Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “LABELLING”

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 1,711 records · Page 95

Visual Detection with Context for Document Layout Analysis

We present 1) a work in progress method to visually segment key regions of scientific articles using an object detection technique augmented with contextual features, and 2) a novel dataset of region-labeled articles. A continuing challenge in scientific literature mining is the difficulty of consistently extracting high-quality text from formatted PDFs. To address this, we adapt the object-detection technique Faster R-CNN for document layout detection, and incorporate contextual information that leverages the inherently localized nature of article contents to improve the region detection performance. Due to the limited availability of region-labels for scientific articles, we also contribute a novel dataset of region annotations, the first version of which covers 9 region classes and 822 article pages. Initial experimental results demonstrate a 23.9% absolute improvement in mean average precision over the baseline by incorporating contextual features, and a processing speed 14x faster than a text-based technique. Ongoing work on further improvements is also discussed.

97 MATHEMATICS AND COMPUTING↗

Devices, systems, and methods for detecting nucleic acids using sedimentation

Embodiments of the present invention are directed toward devices, systems, and method for conducting nucleic acid purification and quantification using sedimentation. In one example, a method includes generating complexes which bind to a plurality of beads in a fluid sample, individual ones of the complexes comprising a nucleic acid molecule such as DNA or RNA and a labeling agent. The plurality of beads including the complexes may be transported through a density media, wherein the density media has a density lower than a density of the beads and higher than a density of the fluid sample, and wherein the transporting occurs, at least in part, by sedimentation. Signal may be detected from the labeling agents of the complexes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Evaluating Cu/BEA Catalyst Performance with Techno-Economic Analysis to Develop a Market-Responsive Biorefinery Concept Around the Conversion of Methanol to High-Octane Hydrocarbons

A recent advancement in the conversion of methanol to hydrocarbon fuels is the development of a Cu-modified beta zeolite catalyst (Cu/BEA) that exhibits a higher productivity and yield than the parent H-form BEA catalyst in the production of branched C4-C8 hydrocarbons. This Cu/BEA catalyst enables process improvements in the High-Octane Gasoline (HOG) pathway, which operates under lower severity conditions than traditional methanol-to-fuels processes, offering a lower coke formation rate, higher selectivity to C5+gasoline-range hydrocarbons, and subsequently, improved process economics based on a recent process model and techno-economic analysis (TEA). Here, a market-responsive biorefinery concept around methanol and the HOG pathway is presented, where the major C4 products, isobutane and isobutene, serve as versatile intermediates that can either be recycled to increase the yield of the HOG product or directed to jet-fuel range hydrocarbons. Based on simulated recycle experiments with 13C-labeled isobutane, the Cu/BEA catalyst achieved increased yield to the HOG products during DME homologation by activating the co-fed isobutane and incorporating it into the chain growth pathway, as evidenced by the presence of isotopically-labeled 13C in C5+ products. Olefin coupling of an isobutene-rich mixture was also investigated, and the fuel properties of the resulting synthetic kerosene product were within the specifications for a typical jet-fuel. These experimental results were incorporated into an updated Aspen-based process model and TEA to quantify their impact on a market-responsive biorefinery concept. The improvements offered by C4 recycling with the Cu/BEA catalyst result in a decrease in the methanol-to-fuel synthesis cost from 101 cents per gallon of gasoline equivalent (GGE) with the baseline BEA catalyst to 72 cents/GGE, due in part to a significantly increased HOG yield from 40 to 55 GGE/dry-ton biomass. Expanding the process model to direct C4 products to a jet-range hydrocarbon product demonstrated a yield of 42 GGE/dry-ton HOG with 12 GGE/dry-ton distillate, resulting in a minor increase in the fuel synthesis cost to 77 cents/GGE versus the HOG-only case. With continued development, these advancements in fuel yield and reductions in cost position a market-responsive biorefinery around the HOG pathway to be competitive against mature fuel-synthesis technologies from a variety of carbon sources through methanol.

09 BIOMASS FUELS↗

Rapid identification of microorganisms

Methods of labeling, identifying and differentiating microorganisms using functionalized Buckyballs are provided herein. The invention further provides methods for imaging or inhibiting gene expression using functionalized Buckyballs of the invention. The invention also provides a system for labeling, identifying and differentiating microorganisms.

59 BASIC BIOLOGICAL SCIENCES↗

Data-Driven Template Discovery Using Graph Convolutional Neural Networks

Modeling adversarial activities is a critical component of developing high-con?dence indicators of efforts to acquire, fabricate, proliferate, and/or deploy weapons of mass terror (WMTs). Current approaches to generating representative patterns of interest (a.k.a templates) from the real-world domains involve a Subject Matter Expert (SME)-guided manual process. The goal of Data-Driven Template Discovery (DDTD) is to use a (potentially small) set of SME generated templates to discover other previously unknown and interesting templates in an attributed graph. A template is an activity pattern describing a set of interactions among a group of nodes in the graph. The motivation behind DDTD is to expand the original set of templates, without having SMEs craft all the templates by hand. DDTD also provides seed templates to SMEs, to help them construct larger, high-?delity, and scenario-oriented templates. In these cases, obtaining a larger set of templates that are related (contain similar signals) to the original set is of great value. In this work, we propose to use Graph Convolutional Neural Networks (GCNs) to discover new templates that are heavily related to the original set. GCNs are a family of Neural Network (NN) architectures especially designed to work directly on graphs. In contrast to the traditional NNs, that require considerable amounts of labeled data, GCNs do not require a big labeled training set because they can directly leverage the graph structure instead. This property makes GCNs the perfect tool for creating activity templates.

Joaristi, Mikel↗

Nuclear Power Fault Diagnostics and Preventative Maintenance Optimization

The nuclear industry is beginning to see reactors shut down—even after their operating licenses have been extended—because they are not economically competitive with other energy sources. These early closures happen primarily due to economic reasons, despite excellent safety records. Therefore, it is imperative to reduce costs in order to prevent these early closures. One of the contributors to these economic reasons is the large operations and maintenance costs. This paper showcases recent research on advanced fault diagnostics techniques and preventative maintenance optimization (PMO) for reducing NPP maintenance costs. Specifically, it focuses on the feedwater and condensate system (FWCS) for both pressurized- and boiling-water reactor (BWR) systems. The computerized maintenance management system (CMMS), which contains the plant’s digital record of all corrective maintenance (CM) and preventative maintenance (PM) work orders, provided the ground truth for locating potential faults and labeling the process data as either healthy or faulted. Various feature extraction techniques were used to further differentiate the faulted data from the healthy data. Through a cross-validation procedure, support vectors machines were used to label other test sets of process data as either healthy or faulted. With relatively few faults identified in the BWR system, the potential for PMO opens up, since an unnecessary amount of PM leads to inflated maintenance costs. The steps for PMO are summarized, from component health determinations to recommendations for action. An example of PMO assessment is presented for condensate pumps, condensate booster pumps, and the respective motors that drive them.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Challenges in Automated Detection of COVID-19 Misinformation

The COVID-19 pandemic has made the dangers of the spread of misinformation obvious but despite much global effort to curbing its spread, fake information about the pandemic keeps proliferating. In this paper, we address the development of automated methods for verification of claims about COVID-19 and discuss the challenges associated with this task. We focus on labeled data collection, limitations of existing models, and difficulties of applying misinformation detection models in practical applications. Our initial analysis indicates label imbalance may be a particular challenge for developing claim verification models and we discuss options for alleviating this issue.

Herrmannova, Dasha↗

Nuclear Power Fault Diagnostics and Preventative Maintenance Optimization NPIC presentation

The nuclear industry is beginning to see reactors shut down—even after their operating licenses have been extended—because they are not economically competitive with other energy sources. These early closures happen primarily due to economic reasons, despite excellent safety records. Therefore, it is imperative to reduce costs in order to prevent these early closures. One of the contributors to these economic reasons is the large operations and maintenance costs. This paper showcases recent research on advanced fault diagnostics techniques and preventative maintenance optimization (PMO) for reducing NPP maintenance costs. Specifically, it focuses on the feedwater and condensate system (FWCS) for both pressurized- and boiling-water reactor (BWR) systems. The computerized maintenance management system (CMMS), which contains the plant’s digital record of all corrective maintenance (CM) and preventative maintenance (PM) work orders, provided the ground truth for locating potential faults and labeling the process data as either healthy or faulted. Various feature extraction techniques were used to further differentiate the faulted data from the healthy data. Through a cross-validation procedure, support vectors machines were used to label other test sets of process data as either healthy or faulted. With relatively few faults identified in the BWR system, the potential for PMO opens up, since an unnecessary amount of PM leads to inflated maintenance costs. The steps for PMO are summarized, from component health determinations to recommendations for action. An example of PMO assessment is presented for condensate pumps, condensate booster pumps, and the respective motors that drive them.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Elucidation of Aromatic Catabolic Pathways in White-Rot Fungi

This project aims to investigate the hypothesis that white-rot fungi can simultaneously depolymerize lignin extracellularly and catabolize depolymerization products intracellularly as carbon and energy sources. Evaluating this hypothesis will provide deeper understanding of the role of white-rot fungi in facilitating carbon sequestration in Nature. Additionally, identifying the most promising fungal strains for lignin turnover and catabolism will catalyze future efforts in genetic tool development to enable metabolic engineering in white-rot fungi for lignin bioconversion to bioproducts. Lignin is the second most abundant plant-based biopolymer on Earth and represents up to 40% of the energy density of lignocellulosic biomass. Even though lignin is a massive natural carbon and energy reservoir, only a small group of basidiomycete fungi, namely white-rot fungi (WRF), have evolved the ability to efficiently depolymerize and mineralize lignin to CO2 and H2O. Considerable research efforts have been undertaken to understand how WRF depolymerize lignin but the biochemical reactions that convert lignin into CO2 have been largely neglected. In fact, it is unclear if WRF intracellularly catabolize lignin-derived aromatic compounds to utilize them as a carbon and energy source, or rather if lignin is depolymerized and mineralized extracellularly merely to facilitate access to cellulose and hemicellulose for use as a primary carbon source. To date, we have employed 13C-isotope labeling, systems biology approaches, and in vitro enzyme assays to definitively demonstrate that two WRF, Trametes versicolor and Gelatoporia (Ceriporiopsis) subvermispora, funnel carbon from lignin-derived aromatic compounds into central carbon metabolism via intracellular catabolic pathways [1]. Specifically, 13C-isotopic labeling approaches showed that these WRF utilize poplar-derived aromatic compounds (e.g. 4-hydroxybenzoic acid (4-HBA)) as a carbon source. In silico genome analysis led us to hypothesize a complete catabolic pathway for 4-HBA and identify multiple homologous sequences for enzymes with putative oxidative decarboxylase, hydroxylase, and ring-opening dioxygenase activities, which are among the main biochemical reactions acting on aromatic compounds. Spatial and differential proteomic and metabolomic analyses supported the proposed catabolic pathways and showed alternative catabolic steps in T. versicolor that were not present in G. subvermispora. Based on the in silico, proteomics, and transcriptomics results, we down-selected enzymes for further in vitro characterization, and we have assigned a function to six fungal enzymes (including oxidative decarboxylases, hydroxylases, and ring-opening dioxygenases). Interestingly, even though we selected homologous enzyme pairs from both WRF with similar -omics trends, in a few cases only one of the studied fungi showed activity for the proposed substrate. Based on all the observations from this study [1], we hypothesized that 4-HBA preferentially undergoes oxidative decarboxylation to hydroquinone and subsequent hydroxylation to 1,2,4-benzenetriol in G. subvermispora before ring cleavage, whereas 4-HBA would preferentially undergo hydroxylation to protocatechuate and further oxidative decarboxylation to 1,2,4-benzenetriol in T. versicolor. Examining additional decarboxylases and hydroxylases as well as enzymes from other protein families that can perform the same or similar oxidative reactions, such as cytochromes P450 with aromatic hydroxylation activity, will also be key for elucidating enzyme preferences for specific substrates. Overall, this work forms the foundation of a new research area based on lignin catabolism by WRF, which could be further exploited to convert the undervalued biopolymer lignin into value-added compounds.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

The CanBikeCO Mini Pilot: Procedure and Preliminary Results

In fall 2020, the Colorado Energy Office, as part of the State of Colorado's "Can Do Colorado" initiative, initiated a project aimed at encouraging energy-efficient transportation during the COVID-19 pandemic. The initial mini-pilot provided e-bikes to 13 low-income households under an individual ownership model. This report assesses the impact of providing this additional mobility option on the travel behavior of participants. It also outlines the lessons learned from deploying a continuous monitoring platform to track the travel behavior. These lessons will influence the evaluation component for the full pilot, which will cover multiple geographic regions, start in summer 2021, and run for 2 years. The continuous data collection was enabled by a customized version of the open-source e-mission platform, called CanBikeCO, configured with a behavioral gamification feature. The Colorado Energy Office used this system to collect a unique data set consisting of 3 months of partially automated travel diaries, combining sensed and surveyed data and linked with demographic information, from 12 participants. The data collection process worked well overall: users generally liked the app, appreciated the game, and did not complain about battery life. The long tracking period introduced behavioral challenges in user engagement, which we plan to address using repeated patterns and automated status checks for the full pilot. The analysis results, based on the subset of trips with user-reported labels (68%), indicate that the e-bike was the dominant commute mode share (31%), in sharp contrast to the census bicycle commute mode share (<1%). E-bike trips primarily replaced single-occupancy vehicle (SOV) trips (28%), followed closely by walking (24%) and regular bike (20%). The non-motorized mode replacement corresponds to lower travel time and increased productivity enabled by the program. The emissions impact analysis of the program, computed using trip-level energy intensity factors, indicates savings of 1,367 lbs. of CO2. Although the results are strongly positive, the narrow demographic profile of study participants, their limited mobility alternatives, and nonuniform labeling indicate caution in broader interpretation. These preliminary results do suggest that such programs, supported by real-time education and support from program managers, can simultaneously meet equity and sustainability goals. The planned full pilot, addressing the data collection challenges and broadening the geographic scope, will provide additional insights into the generality of this approach.

ADVANCED PROPULSION SYSTEMS↗

Efficient Active Learning for Gaussian Process Classification by Error Reduction

Active learning sequentially selects the best instance for labeling by optimizing an acquisition function to enhance data/label efficiency. The selection can be either from a discrete instance set (pool-based scenario) or a continuous instance space (query synthesis scenario). In this work, we study both active learning scenarios for Gaussian Process Classification (GPC). The existing active learning strategies that maximize the Estimated Error Reduction (EER) aim at reducing the classification error after training with the new acquired instance in a onestep-look-ahead manner. The computation of EER-based acquisition functions is typically prohibitive as it requires retraining the GPC with every new query. Moreover, as the EER is not smooth, it can not be combined with gradient-based optimization techniques to efficiently explore the continuous instance space for query synthesis. To overcome these critical limitations, we develop computationally efficient algorithms for EER-based active learning with GPC. Further, we derive the joint predictive distribution of label pairs as a one-dimensional integral, as a result of which the computation of the acquisition function avoids retraining the GPC for each query, remarkably reducing the computational overhead. We also derive the gradient chain rule to efficiently calculate the gradient of the acquisition function, which leads to the first query synthesis active learning algorithm implementing EER-based strategies. Our experiments clearly demonstrate the computational efficiency of the proposed algorithms. We also benchmark our algorithms on both synthetic and real-world datasets, which show superior performance in terms of sampling efficiency compared to the existing state-of-the-art algorithms.

97 MATHEMATICS AND COMPUTING↗

Accelerating the Inference of the Exa.TrkX Pipeline

Recently, graph neural networks (GNNs) have been successfully used for a variety of particle reconstruction problems in high energy physics, including particle tracking. The Exa.TrkX pipeline based on GNNs demonstrated promising performance in reconstructing particle tracks in dense environments. It includes five discrete steps: data encoding, graph building, edge filtering, GNN, and track labeling. All steps were written in Python and run on both GPUs and CPUs. In this work, we accelerate the Python implementation of the pipeline through customized and commercial GPU-enabled software libraries, and develop a C++ implementation for inferencing the pipeline. The implementation features an improved, CUDA-enabled fixed-radius nearest neighbor search for graph building and a weakly connected component graph algorithm for track labeling. GNNs and other trained deep learning models are converted to ONNX and inferenced via the ONNX Runtime C++ API. The complete C++ implementation of the pipeline allows integration with existing tracking software. We report the memory usage and average event latency tracking performance of our implementation applied to the TrackML benchmark dataset.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Panel-Segmentation: A Python Package for Automated Solar Array Metadata Extraction Using Satellite Imagery

The NREL Python Panel-Segmentation package is a toolkit that automates the process of extracting accurate and valuable metadata related to solar array installations, using publicly available Google Maps satellite imagery. Previously published work includes automated azimuth estimation for individual solar installations in satellite images. Our continued research focuses on automated detection and classification of solar installation mounting configuration (tracking or fixed-tilt; rooftop, ground, or carport). Specifically, a Faster-RCNN Resnet-50 feature pyramid network (FPN) model was trained and validated on 862 manually labeled satellite images. This model was used to perform object detection on satellite imagery, locating and classifying individual solar installations' mounting configuration and type. Model results showed a mean average precision score (mAP) of 77.79%, with the model strongest at detecting fixed-tilt ground mount and fixed-tilt carport installations. The object detection model and its outputs have been incorporated into the Panel-Segmentation package's automated metadata extraction pipeline, which returns the mounting configuration and azimuth for individual solar arrays in satellite imagery. The complete image data set with labels has been released on the U.S. Department of Energy (DOE) DuraMAT DataHub, to encourage further research in this area.

deep learning↗

Optical Wavelength Guided Self-Supervised Feature Learning For Galaxy Cluster Richness Estimate

Most galaxies in the nearby Universe are gravitationally bound to a cluster or group of galaxies. Their optical contents, such as optical richness, are crucial for understanding the co-evolution of galaxies and large-scale structures in modern astronomy and cosmology. The determination of optical richness can be challenging. We propose a self-supervised approach for estimating optical richness from multi-band optical images. The method uses the data properties of the multi-band optical images for pre-training, which enables learning feature representations from a large but unlabeled dataset. We apply the proposed method to the Sloan Digital Sky Survey. The result shows our estimate of optical richness lowers the mean absolute error and intrinsic scatter by 11.84% and 20.78%, respectively, while reducing the need for labeled training data by up to 60%. We believe the proposed method will benefit astronomy and cosmology, where a large number of unlabeled multi-band images are available, but acquiring image labels is costly.

79 ASTRONOMY AND ASTROPHYSICS↗

Count Every Trip: Finding the Uncertainty in Energy Estimates Made from Inferred Travel Modes

To properly inform transport policy and infrastructure changes, transportation related metrics need both measured values and uncertainties of those values. Travel monitoring smartphone apps can record people's travel behavior, but trip data quality is limited by sensor errors, user labeling rates and the accuracy of inference algorithms used for travel diary creation. We discuss the use of phone app recorded travel diary data to estimate energy consumption, and propose the use of propagation of variance to find error bars for such estimates. We define energy consumption for one trip as trip length times the energy intensity per distance unit of the travel mode used. We characterize trip length errors with relative error and inferred trip mode errors with confusion matrix columns. The resulting variances of each measurement are then propagated to the final calculated energy consumption. We tested our uncertainty methods on a dataset that used phone app data combined with prompted recall, consisting of 92,234 labeled trips for over 500,000 miles. Accounting for uncertainty using expected energy intensities and variance propagation gives a dataset-wide aggregate energy consumption percent error of about 9%, within one standard deviation from the truth. Future work could involve applying similar methods to other travel diary based metrics.

ADVANCED PROPULSION SYSTEMS,MATHEMATICS AND COMPUT↗

Count Every Trip: Finding the Uncertainty in Energy Estimates Made from Inferred Travel Modes

To properly inform transport policy and infrastructure changes, transportation related metrics need both measured values and uncertainties of those values. Travel monitoring smartphone apps can record people's travel behavior, but trip data quality is limited by sensor errors, user labeling rates and the accuracy of inference algorithms used for travel diary creation. We discuss the use of phone app recorded travel diary data to estimate energy consumption, and propose the use of propagation of variance to find error bars for such estimates. We define energy consumption for one trip as trip length times the energy intensity per distance unit of the travel mode used. We characterize trip length errors with relative error and inferred trip mode errors with confusion matrix columns. The resulting variances of each measurement are then propagated to the final calculated energy consumption. We tested our uncertainty methods on a dataset that used phone app data combined with prompted recall, consisting of 92,234 labeled trips for over 500,000 miles. Accounting for uncertainty using expected energy intensities and variance propagation gives a dataset-wide aggregate energy consumption percent error of about 9%, within one standard deviation from the truth. Future work could involve applying similar methods to other travel diary based metrics.

ADVANCED PROPULSION SYSTEMS↗

Characterizing Signatures of Geothermal Exploration Data with Machine Learning Techniques: An Application to the Nevada Play Fairway Analysis

We are introducing machine learning methods to the play fairway analysis to generate geothermal potential maps to support the evaluation of geothermal resource potential and the exploration for undiscovered blind geothermal systems in the Nevada Great Basin region. Our project aims to identify new ways to combine the play fairway data and empirically organize relationships between feature weights and labels in an improved workflow. As a means of doing this, we introduce machine learning methods to evaluate the influence of certain geological and geophysical features/feature sets in predicting geothermal favorability. This report highlights promising approaches based on supervised and unsupervised learning methods. First, we demonstrate a filter method applied to supervised classification modeling. The supervised filter method is based on permutation analysis to evaluate every possible feature combination/drop out scenario and rank feature influence based on the performance variance of supervised classification models. Additionally, we present an unsupervised factor analysis based on principal component analysis coupled with a semi-supervised kmeans clustering algorithm. This analysis allows us to identify the optimal number of groups/clusters for training sites and structural settings to identify feature patterns including correlation, variance, and latent and dominant feature relationships. The results from these methods offer a promising avenue for identifying favorable sources of predictive information to identify the locations of blind geothermal systems and furthering our understanding of complex geothermal feature and label relationships in the Great Basin region and beyond.

15 GEOTHERMAL ENERGY↗