Search NASA⌕ Search

SEARCH · Search NASA

Results for “Deep”

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 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Deep Learning-based Surrogate Model for Efficient Reservoir Simulation in Large-scale Geological Carbon Storage: Application in IBDP Dataset

This project introduces an advanced deep learning (DL)-based surrogate modeling approach to enhance the efficiency and accuracy of large-scale geological carbon storage (GCS) simulations. Using the Illinois Basin Decatur Project (IBDP) dataset as training data, the study employs a residual U-Net architecture to predict critical state variables such as pressure and CO₂ saturation, as well as CO₂ plume migration. By incorporating key geological parameters (e.g., porosity, permeability, and rock facies) and physics-informed inputs like the diffusive time of flight and time step, the DL model effectively reduces computational complexity while maintaining robust physical constraints. Compared to traditional simulators like Eclipse, the DL model achieves remarkable accuracy, with a root mean square error (RMSE) of 1.57 psi for pressure and 0.007 for saturation, and dramatically reduces computational time from hours to just 69.9 seconds for 50-step simulations. These results demonstrate the potential of innovative DL methodologies to improve the predictivity and operational efficiency of GCS simulations, providing a reliable foundation for decision-making in CCS operations. Supported by the SMART initiative, this project underscores the success of leveraging computational innovations to advance CCS technologies.

advanced deep learning↗
DOE OSTI2026

Measuring the Energy Consumption and Efficiency of Deep Neural Networks: An Empirical Analysis and Design Recommendations

Addressing the "Red-AI" trend of rising energy consumption by large-scale neural networks, this study investigates the measured energy consumption of training various fully connected neural network architectures. We introduce the BUTTER-E dataset, an augmentation to the BUTTER Empirical Deep Learning dataset, containing energy consumption and performance data from 41,129 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (trainable parameters), 8 "shapes", and 14 depths on both CPUs and GPUs using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use. Our analysis uncovers a surprising, hardware-mediated non-linear relationship between energy efficiency and network design, challenging the assumption that reducing the number of parameters or FLOPs is the best way to achieve greater energy efficiency. We propose a straightforward and effective energy model that accounts for network size, computing, and memory hierarchy. Highlighting the need for cache-considerate algorithm development, we suggest a codesign approach to energy efficient network, algorithm, and hardware design. This work contributes to the fields of sustainable computing and Green AI, offering practical guidance for creating more energy-efficient neural networks and promoting sustainable AI.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Diol-enhanced natural deep eutectic solvents for efficient poplar pretreatment

Natural deep eutectic solvents (NDESs) are promising biomass pretreatment media, but their industrial application is often hindered by high viscosity. To address this limitation, diol-enhanced ternary DESs (TDESs) were prepared by incorporating 1,4-butanediol (1,4-BDO) or ethylene glycol (EG) into a choline chloride (ChCl) and 3,4-dihydroxybenzoic acid (DHBA) system. The applied TDESs maintained a liquid state at room temperature and had significantly reduced viscosity compared to the binary DES (BDES). In addition, the applied diols increased lignin solubility and suppressed lignin condensation by intercepting reactive carbocation intermediates. As a result, the recovered lignins from diol-induced TDES pretreatments showed better preservation of β-O-4 linkages and reduced condensation, improving their potential for downstream valorization. The diol-assisted DES systems showed a synergistic effect from the reduced viscosity, enhanced lignin solubility, and suppression of unwanted condensation, resulting in more effective biomass pretreatment performance, including the enhanced delignification and higher enzymatic digestibility compared to BDES. The 1,4-BDO-enhanced DES was also successfully applied to DHBA-enriched transgenic poplar, highlighting its potential for the processing of engineered biomass feedstocks.

3,4-dihydroxybenzoic acid↗
DOE OSTI2025

Insights from Femtosecond Transient Absorption Spectroscopy into the Structure–Function Relationship of Glyceline Deep Eutectic Solvents

This study aimed to determine the structure–function relationship (SFR) for ChCl–glycerol mixtures, a deep eutectic solvent (DES), by investigating their microscopic solvation dynamics and how it relates to their macroscopic properties across varying concentrations of ChCl. Femtosecond transient absorption (fs-TA) spectroscopy revealed two distinct solvation dynamics time constants: τ 1 , governed by glycerol–glycerol interactions, and τ 2 , dominated by the choline response. The τ 2 minimum at 25–30 mol % ChCl closely aligned with the eutectic composition (~33.33 mol % ChCl), where the glycerol network was the most organized and the choline ions exhibited the fastest relaxation. The viscosity decreased sharply up to ~25 mol % ChCl and then plateaued, while the conductivity increased monotonically with ChCl concentration, reflecting enhanced ionic mobility. The density decreased with both increasing ChCl concentration and temperature, indicating disrupted hydrogen bonding and reduced molecular packing. The polarity, measured using betaine-30 (B30) and the E T (30) polarity scale, increased steeply up to approximately 25 mol % ChCl before reaching a plateau. These findings identified the eutectic composition as the optimal concentration range for balancing stability, fluidity, conductivity, and enhanced dynamics within the glycerol system.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Advancing the Limits of InSAR to Detect Crustal Displacement from Low-Magnitude Earthquakes through Deep Learning

Detecting surface deformation associated with low-magnitude (M w ≤ 5) seismicity using interferometric synthetic aperture radar (InSAR) is challenging due to the subtlety of the signal and the often challenging imaging environments. However, low-magnitude earthquakes are potential precursors to larger seismic events, and thus characterizing the crustal displacement associated with them is crucial for regional seismic hazard assessment. We combine InSAR time-series techniques with a Deep Learning (DL) autoencoder denoiser to detect the magnitude and extent of crustal deformation from the M w = 3.4 Gallina, New Mexico earthquake that occurred on 30 July 2020. Although InSAR alone cannot detect event-related deformation from such a low-magnitude seismic event, application of the DL method reveals maximum displacements as small as (±2.5 mm) in the vicinity of both the fault and earthquake epicenter without prior knowledge of the fault system. This finding improves small-scale displacement discernment with InSAR by an order of magnitude relative to previous studies. We additionally estimate best-fitting fault parameters associated with the observed deformation. The application of the DL technique unlocks the potential for low-magnitude earthquake studies, providing new insights into local fault geometries and potential risks from higher-magnitude earthquakes. This technique also permits low-magnitude event monitoring in areas where seismic networks are sparse, allowing for the possibility of global fault deformation monitoring.

58 GEOSCIENCES↗
DOE OSTI2025

Understanding EV Charging Pain Points Through Deep Learning Analysis

Current and potential electric vehicle (EV) owners express concerns about the charging infrastructure, mentioning non-functional chargers, prolonged charging times, inconvenient charger locations, long wait times, and high costs as major barriers. Addressing these issues often requires analyzing actual vehicle charging data, which is typically proprietary and inconsistent due to diverse standards and protocols. To understand and improve the EV charging experience, customer reviews are typically used to identify common customer pain points (CPPs). However, there is not a comprehensive method to map customer reviews to a standardized set of CPPs. In collaboration with the National Charging Experience (ChargeX) Consortium, this study bridges these gaps by proposing a Systematic Categorization and Analysis of Large-scale EV-charging Reviews (SCALER) framework. SCALER is an integrated, deep learning framework that segments, actively labels, analyzes, and classifies EV charging customer reviews into six CPP categories. To test its effectiveness, we used SCALER to analyze over 72,000 reviews from customers charging various EV models on different networks across the United States. SCALER achieves a classification accuracy of 92.5%, with an F1 score exceeding 85.7%. By demonstrating real-world applications of SCALER, we enhance the industry’s ability to understand and address CPPs to improve the EV charging experience.

29 - ENERGY PLANNING, POLICY AND ECONOMY↗
DOE OSTI2026

Deep nonparametric estimation of operators between infinite dimensional spaces

Learning operators between infinitely dimensional spaces is an important learning task arising in machine learning, imaging science, mathematical modeling and simulations, etc. This paper studies the nonparametric estimation of Lipschitz operators using deep neural networks. Non-asymptotic upper bounds are derived for the generalization error of the empirical risk minimizer over a properly chosen network class. Under the assumption that the target operator exhibits a low dimensional structure, our error bounds decay as the training sample size increases, with an attractive fast rate depending on the intrinsic dimension in our estimation. Our assumptions cover most scenarios in real applications and our results give rise to fast rates by exploiting low dimensional structures of data in operator estimation. We also investigate the influence of network structures (e.g., network width, depth, and sparsity) on the generalization error of the neural network estimator and propose a general suggestion on the choice of network structures to maximize the learning efficiency quantitatively.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Deep-Learning-Based Multi-Timescale Load Forecasting in Buildings: Opportunities and Challenges from Research to Deployment

Electricity load forecasting for buildings and campuses is becoming increasingly important as the penetration of distributed energy resources (DERs) grows. Efficient operation and dispatch of DERs require reasonably accurate predictions of future energy consumption in order to conduct near-real-time optimized dispatch of on-site generation and storage assets. Electric utilities have traditionally performed load forecasting for load pockets spanning large geographic areas, and therefore, forecasting has not been a common practice by buildings and campus operators. Given the growing trends of research and prototyping in the grid-interactive efficient buildings domain, characteristics beyond simple algorithm forecast accuracy are important in determining the algorithm's true utility for smart buildings. Other characteristics include the overall design of the deployed architecture and the operational efficiency of the forecasting system. In this work, we present a deep-learning-based load forecasting system that predicts the building load at 1-hour intervals for 18 hours in the future. We also discuss challenges associated with the real-time deployment of such systems as well as the research opportunities presented by a fully functional forecasting system that has been developed within the National Renewable Energy Laboratory's Intelligent Campus program.

building load forecasting↗
DOE OSTI2026

LAF-Net: A Deep Residual and Cross-Attention Framework for Day-Ahead Load Forecasting: Preprint

Accurate day-ahead load forecasting is essential for reliable power system operations and market efficiency. System operators such as the Midcontinent Independent System Operator (MISO) rely on forecasts from multiple vendors, yet combining them effectively remains a persistent challenge due to vendor-specific biases. This paper presents a novel LSTM-Attention Fusion Network with Error Representation (LAF-Net) that enhances day-ahead hourly load forecasting through deep residual learning and multi-modal cross-attention. The proposed model builds a historical error memory from past vendor performance and dynamically queries it with future hour context to generate adaptive, hour-specific trust weights for each vendor. A bounded residual correction further refines forecasts by mitigating systematic and temporally localized errors. Tested on real MISO LBA data with multi-vendor forecasts, LAF-Net consistently outperforms the best vendor baseline across all 38 LBAs, achieving more than a 40% reduction in system-level mean absolute error (MAE) during peak load hours relative to the best vendor baseline.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Socio-Techno-Economic Feasibility of Deep Geothermal with EGS for Residential District Heating in the Northeastern United States

In this paper, we present socio-techno-economic simulation results for enhanced geothermal system (EGS) district heating systems in the northeastern United States. We applied the geospatial Distributed Geothermal Market Demand (dGeo) tool coupled with GEOPHIRES to evaluate the feasibility of geothermal deep direct-use, utilizing EGS reservoirs for providing residential heating with district heating systems. The results were combined with JEDI results to assess social impacts. Simulation results assuming doublet EGS reservoirs with target production temperature of 80 degrees C and newly constructed district heating systems indicate a wide range in system levelized cost of heat (LCOH) values ranging from ~$15/MMBtu to over $1,000/MMBtu, with most the attractive regions urban areas with high total thermal demand and high thermal demand density. We estimate about 60 GWth in potential installed capacity with LCOH values under $50/MMBtu and 100 GWth in potential capacity with LCOH values under $100/MMBtu, respectively.

15 GEOTHERMAL ENERGY↗
DOE OSTI2025

Physics-informed Deep Reinforcement Learning-based Control in Power systems

Incorporating physics information into the deep reinforcement learning (DRL) process is a promising approach for addressing the challenges faced in learning-based control design problems for physical systems. Power grid dynamics, being a physical system, adheres to specific physical laws, constraints, as well as operational and control rules. Therefore, consideration of such physics-based law improves the learning process drastically. In general, traditional grid control schemes rely on rule-based mechanisms that cannot adapt to changing operating conditions. To improve the adaptability and computation time, recent research has seen a surge of DRL-based applications in power grid control. A generic DRL-based control design imposes the system performance requirements through the design of reward functions. In some cases, some of the important physics information is injected through this reward function. However, due to the complex dynamics and large state-action space, learning an optimal DRL policy often becomes challenging. Inspired by the latest developments in general machine learning (ML) research, power system researchers have been investigating more direct ways of incorporating physics knowledge into DRL training. This chapter specifically focuses on these aspects of physics-informed DRL designs in grid control. It discusses the significance, applications, research gaps, and open problems that need to be addressed in future research.

artificial intelligence, machine learning↗
DOE OSTI2025

A Reductively Stable Electrolyte Realizes Deep Cycling Behavior in Anode‐free Sodium Batteries

For anode‐free sodium batteries to achieve practical consideration, highly reversible chemistries require exceptional ≥99.95% coulombic efficiencies that maintain over prolonged cycling periods. To do so, consumption of this severely limited sodium inventory must be restricted while metal nucleation processes that comprise the in situ formed metal anode are improved. Herein, we describe a fluorine‐free carborane electrolyte that satisfies these criteria by emphasizing reductive stability and weakly coordinating anion behavior as design principles. We find this approach promotes the development of a thin, robust SEI chemistry rich in both organic speciation and boron. The electrolyte described herein exhibits ideal metal nucleation behavior on one‐micron thin carbonaceous current collector surfaces and achieves a metal deposition/stripping efficiency near parity for 400 cycles. This novel anode chemistry is introduced to anode‐free full cell configurations where 87% of the initial discharge capacity is retained after 1000 cycles at 2.0 C. In conclusion, post‐test characterization of deep‐cycled anode‐free cells reveals suppressed capacity fade in these systems is attributed to the chemical stability of the carborane anion.

anode-free↗
DOE OSTI2024

Deep Learning for Spectroscopic X-ray Nano-Imaging Denoising

Synchrotron transmission X-ray microscopy with absorption near edge structure (TXM-XANES) is a powerful tool for investigating the structure and composition of materials at nano- to meso-scales. It is, however, often challenged by high levels of noise that obscure critical details at the single-pixel level. To address this issue, a deep learning-based algorithm is developed for suppressing the image noise, grounded in self-supervised learning principles. In contrast to traditional image denoising methods, this approach successfully enhances the visibility of fine details while significantly reducing the noise in the X-ray images. Through this advancement, the potential of the approach for improving the accuracy and interpretability of the TXM-XANES data is demonstrated, thereby enabling more precise detection of nanoscale phenomena such as inhomogeneous cation redox and metal segregation in battery cathode materials. This technique offers an effective new avenue for harnessing the full potential of synchrotron TXM-XANES imaging, paving the way for a range of exciting new studies in materials science and beyond.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Thermodynamic Framework for Water in Deep Eutectic Solvents and Ionic Liquids

Water is widely recognized as critical to the tunability and electrochemical stability of nonaqueous solvents such as deep eutectic solvents (DESs) and ionic liquids (ILs). Traditionally, the water content of these solvents has been controlled by either drying or adding small amounts of water to control their bulk properties to meet specific application requirements. The total water content by itself, does not provide sufficient information about the chemical reactivity and molecular interactions within DES– and IL–water mixtures. In this concept article, water activity is highlighted as a thermodynamically more rigorous descriptor to quantify the influence of the co‐solvent water on DES– and IL–water mixtures. Water activity relates measurable physical properties, such as vapor pressure, density, viscosity, electrochemical stability, and conductivity of DESs and ILs to the underlying molecular interactions between their components. Furthermore, water activity of DESs and ILs correlates with changes in local solvent structures and thermodynamic excess properties, including excess molar volume, enthalpy, and Gibbs energy.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Tuning the Properties of Lignin‐Derived Deep Eutectic Solvents for Biomass Processing

Lignin-derived deep eutectic solvents (DESs) have been investigated as sustainable green media for biomass processing. However, the properties and processability of DESs have not been fully understood with the chemical structures of their constituents for biomass fractionation. Here, in this article, the properties of the phenolic DESs are discussed with different numbers of functional groups, such as –OCH 3 and –CHO in their hydrogen bond donor (HBD) structures. The formation of DES is significantly related to the hydrogen bond between its constituents, identified by nuclear magnetic resonance (NMR) analysis and density functional theory calculation (DFT). Lower viscosity and net basicity of DES are achieved with fewer –OCH 3 on HBD structures, resulting in enhanced processability and fractionation efficiency. The thermal stability of the DES is also influenced by the –OCH 3 and –CHO of HBD, as indicated by its onset temperature. The recyclability of the phenolic DES is confirmed by the fractionation performance of the recycled DES. Understanding the structural impacts of DES constituents on the properties and performance is crucial for designing solvents in biorefinery applications.

Ryu, Jiae [State University of New York (SUNY), Sy↗
DOE OSTI2024

Development of AlGaN‐Based Deep‐Ultraviolet Avalanche Photodetectors—Toward Their Intrinsic Characteristics

Al 0.6 Ga 0.4 N deep‐UV p–i–n avalanche photodiodes (APDs) are demonstrated grown by metalorganic chemical vapor deposition (MOCVD) on a (0001) c‐plane aluminum nitride (AlN) bulk substrate and fabricated both with and without an ion implantation process for mesa‐edge electric‐field termination. Three design considerations are taken into account: the optical absorption of the AlN bulk substrate, the photosensitivity of the APD, and the breakdown electric field. The MOCVD growth conditions of the AlGaN APD layers are optimized. Then, a detailed description of the seven device fabrication steps of the APDs is provided, which includes nitrogen ion implantation. The APDs fabricated by the ion implant process exhibit a dark‐current density under low reverse bias ≈1 × 10 −9 A cm −2 . This is one order of magnitude lower than the APD fabricated without ion implantation from the same wafer. The breakdown voltage of the APD is ≈−140 V. The calculated optical gain of the ion‐implanted APD beyond avalanche breakdown is ≈5.2 × 10 5 (current limited) and the device has an average zero‐bias photoresponse of ≈68 mA W −1 at a wavelength of 250 nm.

Jeong, Hoon [Department of Electronic Engineering ↗
DOE OSTI2024

A Deep Learning Approach for Detection and Localization of Leaf Anomalies

The detection and localization of possible diseases in crops are usually automated by resorting to supervised deep learning approaches. In this work, we tackle these goals with unsupervised models, by applying three different types of autoencoders to a specific open-source dataset of healthy and unhealthy pepper and cherry leaf images. CAE, CVAE and VQ-VAE autoencoders are deployed to screen unlabeled images of such a dataset, and compared in terms of image reconstruction, anomaly removal, detection and localization. The vector-quantized variational architecture turns out to be the best performing one with respect to all these targets.

Calabro', Davide↗
DOE OSTI2025

Benchmarking Operators in Deep Neural Networks for Improving Performance Portability of SYCL

SYCL is a portable programming model for heterogeneous computing, so it is important to obtain reasonable performance portability of SYCL. Towards the goal of better understanding and improving performance portability of SYCL for machine learning workloads, we have been developing benchmarks for basic operators in deep neural networks (DNNs). These operators could be offloaded to heterogeneous computing devices such as graphics processing units (GPUs) to speed up computation. In this paper, we introduce the benchmarks, evaluate the performance of the operators on GPU-based systems, and describe the causes of the performance gap between the SYCL and Compute Unified Device Architecture (CUDA) kernels. We find that the causes are related to the utilization of the texture cache for read-only data, optimization of the memory accesses with strength reduction, use of local memory, and register usage per thread. We hope that the efforts of developing benchmarks for studying performance portability will stimulate discussion and interactions within the community.

Jin, Zheming [ORNL] (ORCID:000000027197780X)↗