors to operate their power conversion cycles more flexibly will enhance their value to energy grids with variable pricing. Current nuclear control systems are typically classical controllers that are often based on proportional-integral-derivative (PID) control. This paper presents a method of augmenting the existing PID control for difficult transient operations in nuclear power plants using a reinforcement learning–derived feedforward signal applied in real time. The agents, which are trained on a test thermal load-following problem, are designed to improve steam generator outlet temperature control for a range of fast load-following scenarios covering ramp rates from 9%/min to 15%/min. Several reinforcement learning algorithms were initially investigated for the training of the feedforward agents with deep Q-learning (DQN) and proximal policy optimization (PPO) networks, which were found to be the most promising. The DQN controllers utilize discrete actions, giving them a better disturbance rejection at steady state but inconsistent response to initial temperature deviations. In contrast, PPO-trained agents, which take continuous actions except for a dead zone around zero, were shown to have the best combination of high disturbance rejection at steady state and good tracking of the desired temperature value. The ability of the PPO agent was also examined, with the average time of decision making found to be on the order of 1 ms. The fault properties of the controller under the loss of the reinforcement learning agent feedforward signal were also examined. The controller showed strong performance in situations of “no-signal” faults. but was less good at handling “stuck-at” faults, where the feedforward signal remains at a set value. In both cases, however, the PID was able to successfully maintain stability, eventually returning the system to a steady state. It is hoped that this work will allow for the proposed control architecture to be examined for more difficult control problems such that it may eventually be used to adapt existing nuclear plants for more aggressive load-following on grids of the future.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2026

Operating advanced scientific instruments with AI agents that learn on the job

Advanced scientific user facilities, such as next generation X-ray light sources and self-driving laboratories, are revolutionizing scientific discovery by automating routine tasks and enabling rapid experimentation and characterizations. However, these facilities must continuously evolve to support new experimental workflows, adapt to diverse user projects, and meet growing demands for more intricate instruments and experiments. This continuous development introduces significant operational complexity, necessitating a focus on usability, reproducibility, and intuitive human-instrument interaction. In this work, we explore the integration of agentic AI, powered by Large Language Models (LLMs), as a transformative tool to achieve this goal. We present our approach to developing a human-in-the-loop pipeline for operating advanced instruments including an X-ray nanoprobe beamline and an autonomous robotic station dedicated to the design and characterization of materials. Specifically, we evaluate the potential of various LLMs as trainable scientific assistants for orchestrating complex, multi-task workflows, which also include multimodal data, optimizing their performance through optional human input and iterative learning. We demonstrate the ability of AI agents to bridge the gap between advanced automation and user-friendly operation, paving the way for more adaptable and intelligent scientific facilities.

Large Language Models↗
DOE OSTI2026

Latent Twins

Over the past decade, scientific machine learning has transformed the development of mathematical and computational frameworks for analyzing, modeling, and predicting complex systems. From inverse problems to numerical partial differential equations (PDEs), dynamical systems, and model reduction, these advances have pushed the boundaries of what can be simulated. Yet they have often progressed in parallel, with representation learning and algorithmic solution methods evolving largely as separate pipelines. With Latent Twins, we propose a unifying mathematical framework that creates a hidden surrogate in latent space for the underlying equations. Whereas digital twins mirror physical systems in the digital world, Latent Twins mirror mathematical systems in a learned latent space governed by operators. Through this lens, classical modeling, inversion, model reduction, and operator approximation all emerge as special cases of a single principle. We establish the fundamental approximation properties of Latent Twins for both ordinary differential equations (ODEs) and PDEs and demonstrate the framework across three representative settings: (i) canonical ODEs, capturing diverse dynamical regimes; (ii) a PDE benchmark using the shallow-water equations, contrasting Latent Twin simulations with deep operator network and forecasts with a four-dimensional variational method baseline; and (iii) a challenging real-data geopotential reanalysis dataset, reconstructing and forecasting from sparse, noisy observations. Latent Twins provide a compact, interpretable surrogate for solution operators that evaluate across arbitrary time gaps in a single-shot, while remaining compatible with scientific pipelines such as assimilation, control, and uncertainty quantification. Looking forward, this framework offers scalable, theory-grounded surrogates that bridge data-driven representation learning and classical scientific modeling across disciplines.

Latent Twins↗
DOE OSTI2025

Reinforcement Learning for Anomaly Detection in Nuclear Power Plant Operation and Maintenance

In nuclear power plants (NPPs), timely identification of sensor and human errors is critical to ensure safe and efficient plant operations. Anomaly detection models can be employed for this task. However, traditional anomaly detection approaches may have high dependency on labeled datasets and struggle with adaptability in complex, dynamic environments. Reinforcement learning (RL) has demonstrated significant potential in fault diagnosis and anomaly detection; however, its application to anomaly detection in NPPs remains a relatively underexplored research direction. Hence, to address this gap, in this study, we present a novel physics-informed reinforcement learning model, PIRL-AD: Physics-Informed Reinforcement Learning for Anomaly Detection, that integrates domain knowledge from calorimetric equations into the RL framework for enhanced sensor and human error anomaly detection. We evaluate the performance of PIRL-AD against a non-physics informed RL benchmark and a support vector machine (SVM) on data collected from a forced flow loop testbed. Experimental results suggest that PIRL-AD outperforms other baselines on a range of anomalous datasets that include both sensor and human-induced anomalies across key performance metrics, statistically outperforming the RL and SVM benchmarks with respect to geometric mean (respectively, 92.96% vs. 91.06% vs. 83.01%) and F1-score (respectively, 89.23% vs. 86.98% vs. 77.01%). Furthermore, the findings suggest the potential of physics-integrated reinforcement learning models for enhanced anomaly detection performance in NPPs.

Reinforcement learning↗
DOE OSTI2026

MSD CoP Webinar: Modeling the Operations of Reservoir Systems with LLMs and Inverse Reinforcement Learning

Context: This panel featured three presentations centered on the common theme of applying LLMs and inverse reinforcement learning (IRL) to capture the complex human-environment interactions that are central to the operation of reservoir systems. Dr. Wyatt Arnold will kick off the webinar with a talk on how analyzing LLM chain-of-thought reasoning reveals sophisticated quantitative justification and risk awareness, showing promise as a bridge between quantitative models and value-driven water management decisions. Next, Dr. Matteo Giuliani will build on this with a discussion demonstrating that AI- and IRL-driven approaches can infer the trade-offs between flood control and water supply using historical observations. Finally, Dr. Rohan Singh Wilkho will close the webinar with a talk establishing IRL as a generalizable diagnostic tool for decoding decision-making in managed hydrologic and human-infrastructure systems. Across the three presentations, the application of LLMs and IRL opens new possibilities for the development of adaptive, transparent, and human-aware models supporting water management in an increasingly uncertain future. Presenters: Wyatt Arnold (Politecnico di Milano); Matteo Giuliani (Politecnico di Milano); Rohan Singh Wilkho (Cornell University) Moderator: Patrick M. Reed (MSD CoP Facilitation Team); Stefano Galelli (MSD CoP AI Working Group Co-Chair); David Gold (MSD CoP AI Working Group Co-Chair) This webinar was held on: June 23rd, 2026 from 12-1 PM EST.

Arnold, Wyatt [Politecnico di Milano]↗
DOE OSTI2026

Structure-aware Initialization via Numerical Continuation and Informed Priors

Scientific machine learning (SciML) often operates in ill-conditioned, weakly identifiable regimes due to limited data or indirect observations. In such settings, optimization and inference are highly sensitive to the starting point, making initialization--often under-reported--a consequential degree of freedom. Random initialization is not a neutral default as it induces an implicit prior over candidate solutions and can systematically bias the result, producing large run-to-run variability. Here, we formalize this view by treating initialization as a hidden confounder in SciML and develop a unifying theory for structure-aware initialization via numerical continuation, constructing warm starts from related problem instances. Across representative tasks, including physics-informed neural networks, maximum likelihood estimation, and variational inference, warm starts have been shown to consistently reduce optimization effort and improve reliability.

Data integrity↗
DOE OSTI2024

Non-Intrusive Parallel-in-Time Solvers for Partial Differential Equations (Final Report)

Many time-dependent problems and simulations are often modeled using Partial Differential Equations. Traditional modeling approaches that use sequential time-stepping are reaching a bottleneck in optimizing efficiency. The Center of Applied Science and Computing at Lawrence Livermore National Laboratory extensively works on parallelizing these algorithms to leverage the increasing computational power from the growing number of processors in computer hardware. In particular, they aim to design non-intrusive algorithms that can generalize to a variety of problems and sizes without requiring additional information from or modifications on the original problems. Multigrid Reduction in Time (MGRIT) is a parallel-in-time algorithm that is designed to be non-intrusive. This project focuses on increasing the efficiency of MGRIT by approximating the coarse-grid operator using machine learning approaches as a means to find the most non-intrusive, or general, solution.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Novel Carbon Storage Monitoring Methods

Conference presentation at American Institute of Chemical Engineers (AIChE) Annual Meeting, San Diego, California, October 27–31, 2024. We presented a high-level overview of many novel and sustainable carbon storage monitoring methods that are at various stages of planning or demonstration to accelerate the deployment of CCS technologies at future CCS sites across the United States. Our work impacts the general CCS industry by providing novel and more sustainable methods for tracking CO 2 plume migration and performing assurance monitoring. Our work primarily benefits three parties: 1) CCS community members through knowledge sharing of lessons learned; 2) CCS operators through commercialization of additional methods, including improvements on workflows and simplification of fieldwork; and 3) stakeholders of CCS projects through implementation of low-impact and more autonomous monitoring solutions.

02 PETROLEUM↗
DOE OSTI2025

Recent Collaborations and Innovations to Demonstrate Next-Generation Techniques for Monitoring Subsurface Carbon Storage

World Carbon Capture, Utilization, and Storage (CCUS) Conference, Bergen, Norway, September 1–4, 2025. This talk provides a high-level overview of many novel and sustainable carbon storage-monitoring methods to accelerate the deployment of CCUS technologies at future CCUS sites across the United States. The Energy & Environmental Research Center’s work impacts the general CCUS industry by providing novel low-impact methods for tracking the injected plume’s migration and more autonomous data collection and processing techniques for performing assurance monitoring. Specifically, the results benefit 1) CCUS community members through knowledge sharing of lessons learned; 2) CCUS operators through commercialization of additional methods, including improvements to workflows and simplification of fieldwork; and 3) CCUS project stakeholders through implementation of low-impact and more autonomous monitoring solutions.

02 PETROLEUM↗
DOE OSTI2026

Mark-18A Target Material Recovery Program: Initial Hot Startup

The Savannah River National Laboratory (SRNL) has been tasked by the National Nuclear Security Administration to recover highly valued isotopes from irradiated Mark-18A (Mk-18A) targets. The Savannah River Site (SRS) has sixty-five Mk-18A targets available for the recovery of the high valued materials. The Mk-18A targets were manufactured with Pu-242 then irradiated under high neutron flux in K-Reactor at the SRS from 1968 to 1978. The sixty-five Mk-18A targets are currently stored in the SRS L-Area Basin and will be removed one at a time and individually transported to SRNL. Upon receipt at SRNL, the Mk-18A target material will be removed from the confinement, dissolved, chemically separated, and calcined to a stable oxide. The flowsheet is designed to recover the plutonium as well as the trivalent actinides. The original targets were manufactured with varying quantities of Pu-242, ranging from 5.2 grams to 121.5 grams. Taking a graded approach to process start up, the lowest loaded target (FT-80-03) was selected as the first target to be received and processed at SRNL. As operational experience and knowledge is gained from processing targets, higher loaded targets will be selected for processing to increase the quantities of valuable isotopes recovered. Due to dose concerns, the receipt and processing of the Mk-18A targets is performed in the SRNL Shielded Cells Facility. The targets are stored in a double J-can confinement in the L-Area Basin. A specially designed cask was procured for transport of the targets from L-Area to SRNL. Once received at SRNL, the targets are loaded into the back of Cell 7 and resized as they enter the cell. The resized targets (1/4 length) are then processed one at a time through the following processes: caustic dissolution and filtration, acidic dissolution and filtration, elutable Reillex anion exchange, diglycolamide (DGA) resin extraction, and DGA calcination. This processing results in two product streams. The first is an aqueous plutonium solution which is removed from the shielded cells and taken to a glovebox for further purification and conversion to an oxide. The second is a calcined oxide product containing the Am and Cm as well as other lanthanide fission products which is removed from the shielded cells using a bagless transfer system. Both materials are packaged for shipment to Oak Ridge National Laboratory (ORNL). This paper will discuss the operating experience, lessons learned, and results from initial process hot startup.

Armstrong, Christopher [Savannah River National La↗
DOE OSTI2025

Block-Structured Operator Inference for Coupled Multiphysics Model Reduction

This work presents a block-structured formulation of Operator Inference as a way to learn structured reduced-order models for multiphysics systems. The approach specifies the governing equation structure for each physics component and the structure of the coupling terms. Once the multiphysics structure is specified, the reduced-order model is learned from snapshot data following the nonintrusive Operator Inference methodology. In addition to preserving physical system structure, which in turn permits preservation of system properties such as stability and second-order structure, the block-structured approach has the advantages of reducing the overall dimensionality of the learning problem and admitting tailored regularization for each physics component. The numerical advantages of the block-structured formulation over a monolithic Operator Inference formulation are demonstrated for aeroelastic analysis, which couples aerodynamic and structural models. For the benchmark test case of the AGARD 445.6 wing, block-structured Operator Inference provides an average 20% online prediction speedup over monolithic Operator Inference across subsonic and supersonic flow conditions in both the stable and fluttering parameter regimes while preserving the accuracy achieved with monolithic Operator Inference.

42 ENGINEERING↗
DOE OSTI2025

Machine learning for reparameterization of multi-scale closures

Scientific machine learning (ML) is becoming increasingly useful in learning closure models for multi-scale physics problems; however, many ML approaches require a vast array of training data and can struggle with generalization and interpretability. Here, rather than learning an entire closure operator, we adopt an existing reduced-dimension model of the microphysics and learn an optimal re-parameterization of the solver. We demonstrate two approaches for training the reduced dimension closure model (1) an a priori method that optimizes the closure parameterization and the neural network parameters separately and (2) an a posteriori method that simultaneously optimizes both. Using the simulation of biomass pyrolysis as a motivating example, we show that the a posteriori method achieves better target losses and is less dependent on training dataset size for generalizability. We then demonstrate the impact that implementing this reparameterization has at the macroscale, showing improved predictive performance with no modification to the underlying macroscale solvers.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Continual Learning for Production-Level Machine Learning in Particle Accelerators

Particle accelerators operate in complex environments where data distribution can change dynamically, leading to data drifts that significantly challenge Machine Learning (ML) models. These non-stationary conditions often cause ML models to deteriorate in performance, making it difficult to maintain reliable predictions in operation. The primary sources of data drifts are changes in accelerator settings and changes in equipment performance which cannot be measured directly. To bridge this gap between ML development and long-term deployment in operational settings, we identify key areas within particle accelerators where continual learning can help mitigate drift-induced performance degradation. We will provide a practical guide on selecting the appropriate method given resource constraints and desired stability plasticity trade offs. As a concrete example, we will present a real-world use case for anomaly detection to predict errant beams at the Spallation Neutron Source accelerator, where continual learning has been employed to demonstrate stable performance on drifting data streams. We will present practical challenges, lessons learned, and the results from the deployed ML model.

Rajput, Kishansingh [Thomas Jefferson National Acc↗