Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “TRAINING PLATFORMS”

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 649 records · Page 36

Data augmentation and multimodal learning for predicting drug response in patient-derived xenografts from gene expressions and histology images

Patient-derived xenografts (PDXs) are an appealing platform for preclinical drug studies. A primary challenge in modeling drug response prediction (DRP) with PDXs and neural networks (NNs) is the limited number of drug response samples. We investigate multimodal neural network (MM-Net) and data augmentation for DRP in PDXs. The MM-Net learns to predict response using drug descriptors, gene expressions (GE), and histology whole-slide images (WSIs). We explore whether combining WSIs with GE improves predictions as compared with models that use GE alone. We propose two data augmentation methods which allow us training multimodal and unimodal NNs without changing architectures with a single larger dataset: 1) combine single-drug and drug-pair treatments by homogenizing drug representations, and 2) augment drug-pairs which doubles the sample size of all drug-pair samples. Unimodal NNs which use GE are compared to assess the contribution of data augmentation. The NN that uses the original and the augmented drug-pair treatments as well as single-drug treatments outperforms NNs that ignore either the augmented drug-pairs or the single-drug treatments. In assessing the multimodal learning based on the MCC metric, MM-Net outperforms all the baselines. Our results show that data augmentation and integration of histology images with GE can improve prediction performance of drug response in PDXs.

60 APPLIED LIFE SCIENCES↗

The State of NOS3

The NASA Operational Simulator for Small Satellites (NOS3) showcases some of the Jon McBride Software Testing and Research (JSTAR) laboratories technologies on an open-source platform. NOS3 is a software digital twin providing a virtualized platform inside which you have your traditional flight software, ground software, environmental simulators, and middleware to keep all pieces in sync. NOS3 leverages the core Flight System (cFS), OpenC3 COSMOS, and NASA GSFC’s 42 software as the baseline to which additional research technologies can be developed. Current technologies to be demonstrated include NOS3 Igniter, constellation support, NASA JPL’s SYNOPSIS integration, and NASA GSFC’s OnAir. NOS3 Igniter is a GUI in which you can configure, build, and run your simulation. This along with improvements to the documentation and training available open source aims to reduce the ramp up time with new users and improve accessibility. As constellations introduce another level of complexity, it is important to ensure the baseline design reference mission covers all the basics required and allows users to experiment, understand, and test at all levels of the system. The Science Yield improvement via Onboard Prioritization and Summary of Information Systems (SYNOPSIS) is an open-source tool developed by NASA JPL to enable data prioritization and planning. GSFC’s Onboard Artificial Intelligence Research (OnAIR) enables custom algorithm development written in python to interface with the flight software allowing scientists to develop what they need for the next generation of missions and easily interface back to the traditional flight software. During the presentation, a review and demonstration of the above technologies is planned along with a roadmap.

NOS3↗

Evolutionary Optimization of a Charge Transfer Ionic Potential Model for Ta/Ta-Oxide Heterointerfaces

Heterostructures of tantalum and its oxide are of tremendous technological interest for a myriad of technological applications, including electronics, thermal management, catalysis and biochemistry. In particular, local oxygen stoichiometry variation in TaO x memristors comprising of thermodynamically stable metallic (Ta) and insulating oxide (Ta 2 O 5 ) have been shown to result in fast switching on the subnanosecond timescale over a billion cycles. This rapid switching opens up the potential for advanced functional platforms such as stateful logic operations and neuromorphic computation. Despite its broad importance, an atomistic scale understanding of oxygen stoichiometry variation across Ta/TaO x heterointerfaces, such as during early stages of oxidation and oxide growth, is not well understood. This is mainly due to the lack of a unified interatomic potential model for tantalum oxides that can accurately describe metallic (Ta), ionic (TaO x ) as well as mixed (Ta/TaO x interfaces) bonding environments simultaneously. To address this challenge, we introduce a Charge Transfer Ionic Potential (CTIP) model for Ta/Ta-oxide system by training against lattice parameters, cohesive energies, equations of state (EOS), elastic properties, and surface energies of the various experimentally observed Ta 2 O 5 polymorphs (hexagonal, orthorhombic and monoclinic) obtained from density functional theory (DFT) calculations. The best CTIP parameters are determined by employing a global optimization scheme driven by genetic algorithms followed by local Simplex optimization. Our newly developed CTIP potential accurately predicts structure, thermodynamics, energetic ordering of polymorphs, as well as elastic and surface properties of both Ta and Ta 2 O 5 , in excellent agreement with DFT calculations and experiments. We employ our newly parameterized CTIP potential to investigate the early stages of oxidation and atomic scale mechanisms associated with oxide growth on Ta surface at various temperatures. Furthermore, the CTIP potential developed in this work is an invaluable tool to investigate atomic-scale mechanisms and transport phenomena underlying the response of Ta/TaO x interfaces to external stimuli (e.g, temperature, pressure, strain, electric field etc.), as well as other interesting dynamical phenomena including the physics of switching dynamics in TaO x based memristors and neuromorphic devices.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Multiresonant metasurfaces for arbitrarily broad bandwidth pulse chirping and dispersion compensation

Here, we show that ultrathin metasurfaces with a specific multiresonant response can enable simultaneously arbitrarily strong and arbitrarily broadband dispersion compensation, pulse (de)chirping, and compression or broadening. This breakthrough overcomes the fundamental limitations of both conventional nonresonant approaches (bulky) and modern singly resonant metasurfaces (narrowband) for quadratic phase manipulations of electromagnetic signals. The required nonuniform trains of resonances in the electric and magnetic sheet conductivities that completely control phase delay, group delay, and chirp are rigorously derived and the limitations imposed by fundamental physical constraints are thoroughly discussed. Subsequently, a practical, truncated approximation by finite sequences of physically realizable linear resonances is constructed and the associated error is quantified. By appropriate spectral ordering of the resonances, operation can be achieved either in transmission or reflection mode, enabling full space coverage. The proposed concept is not limited to dispersion compensation, but introduces a generic and powerful ultrathin platform for the spatiotemporal control of broadband real-world signals with a myriad of applications in modern optics, microwave photonics, radar, and communication systems.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

SU-C-304-04: A Compact Modular Computational Platform for Automated On-Board Imager Quality Assurance

Purpose: Traditionally, the assessment of X-ray tube output and detector positioning accuracy of on-board imagers (OBI) has been performed manually and subjectively with rulers and dosimeters, and typically takes hours to complete. In this study, we have designed a compact modular computational platform to automatically analyze OBI images acquired with in-house designed phantoms as an efficient and robust surrogate. Methods: The platform was developed as an integrated and automated image analysis-based platform using MATLAB for easy modification and maintenance. Given a set of images acquired with the in-house designed phantoms, the X-ray output accuracy was examined via cross-validation of the uniqueness and integration minimization of important image quality assessment metrics, while machine geometric and positioning accuracy were validated by utilizing pattern-recognition based image analysis techniques. Results: The platform input was a set of images of an in-house designed phantom. The total processing time is about 1–2 minutes. Based on the data acquired from three Varian Truebeam machines over the course of 3 months, the designed test validation strategy achieved higher accuracy than traditional methods. The kVp output accuracy can be verified within +/−2 kVp, the exposure accuracy within 2%, and exposure linearity with a coefficient of variation (CV) of 0.1. Sub-millimeter position accuracy was achieved for the lateral and longitudinal positioning tests, while vertical positioning accuracy within +/−2 mm was achieved. Conclusion: This new platform delivers to the radiotherapy field an automated, efficient, and stable image analysis-based procedure, for the first time, acting as a surrogate for traditional tests for LINAC OBI systems. It has great potential to facilitate OBI quality assurance (QA) with the assistance of advanced image processing techniques. In addition, it provides flexible integration of additional tests for expediting other OBI quality assurance tests, such as 2D/3D image quality, making completely automated QA possible. Research Funding from Varian Medical Systems Inc. . Dr. Sasa Mutic receives compensation for providing patient safety training services from Varian Medical Systems, the sponsor of this study.

60 APPLIED LIFE SCIENCES↗

Tailoring coherent beam combined laser pulse trains for high peak and average power applications

Laser systems based on coherent beam combination (CBC) that rely on tiled pupil architecture intrinsically carry digital capabilities independently applicable to all three essential characteristics of a laser pulse: amplitude, phase and polarization. Those capabilities allow the far-field energy distribution to be flexibly tailored in real time. Operation in the femtosecond regime at high repetition rates gives access to a wide range of applications requiring both high peak and average powers. We address the task of independent peak versus average power adjustment needed for applications seeking to decouple nonlinear phenomena associated with GW peak power from the thermal load inherent to kW average power operation. The technical solutions proposed are presented in the framework of the Ecole Polytechnique XCAN CBC laser platform (61 independent channels) with an emphasis on thermal management measures implemented to ensure its nominal operation.

47 OTHER INSTRUMENTATION↗

Object and Gas Source Detection with Robotic Platforms in Perceptually-Degraded Environments

In exploration-oriented robotic missions for disaster relief in unknown subterranean environments, it is of prime importance for a human supervisor to rapidly gain situational awareness of salient objects within the environment. In this paper we present an automated object detection pipeline that is adaptable to heterogeneous robots with arbitrary sensor configurations. It has been deployed in time-critical scenarios with multiple collaborative robots in a variety of demanding underground environments. For visually observable objects, detections are made in both the visible and thermal spectra using a state-of-the-art machine learning framework for object detection and classification. Our pipeline can be rapidly adapted to a specific task by using a small, structured dataset to fine-tune a pre-trained convolutional neural network (CNN). Relative localization is separated from the CNN for speed of operation. A robust architecture for localization is used with outlier rejection and a hierarchy of fall-back distance measurement methods. Point-source objects such as gas and WiFi hotspots can also be detected, by tracking signal strength over time and presenting an intuitive visualization on a map. Observations of each object types are presented to the operator in ranked confidence order for final evaluation.

Agha-mohammadi, Ali-akbar↗

Increasing the Scale of the Mass Spectrometry Query Language Compendium with Explainable AI

A significant bottleneck in metabolomics data interpretation is the effective use of domain knowledge to assign structural information based on fragmentation patterns. The mass spectrometry query language (MassQL) aims to make this process accessible and applicable across multiple analysis platforms. While advanced computational methods are capable of predicting compound structures from fragmentation data, AI/ML approaches often rely on complex, opaque criteria that are difficult to interpret or modify. As a result, their predictive patterns cannot be readily translated into human-readable rules, such as those used in MassQL. Here, in this study, we introduce ChemEcho, a machine learning embedding method that converts tandem mass spectrometry data into sparse feature vectors containing peak and neutral mass subformulae to enhance explainable AI/ML-based methods. An advantage of this approach is that decision trees trained using these feature vectors can be directly translated to MassQL. Using a battery of decision trees trained using ChemEcho embeddings to predict molecular attributes, we generated over 1500 MassQL queries for 765 molecular features and evaluated their precision and recall. From these queries, the 50 highest-performing queries were integrated into the MassQL compendium. This set of generated MassQL queries included environmentally and biologically relevant classes such as PFAS and molecules containing phosphate or sulfate substructures. To illustrate the impact these queries would have on a typical metabolomics experiment, these MassQL queries were applied to a public metabolomics data set─resulting in a marked increase in the structural information derived from tandem mass spectra. Access and reuse of these queries is expected to enhance structural annotation in untargeted experiments, leading to more specific claims and advancing many applications in metabolomics.

Harwood, Thomas V. [USDOE Joint Genome Institute (↗

SARS-CoV2 billion-compound docking

Abstract This dataset contains ligand conformations and docking scores for 1.4 billion molecules docked against 6 structural targets from SARS-CoV2, representing 5 unique proteins: MPro, NSP15, PLPro, RDRP, and the Spike protein. Docking was carried out using the AutoDock-GPU platform on the Summit supercomputer and Google Cloud. The docking procedure employed the Solis Wets search method to generate 20 independent ligand binding poses per compound. Each compound geometry was scored using the AutoDock free energy estimate, and rescored using RFScore v3 and DUD-E machine-learned rescoring models. Input protein structures are included, suitable for use by AutoDock-GPU and other docking programs. As the result of an exceptionally large docking campaign, this dataset represents a valuable resource for discovering trends across small molecule and protein binding sites, training AI models, and comparing to inhibitor compounds targeting SARS-CoV-2. The work also gives an example of how to organize and process data from ultra-large docking screens.

60 APPLIED LIFE SCIENCES↗

Expert Connections Empower Bankable Solar Projects

The International Solar Alliance (ISA) formed in 2015 to serve as a collaborative platform for increasing solar energy technology deployment worldwide. Currently, the ISA includes more than 65 signatory countries that are working with global partners to improve their access to affordable and reliable clean energy. In support of ISA member countries, the U.S. Department of Energy and the Australian Government's Department of Environment and Energy selected the Clean Energy Solutions Center (Solutions Center) to help mobilize private investment in solar technologies associated with this global effort. The Solutions Center is partnering with ISA selected partners to provide a range of technical assistance services, training programs, and other capacity building resources. The U.S. Department of Energy's National Renewable Energy Laboratory serves as the operating agent for the Clean Energy Solutions Center. The International Solar Alliance is Headquartered in India where the National Institute of Solar Energy serves as interim Secretariat.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Energy Efficient Stochastic-Based Deep Spiking Neural Networks for Sparse Datasets

With large deep neural networks (DNNs) neces- sary to solve complex and data-intensive problems, energy efficiency is a key bottleneck for effectively deploying DL in the real world. Deep spiking NNs have gained much research at- tention recently due to the interest in building biological neural networks and the availability of neuromorphic platforms, which can be orders of magnitude more energy efficient compared to CPUs and GPUs. Although spiking NNs have proven to be an efficient technique for solving many machine learning and computer vision problems, to the best of our knowledge, this is the first attempt to adapt spiking NNs to sparse datasets. In this paper, we study the behaviour of spiking NNs in handling NLP datasets and the sparsity in their data representation. Then, we propose a novel framework for spiking NN using the concept of stochastic computing. Specifically, instead of generating spike trains with firing rates proportional to the intensity of each value in the feature set separately, the whole feature set is treated as a distribution function and a stochastic spiking train that follow this distribution is generated. This framework reduces the connectivity between NN layers from O(N) to O(logN). Also, it encodes input data differently and make suitable to handle sparse datasets. Finally, the framework achieves high energy efficiency since it uses Integrate and Fire neurons same as conventional spiking NNs. The results show that our proposed stochastic-based SNN achieves nearly the same accuracy as the original DNN on MNIST dataset, and it has better performance than state- of-the-art SNN. Besides that stochastic-based SNN is energy efficient, where the fully connected DNN, the conventional SNN, and the data normalized SNN consume 38.24, 1.83, and 1.85 -times more energy than the stochastic-based SNN, respectively. For sparse datasets, including IMDb and In-House clinical datasets, stochastic-based SNN achieves performance comparable to that of the conventional DNN. However, the conventional spiking NN has a significant decline in classification accuracy.

Alawad, Mohammed↗

NEON AOP foliar trait maps, maps of model uncertainty estimates, and conifer map, East River, CO 2018

This data package contains mapped trait estimates and their uncertainties, and conifer map, for the National Ecological Observatory Network's Airborne Observation Platform survey data acquired over the Upper East River, Colorado in 2018. For full details, please see associated reference. in brief, trait models were developed independently for needle and non-needle leaf species using partial least squares regression (PLSR) using ground data from additional datasets: doi:10.15485/1618130, doi:10.15485/1618132, and doi:10.15485/1631278, merged with extracted spectral data from doi:10.15485/1618131. We separated vegetated pixels into needle and non-needle classes in order to generate a classification map based on the spectral differences between these leaf types (conifer.tif). We trained a deep learning model with custom architecture, detailed in Chadwick et al. In Press. The model performed with 0.998 true positive rate and 0.982 true negative rate, with ‘positives’ being non-needle identification. We then utilized PLSR to generate models of foliar traits for each leaf type. So that we could also map uncertainty in these predictions, we generated ten different models for needle and non-needle leaf species using different testing holdout sets of discrete sites. Each of these models was developed with a 100-fold cross validation procedure that utilized a 70% training set and 30% validation set with each fold, and then assessed based on the 10% of testing sites that were not included in that model’s development. The mean predicted value across the 10 models is used for the trait estimate in each pixel across the study area. The models are applied according to the leaf type designation in the conifer.tif map. The errors are the standard deviation across the 10 different models developed, with high error suggesting instability in model prediction and areas where values may not be reliable for ecological inference. These maps are only applied to areas with a NDVI > 0.5 to exclude non-vegetated areas. Shade masks could be applied to these data (doi:10.15485/1618131), but have not been for this data package. These data are also available on Google Earth Engine: https://code.earthengine.google.com/?asset=users/kdc/ER_NEON

54 ENVIRONMENTAL SCIENCES↗

Continuous integration data-driven platform of industrial-scale subsurface storage for real-time analytics

This project helped address the growing need for efficient and scalable models to support geological carbon and energy storage, which are crucial for achieving net-zero emissions. Traditionally accurate high-fidelity numerical models have been used to simulate relevant storage processes under a handful of processes, however such models are computationally demanding, making uncertainty quantification impractical. Consequently, we first developed a machine learning framework, based on Graph Neural Operators (GNOs), to improving the accuracy of model predictions for a fixed computational budget. We then developed an Ensemble of Improved Neural Operators (ENO), which uses bagging and Monte Carlo dropout techniques, to further improve prediction accuracy. Lastly, we developed the way to explain progressive transfer learning methods to reduce the amount of training data and computational cost of training (i.e., reduce trainable parameters) when using our models for multiple storage sites. Our numerical investigation, which used real-world case studies, demonstrated that our framework can significantly improve the safety and efficiency of geological storage operations, with potential applications in other domains such as geothermal reservoirs and climate modeling.

54 ENVIRONMENTAL SCIENCES↗

Concurrent Relaxation through Accelerated Deep Learning

CRADL captures performance metrics of machine learning algorithms operating on mesh data from multiphysics codes This proxy application is a tool to explore scalability of inference on HPC platforms, and also gather performance metrics for inference on new machine learning specific hardware. CRADL is designed to give users as fine a control as possible over an inference simulation. Users may select the number of cycles, amount of data, and batch size to pass to the accelerator of choice. Additionally the user may select a number of performance optimization libraries and flags. CRADL comes packaged with a repository of anonymized multi-physics simulation data, as well as a pretrained model for inference. The code allows a user to load their own pre-trained model and data if they wish. The code can operate in multiple parallelization schemes, with performance enhancing options such as half-precision libraries, PyTorch benchmarking, and pinned memory with non-blocking data transfers.

Zieb, KristoferJ.↗

Final report for LDRD project 11-0783 : directed robots for increased military manpower effectiveness.

The purpose of this LDRD is to develop technology allowing warfighters to provide high-level commands to their unmanned assets, freeing them to command a group of them or commit the bulk of their attention elsewhere. To this end, a brain-emulating cognition and control architecture (BECCA) was developed, incorporating novel and uniquely capable feature creation and reinforcement learning algorithms. BECCA was demonstrated on both a mobile manipulator platform and on a seven degree of freedom serial link robot arm. Existing military ground robots are almost universally teleoperated and occupy the complete attention of an operator. They may remove a soldier from harm's way, but they do not necessarily reduce manpower requirements. Current research efforts to solve the problem of autonomous operation in an unstructured, dynamic environment fall short of the desired performance. In order to increase the effectiveness of unmanned vehicle (UV) operators, we proposed to develop robots that can be 'directed' rather than remote-controlled. They are instructed and trained by human operators, rather than driven. The technical approach is modeled closely on psychological and neuroscientific models of human learning. Two Sandia-developed models are utilized in this effort: the Sandia Cognitive Framework (SCF), a cognitive psychology-based model of human processes, and BECCA, a psychophysical-based model of learning, motor control, and conceptualization. Together, these models span the functional space from perceptuo-motor abilities, to high-level motivational and attentional processes.

99 GENERAL AND MISCELLANEOUS↗

Summary of NDC Capacity Building Workshop and Regional Seismic Travel Time (RSTT) in combination with Data Sharing and Integration Training

The NNSA Seismic Cooperation Program (SCP) sponsored Stephen Myers (LLNL), Michael Begnaud (LANL), Brian Young (SNL) and Istvan Bondar (Research Center for Astronomy and Earth Sciences, Hungary) to serve as a presenters/trainers at the “NDC Capacity Building Workshop and Regional Seismic Travel Time (RSTT) in combination with Data Sharing and Integration Training” September 4-8 2022 in Muscat, Oman (See Appendix A for the agenda). The workshop and training (workshop from here forward) was organized by the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) Provisional Technical Secretariat (PTS). The first half of the week was devoted to NDC workshop activities, and the second half was devoted to RSTT training. Fifty-five participants from 27 countries and the CTBTO-PTS attended the 5-day workshop (See Appendix B for list of participants and countries of origin). Presentations from the PTS described the International Monitoring System (IMS), International Data Centre (IDC) products, and metrics of regional data utilization. Contributed presentations from each country’s scientists included descriptions of regional and national networks, methods of data analysis, and needs for material and technical assistance. Training included an overview of the RSTT method and instruction on how to locate seismic events with the iLoc program, which utilizes RSTT travel times to reduce bias in event location estimates. Methods of seismic tomography and the need for a high-quality tomographic set, including seismological “ground truth”, were emphasized. Seismological “ground truth” or “GT” is a term that has come to mean both events with known location and events with well-characterized locations that are estimated using seismological data, typically with epicenter accuracy of 5 km or better. Notably, the instructional platform has migrated from UNIX shell scripts to Jupyter Notebooks. Jupyter Notebooks have the advantage being more visually intuitive, including display of graphics within the notebook. Each notebook includes every processing step that participants need to reproduce the entire exercise.

58 GEOSCIENCES↗

SetBERT: the deep learning platform for contextualized embeddings and explainable predictions from high-throughput sequencing

MOTIVATION: High-throughput sequencing (HTS) is a modern sequencing technology used to profile microbiomes by sequencing thousands of short genomic fragments from the microorganisms within a given sample. This technology presents a unique opportunity for artificial intelligence to comprehend the underlying functional relationships of microbial communities. However, due to the unstructured nature of HTS data, nearly all computational models are limited to processing DNA sequences individually. This limitation causes them to miss out on key interactions between microorganisms, significantly hindering our understanding of how these interactions influence the microbial communities as a whole. Furthermore, most computational methods rely on post-processing of samples which could inadvertently introduce unintentional protocol-specific bias. RESULTS: Addressing these concerns, we present SetBERT, a robust pre-training methodology for creating generalized deep learning models for processing HTS data to produce contextualized embeddings and be fine-tuned for downstream tasks with explainable predictions. By leveraging sequence interactions, we show that SetBERT significantly outperforms other models in taxonomic classification with genus-level classification accuracy of 95%. Furthermore, we demonstrate that SetBERT is able to accurately explain its predictions autonomously by confirming the biological-relevance of taxa identified by the model. AVAILABILITY AND IMPLEMENTATION: All source code is available at https://github.com/DLii-Research/setbert. SetBERT may be used through the q2-deepdna QIIME 2 plugin whose source code is available at https://github.com/DLii-Research/q2-deepdna.

Ludwig, David W↗

Advanced Gas Turbine (AGT) power-train system development

Technical work on the design and component testing of a 74.5 kW (100 hp) advanced automotive gas turbine is described. Selected component ceramic component design, and procurement were tested. Compressor tests of a modified rotor showed high speed performance improvement over previous rotor designs; efficiency improved by 2.5%, corrected flow by 4.6%, and pressure ratio by 11.6% at 100% speed. The aerodynamic design is completed for both the gasifier and power turbines. Ceramic (silicon carbide) gasifier rotors were spin tested to failure. Improving strengths is indicated by burst speeds and the group of five rotors failed at speeds between 104% and 116% of engine rated speed. The emission results from combustor testing showed NOx levels to be nearly one order of magnitude lower than with previous designs. A one piece ceramic exhaust duct/regenerator seal platform is designed with acceptable low stress levels.

Helms, H. E.↗