Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep potential”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Deep-learning-driven simulations of boundary layer clouds over the Southern Great Plains

Abstract. Based on long-term observations at the Southern Great Plains site by the Atmospheric Radiation Measurement (ARM) program for training and validation, a deep-learning model is developed to simulate the daytime evolution of boundary layer clouds (BLCs) from the perspective of land–atmosphere coupling. The model takes ARM measurements (including early-morning soundings and diurnally varying surface meteorological conditions and heat fluxes) as inputs and predicts hourly estimates (including cloud occurrence, the positions of cloud boundaries, and the vertical profile of the cloud fraction) as outputs. The deep-learning model offers good agreement with the observed cloud fields, especially in the accuracy with which cloud occurrence and base height are reproduced. When the inputs are substituted by reanalysis data from ERA5 and MERRA-2, the outputs of the deep-learning model provide a better agreement with observation than the cloud fields extracted from ERA5 and MERRA-2 themselves. Thus, the deep-learning model shows great potential to serve as a diagnostic tool for the performance of physics-based models in simulating stratiform and cumulus clouds. By quantifying biases in clouds and attributing them to the simulated atmospheric state variables versus the model-parameterized cloud processes, this observation-based deep-learning model may offer insights into the directions needed to improve the simulation of BLCs in physics-based models for weather forecasting and climate prediction.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Neurogenesis Deep Learning: Extending deep networks to accommodate new classes

Neural machine learning methods, such as deep neural networks (DNN), have achieved remarkable success in a number of complex data processing tasks. These methods have arguably had their strongest impact on tasks such as image and audio processing – data processing domains in which humans have long held clear advantages over conventional algorithms. In contrast to biological neural systems, which are capable of learning continuously, deep artificial networks have a limited ability for incorporating new information in an already trained network. As a result, methods for continuous learning are potentially highly impactful in enabling the application of deep networks to dynamic data sets. Here, inspired by the process of adult neurogenesis in the hippocampus, we explore the potential for adding new neurons to deep layers of artificial neural networks in order to facilitate their acquisition of novel information while preserving previously trained data representations. Our results on the MNIST handwritten digit dataset and the NIST SD 19 dataset, which includes lower and upper case letters and digits, demonstrate that neurogenesis is well suited for addressing the stability-plasticity dilemma that has long challenged adaptive machine learning algorithms.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Probing the state of hydrogen in δ − AlOOH at mantle conditions with machine learning potential

Hydrous and nominally anhydrous minerals are a fundamental class of solids of enormous significance to geophysics. They are the water carriers in the deep geological water cycle and impact structural, elastic, plastic, and thermodynamic properties and phase relations in Earth's forming aggregates (rocks). They play a critical role in the geochemical and geophysical processes that shape the planet. Their complexity has prevented predictive calculations of their properties, but progress in materials simulations ushered by machine-learning potentials is transforming this state of affairs. Here, we adopt a hybrid approach that combines deep learning potentials (DPs) with the strongly constrained and appropriately normed meta-generalized gradient approximation functional to simulate a prototypical hydrous system. We illustrate the success of this approach to simulate δ − AlOOH ( δ ) , a phase capable of transporting water down to near the core-mantle boundary of the Earth ( ∼ 2900 km depth and ∼ 135 GPa ) in subducting slabs. A high-throughput sampling of phase space using molecular dynamics simulations with DPs sheds light on the hydrogen-bond behavior and proton diffusion at geophysical conditions. These simulations provide a pathway for a deeper understanding of these crucial components that shape Earth's internal state. Published by the American Physical Society 2024

Luo, Chenxing (ORCID:0000000341166851)↗
NASA NTRS2009

Mud Volcanoes in the Martian Lowlands: Potential Windows to Fluid-Rich Samples from Depth

The regional setting of the Chryse-Acidalia area augurs well for a fluid-rich subsurface, accumulation of diverse rock types reflecting the wide catchment area, astrobiological prospectivity, and mud volcanism. This latter provides a mechanism for transporting samples from relatively great depth to the surface. Since mud volcanoes are not associated with extreme heat or shock pressures, materials they transport to the surface are likely to be relatively unaltered; thus such materials could contain interpretable remnants of potential martian life (e.g., organic chemical biomarkers, mineral biosignatures, or structural remains) as well as unmetamorphosed rock samples. None of the previous landings on Mars was located in an area with features identified as potential mud volcanoes (Fig. 3), but some of these features may offer targets for future missions aimed at sampling deep fluid-rich strata with potential habitable zones.

Oehler, Dorothy Z.↗
NASA NTRS2016

The Auto-Gopher: A Wireline Rotary-Percussive Deep Sampler

Accessing regions on planetary bodies that potentially preserved biosignatures or are presently habitable is vital to meeting NASA solar system "Search for Life" exploration objectives. To address these objectives, a wireline deep rotary-percussive corer called Auto-Gopher was developed. The percussive action provides effective material fracturing and the rotation provides effective cuttings removal. To increase the drill's penetration rate, the percussive and rotary motions are operated simultaneously. Initially, the corer was designed as a percussive mechanism for sampling ice and was demonstrated in 2005 in Antarctica reaching about 2 m deep. The lessons learned suggested the need to use a combination of rotation and hammering to maximize the penetration rate. This lesson was implemented into the Auto-Gopher-I deep drill which was demonstrated to reach 3-meter deep in gypsum. The average drilling power that was used has been in the range of 100-150 Watt, while the penetration rate was approximately 2.4 m/hr. Recently, a task has started with the goal to develop Auto-Gopher-II that is equipped to execute all the necessary functions in a single drilling unit. These functions also include core breaking, retention and ejection in addition drilling. In this manuscript, the Auto-Gopher-II, its predecessors and their capability are described and discussed.

drilling↗
DOE OSTI2023

Many but not all deep neural network audio models capture brain responses and exhibit correspondence between model stages and brain regions

Models that predict brain responses to stimuli provide one measure of understanding of a sensory system and have many potential applications in science and engineering. Deep artificial neural networks have emerged as the leading such predictive models of the visual system but are less explored in audition. Prior work provided examples of audio-trained neural networks that produced good predictions of auditory cortical fMRI responses and exhibited correspondence between model stages and brain regions, but left it unclear whether these results generalize to other neural network models and, thus, how to further improve models in this domain. We evaluated model-brain correspondence for publicly available audio neural network models along with in-house models trained on 4 different tasks. Most tested models outpredicted standard spectromporal filter-bank models of auditory cortex and exhibited systematic model-brain correspondence: Middle stages best predicted primary auditory cortex, while deep stages best predicted non-primary cortex. However, some state-of-the-art models produced substantially worse brain predictions. Models trained to recognize speech in background noise produced better brain predictions than models trained to recognize speech in quiet, potentially because hearing in noise imposes constraints on biological auditory representations. The training task influenced the prediction quality for specific cortical tuning properties, with best overall predictions resulting from models trained on multiple tasks. The results generally support the promise of deep neural networks as models of audition, though they also indicate that current models do not explain auditory cortical responses in their entirety.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Peak2Patch: High-Fidelity Functional Group Identification through Attention-Based Fusion of Infrared and Mass Spectra

Identifying molecular structure based on spectroscopic readings is a key task in a variety of chemical and biological applications. Common spectroscopy techniques, such as Infrared (IR) Spectroscopy and Mass Spectrometry (MS), provide detailed information on the structure of molecular compounds but nonetheless require expert-level knowledge to decode. Machine learning has emerged as a potential solution for automating structure prediction from chemical spectra; however, current approaches generally focus on single sensor modalities, neglecting to leverage the complementary information contained within differing spectra. In this paper, we introduce Peak2Patch, a novel approach to fusion-enhanced prediction of functional groups from IR and mass spectra. First, we perform a detailed comparison of backbone networks for encoding both sparse mass spectra and dense IR spectra and demonstrate the superior performance of transformer neural networks over current state-of-the-art convolutional neural networks. Second, we evaluate three broad categories of fusion: early (raw feature), middle (deep feature), and late (decision) fusion, demonstrating the potential of a deep feature fusion-based approach. Lastly, we present Peak2Patch, our attention-based fusion scheme, which leverages cross-attention to mix features between encoded tokens of the two modalities. We validate our approach on a publicly available multimodal spectroscopic data set of 790k simulated molecules, demonstrating a large improvement in functional group prediction over both the previous state-of-the-art and our own strong single-modal baselines.

Jacobson, Philip [Sandia National Laboratories (SN↗
DOE OSTI2017

ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost

Deep learning is revolutionizing many areas of science and technology, especially image, text, and speech recognition. In this paper, we demonstrate how a deep neural network (NN) trained on quantum mechanical (QM) DFT calculations can learn an accurate and transferable potential for organic molecules. We introduce ANAKIN-ME (Accurate NeurAl networK engINe for Molecular Energies) or ANI for short. ANI is a new method designed with the intent of developing transferable neural network potentials that utilize a highly-modified version of the Behler and Parrinello symmetry functions to build single-atom atomic environment vectors (AEV) as a molecular representation. AEVs provide the ability to train neural networks to data that spans both configurational and conformational space, a feat not previously accomplished on this scale. We utilized ANI to build a potential called ANI-1, which was trained on a subset of the GDB databases with up to 8 heavy atoms in order to predict total energies for organic molecules containing four atom types: H, C, N, and O. To obtain an accelerated but physically relevant sampling of molecular potential surfaces, we also proposed a Normal Mode Sampling (NMS) method for generating molecular conformations. Through a series of case studies, we show that ANI-1 is chemically accurate compared to reference DFT calculations on much larger molecular systems (up to 54 atoms) than those included in the training data set.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Potential Activities and Long Lifetimes of Organic Carbon-Degrading Extracellular Enzymes in Deep Subsurface Sediments of the Baltic Sea

Heterotrophic microorganisms in marine sediments produce extracellular enzymes to hydrolyze organic macromolecules, so their products can be transported inside the cell and used for energy and growth. Therefore, extracellular enzymes may mediate the fate of organic carbon in sediments. The Baltic Sea Basin is a primarily depositional environment with high potential for organic matter preservation. The potential activities of multiple organic carbon-degrading enzymes were measured in samples obtained by the International Ocean Discovery Program Expedition 347 from the Little Belt Strait, Denmark, core M0059C. Potential maximum hydrolysis rates (V max ) were measured at depths down to 77.9mbsf for the following enzymes: alkaline phosphatase, β- d -xylosidase, β- d -cellobiohydrolase, N-acetyl-β- d -glucosaminidase, β-glucosidase, α-glucosidase, leucyl aminopeptidase, arginyl aminopeptidase, prolyl aminopeptidase, gingipain, and clostripain. Extracellular peptidase activities were detectable at depths shallower than 54.95mbsf, and alkaline phosphatase activity was detectable throughout the core, albeit against a relatively high activity in autoclaved sediments. β-glucosidase activities were detected above 30mbsf; however, activities of other glycosyl hydrolases (β-xylosidase, β-cellobiohydrolase, N-acetyl-β-glucosaminidase, and α-glucosidase) were generally indistinguishable from zero at all depths. These extracellular enzymes appear to be extremely stable: Among all enzymes, a median of 51.3% of enzyme activity was retained after autoclaving for an hour. We show that enzyme turnover times scale with the inverse of community metabolic rates, such that enzyme lifetimes in subsurface sediments, in which metabolic rates are very slow, are likely to be extraordinarily long. A back-of-the-envelope calculation suggests enzyme lifetimes are, at minimum, on the order of 230days, and may be substantially longer. These results lend empirical support to the hypothesis that a population of subsurface microbes persist by using extracellular enzymes to slowly metabolize old, highly degraded organic carbon.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Towards deep computer vision for in-line defect detection in polymer electrolyte membrane fuel cell materials

Polymer Electrolyte Membrane (PEM) fuel cells are a promising source of alternative energy. However, their production is limited by a lack of well-established methods for quality control of their constituent materials like the membrane-electrode assembly during roll-to-roll manufacturing. One potential solution is the implementation of deep learning methods to detect unwanted defects through their detection in scanned images. Here we explore the detection of defects like scratches, pinholes, and scuffs in a sample dataset of PEM optical images using two deep learning algorithms: Patch Distribution Modeling (PaDiM) for unsupervised anomaly detection and Faster-RCNN for supervised object detection. Both methods achieve scores on performance metrics (ROC-AUC and PRO-AUC for PaDiM and AP for Faster-RCNN) that are comparable to their scores on benchmark datasets. These methods also have the potential to detect a wider range of defects compared to IR thermography and previous optical inspection methods. Overall, deep learning shows promise at detecting relevant defects of interest and has the potential to achieve real-time defect detection.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2019

Learning Interactions in Complex Biological Systems

In this project, we developed novel methods for integrating simulation and data-driven methods to accelerate biomedical discovery. In particular, we exploit agent-based models of disease processes and deep reinforcement learning to identify potential novel multi-drug, patient-specific, and adaptive therapeutic strategies. We use the Innate Immune Response agent-based model (IIRABM) to model sepsis as a demonstration of our approach. We first calibrate the agent-based model using novel stochastic optimization algorithms develop in this project, then use the calibrated model to identify optimal cytokine mediation strategies using deep reinforcement learning (DRL). The learned policy achieves a dramatic reduction in mortality over a set of 500 simulated patients relative to standalone antibiotic therapy. Advantages of our approach are three-fold: 1) the use of simulation allows exploring therapeutic strategies beyond clinical practice and available data, 2) advances in DRL accommodate learning complex therapeutic strategies for complex biological systems, and 3) optimized treatments respond to a patient’s individual disease progression over time, therefore capturing both differences across patients and the inherent randomness of disease progression within a single patient. We hope that this work motivates both considering adaptive, personalized multi-cytokine mediation therapy for sepsis and exploiting simulation with DRL for precision medicine more broadly.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2017

Addressing Technical and Policy Challenges for Remediation of Metals and Radionuclides within Complex Deep Vadose Zone Environments - 17028

Deep vadose zone contamination is a significant issue in many regions of the world. Although remediation of deep vadose zone contamination is important, direct contact with the contamination does not threaten human health or the environment due to its depth in the vadose zone. Yet, movement of contamination from the deep vadose zone to the groundwater creates the potential for exposure and risk to receptors. Therefore, the challenge of deep vadose zone remediation is to maintain a minimal rate of contaminant transport such that flux from the contaminated vadose zone is limited to protect groundwater resources and down-gradient receptors. We herein review major processes for viable remediation of deep vadose zone metal and radionuclides that form the practical constraints on remedial actions and provide an overview of how these principles are being practiced to address key contaminants at the U.S. Department of Energy Hanford Site. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
NASA NTRS2016

NASA's Space Launch System: A Transformative Capability for Exploration

Currently making rapid progress toward first launch in 2018, NASA's exploration-class Space Launch System (SLS) represents a game-changing new spaceflight capability, enabling mission profiles that are currently impossible. Designed to launch human deep-space missions farther into space than ever before, the initial configuration of SLS will be able to deliver more than 70 metric tons of payload to low Earth orbit (LEO), and will send NASA's new Orion crew vehicle into lunar orbit. Plans call for the rocket to evolve on its second flight, via a new upper stage, to a more powerful configuration capable of lofting 105 t to LEO or comanifesting additional systems with Orion on launches to the lunar vicinity. Ultimately, SLS will evolve to a configuration capable of delivering more than 130 t to LEO. SLS is a foundational asset for NASA's Journey to Mars, and has been recognized by the International Space Exploration Coordination Group as a key element for cooperative missions beyond LEO. In order to enable human deep-space exploration, SLS provides unrivaled mass, volume, and departure energy for payloads, offering numerous benefits for a variety of other missions. For robotic science probes to the outer solar system, for example, SLS can cut transit times to less than half that of currently available vehicles, producing earlier data return, enhancing iterative exploration, and reducing mission cost and risk. In the field of astrophysics, SLS' high payload volume, in the form of payload fairings with a diameter of up to 10 meters, creates the opportunity for launch of large-aperture telescopes providing an unprecedented look at our universe, and offers the ability to conduct crewed servicing missions to observatories stationed at locations beyond low Earth orbit. At the other end of the spectrum, SLS opens access to deep space for low-cost missions in the form of smallsats. The first launch of SLS will deliver beyond LEO 13 6U smallsat payloads, representing multiple disciplines, including three spacecraft competitively chosen through NASA's Centennial Challenges competition. Private organizations have also identified benefits of SLS for unique public-private partnerships. This paper will give an overview of SLS' capabilities and its current status, and discuss the vehicle's potential for human exploration of deep space and other game-changing utilization opportunities.

Robinson, Kimberly F.↗
NASA NTRS2016

NASA's Space Launch System: A Transformative Capability for Exploration

Currently making rapid progress toward first launch in 2018, NASA's exploration-class Space Launch System (SLS) represents a game-changing new spaceflight capability, enabling mission profiles that are currently impossible. Designed to launch human deep-space missions farther into space than ever before, the initial configuration of SLS will be able to deliver more than 70 metric tons of payload to low Earth orbit (LEO), and will send NASA's new Orion crew vehicle into lunar orbit. Plans call for the rocket to evolve on its second flight, via a new upper stage, to a more powerful configuration capable of lofting 105 tons to LEO or co-manifesting additional systems with Orion on launches to the lunar vicinity. Ultimately, SLS will evolve to a configuration capable of delivering more than 130 tons to LEO. SLS is a foundational asset for NASA's Journey to Mars, and has been recognized by the International Space Exploration Coordination Group as a key element for cooperative missions beyond LEO. In order to enable human deep-space exploration, SLS provides unrivaled mass, volume, and departure energy for payloads, offering numerous benefits for a variety of other missions. For robotic science probes to the outer solar system, for example, SLS can cut transit times to less than half that of currently available vehicles, producing earlier data return, enhancing iterative exploration, and reducing mission cost and risk. In the field of astrophysics, SLS' high payload volume, in the form of payload fairings with a diameter of up to 10 meters, creates the opportunity for launch of large-aperture telescopes providing an unprecedented look at our universe, and offers the ability to conduct crewed servicing missions to observatories stationed at locations beyond low Earth orbit. At the other end of the spectrum, SLS opens access to deep space for low-cost missions in the form of smallsats. The first launch of SLS will deliver beyond LEO 13 6-unit smallsat payloads, representing multiple disciplines, including three spacecraft competitively chosen through NASA's Centennial Challenges competition. Private organizations have also identified benefits of SLS for unique public-private partnerships. This paper will give an overview of SLS' capabilities and its current status, and discuss the vehicle's potential for human exploration of deep space and other game-changing utilization opportunities.

Robinson, Kimberly F.↗
DOE OSTI2019

Performance Analysis of Deep Learning Workloads on Leading-edge Systems

This work examines the performance of leading-edge systems designed for machine learning computing, including the NVIDIA DGX-2, Amazon Web Services (AWS) P3, IBM Power System Accelerated Compute Server AC922, and a consumer-grade Exxact TensorEX TS4 GPU server. Representative deep learning workloads from the fields of computer vision and natural language processing are the focus of the analysis. Performance analysis is performed along with a number of important dimensions. Performance of the communication interconnects and large and high-throughput deep learning models are considered. Different potential use models for the systems as standalone and in the cloud also are examined. The effect of various optimization of the deep learning models and system configurations is included in the analysis.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Crystallization of the P 3 Sn 4 Phase upon Cooling P 2 Sn 5 Liquid by Molecular Dynamics Simulation Using a Machine Learning Interatomic Potential

We performed molecular dynamics simulations to study the crystallization of the P3Sn4 phase from P 2 Sn 5 liquid using a machine learning (ML) interatomic potential with desirable efficiency and accuracy. Our results capture the liquid properties of P 2 Sn 5 at 1300 K, which is well above the melting temperature. The phase separation and crystallization are observed when P 2 Sn 5 liquid is cooled down below 832 and 505 K, respectively. The simulation results are in good agreement with the experimentally observed phase transformation behaviors and provide useful insights into the complex nucleation and crystallization process at the details of atomistic scale. Our work also demonstrated that ML interatomic potentials based on neural network deep learning are robust and capable of accurately describing the energetics and kinetics of complex materials through molecular dynamics simulations.

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

Reproducibility of Radiokrypton in Deep Desert Aquifers: Insights from a Decade of Research

Great technical advances have been achieved since the first atom-trap trace analysis (ATTA) -based radiokrypton application in Egypt, where 1 Myr old groundwater was discovered. Beyond advances in ATTA measurement capabilities, including reduction in sample size, analysis duration, and analytical uncertainty, major progress has been achieved over the past two decades in the sample collection and preparation techniques. These advances paved the expansion of ATTA-based noble gas applications to many other aquifers worldwide, illuminating the nature and flow pattern of deep groundwater systems. While the potential of this new analytical technique for old groundwater dating is well recognized, another important aspect yet to be examined is the reproducibility of radiokrypton in aquifers over time, i.e., how representative is a discrete groundwater sample, collected at a specific time and location, for the natural groundwater system? The likelihood of a negative answer is increased by flow-field disturbance in aquifers following massive groundwater abstraction. Here, in this work, we present repeated 81 Kr sampling and measurements in twenty-one sites over Israel, mostly of deep (up to 1 km) wells tapping confined aquifers in the arid to hyperarid Negev desert. The results demonstrate that radiokrypton measurements are indeed reproducible, even in cases where samples were collected as long as nine years apart and from highly productive (∼1 Mm 3 /yr order) pumping wells. Furthermore, many of the repeated measurements in this study (17 out of the 21 sites) were conducted with different ATTA Instruments in two different laboratories using slightly different sampling, preparation, and analysis techniques, yet with an overall good agreement. The consistency in the ATTA-based 81 Kr-dating results over time highlights the robustness of this state-of-the-art technique as a tool to unravel groundwater flow patterns and encourages further applications to many other yet-to-be-explored deep aquifers.

atom-trap trace analysis↗
NASA NTRS2017

NASA's Space Launch System: A Transformative Capability for Deep Space Missions

Already making substantial progress toward its first launches, NASA’s Space Launch System (SLS) exploration-class launch vehicle presents game-changing new opportunities in spaceflight, enabling human exploration of deep space, as well as a variety of missions and mission profiles that are currently impossible. Today, the initial configuration of SLS, able to deliver more than 70 metric tons of payload to low Earth orbit (LEO), is well into final production and testing ahead of its planned first flight, which will send NASA’s new Orion crew vehicle around the moon and will deploy 13 CubeSats, representing multiple disciplines, into deep space. At the same time, production work is already underway toward the more-capable Block 1B configuration, planned to debut on the second flight of SLS, and capable of lofting 105 tons to LEO or of co-manifesting large exploration systems with Orion on launches to the lunar vicinity. Progress being made on the vehicle for that second flight includes initial welding of its core stage and testing of one of its engines, as well as development of new elements such as the powerful Exploration Upper Stage and the Universal Stage Adapter “payload bay.” Ultimately, SLS will evolve to a configuration capable of delivering more than 130 tons to LEO to support humans missions to Mars. In order to enable human deep-space exploration, SLS provides unrivaled mass, volume, and departure energy for payloads, offering numerous benefits for a variety of other missions. For robotic science probes to the outer solar system, for example, SLS can cut transit times to less than half that of currently available vehicles or substantially increased spacecraft mass. In the field of astrophysics, SLS’ high payload volume, in the form of payload fairings with a diameter of up to 10 meters, creates the opportunity for launch of large-aperture telescopes providing an unprecedented look at our universe. This presentation will give an overview of SLS’ capabilities and its current status, and discuss the vehicle’s potential for human exploration of deep space and other game-changing utilization opportunities.

Creech, Stephen D.↗