Search NASA⌕ Search

SEARCH · Search NASA

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 721 records · Page 40

Maximizing Spaceflight Biological Data with Omics Analytics: The NASA GeneLab Database

NASA’s GeneLab includes an open-access repository of some 250+ omics datasets generated by biological experiments relevant to spaceflight including simulated cosmic radiation and microgravity. In order to maximize the intelligibility of these data, particularly for users with limited bioinformatics background, GeneLab has become a knowledgebase platform converting raw genetic and proteomic signatures found in flight samples into biological and physiological meanings. A large community of more than 100 scientists has rallied behind GeneLab and organized into four Analysis Working Groups (AWGs: Animal, Plant, Microbe, and Multi-Omics). Together, the AWGs have gained scientific recognition worldwide by establishing a consortium in charge of adopting new complex standards for data analysis workflows and omics sample processing in a rapidly evolving field. We will demonstrate the usage of the repository with smart search capability, an online controlled-access toolshed "Galaxy" to process user data with vetted standard workflows, a workspace for data sharing and a data submission portal with ontology control for better metadata curation. The GeneLab visualization portal will also be demonstrated, showing how anyone without formal training in bioinformatics can now browse the space biology omics data to discover new biology and potential solutions to improve life in space.

Sylvain Vincent Costes↗

GeneLab: The NASA System Biology Platform for Space Omics Repository, Analysis and Visualization

NASA’s GeneLab includes an open-access repository of some 250+ omics datasets generated by biological experiments relevant to spaceflight including simulated cosmic radiation and microgravity. In order to maximize the intelligibility of these data, particularly for users with limited bioinformatics background, GeneLab has become a knowledgebase platform converting raw genetic and proteomic signatures found in flight samples into biological and physiological meanings. A large community of more than 100 scientists has rallied behind GeneLab and organized into four Analysis Working Groups (AWGs: Animal, Plant, Microbe, and Multi-Omics). Together, the AWGs have gained scientific recognition worldwide by establishing a consortium in charge of adopting new complex standards for data analysis workflows and omics sample processing in a rapidly evolving field. We will demonstrate the usage of the repository with smart search capability, an online controlled-access toolshed "Galaxy" to process user data with vetted standard workflows, a workspace for data sharing and a data submission portal with ontology control for better metadata curation. The GeneLab visualization portal will also be demonstrated, showing how anyone without formal training in bioinformatics can now browse the space biology omics data to discover new biology and potential solutions to improve life in space.

GeneLab↗

Effects of Replacing Treadmill Running with Alternative Exercise Countermeasures During Long-Duration Spaceflight on Astronaut Health and Performance

INTRODUCTION Current exercise countermeasures on the International Space Station (ISS) include treadmill running, cycle ergometry, and resistive exercise, which are used to protect crewmember health and performance during long-duration spaceflight. However, exploration vehicles for Artemis missions to the Moon and beyond will have volume and power restrictions, requiring exercise hardware to have a smaller footprint and use fewer resources. Thus, recent efforts have focused on developing exercise devices that provide both aerobic and resistive capabilities on one platform without including a treadmill. The European Enhanced Exploration Exercise Device [E4D] is one such apparatus. It is critical to validate the efficacy of using exploration-focused exercise modalities to preserve muscle strength, aerobic fitness, bone density, and sensorimotor performance. Thus, the aim of this study is to determine the effectiveness of using nominal ISS exercise devices for an entire mission compared to exploration-forward exercise modalities to determine if a treadmill is required to maintain current levels of protection during long-duration missions. METHODS Crewmembers are assigned to one of three groups: 1) Controls (n ≥ 40), who partake in nominal exercise on the ISS, including running on the Treadmill with Vibration Isolation and Stabilization 2 (T2), cycling on the Cycle Ergometer with Vibration Isolation and Stabilization (CEVIS), and strength training on the Advanced Resistive Exercise Device (ARED); 2) Active Group 1, who exercise on the CEVIS and ARED only (n = 8); and 3) Active Group 2, who perform aerobic and resistive exercise on the E4D only (n = 8). For Active Group 1, nominal exercise on T2 will be replaced with corresponding exercise on CEVIS. For Active Group 2, a dedicated exercise prescription will be designed to maximize the capabilities of the E4D to include resistive exercise, cycle ergometry, rowing, and rope pulling. Crewmembers in both Active groups are not permitted to perform treadmill exercise for the entirety of their flight. Health and performance markers, including bone mineral density (dual-energy x-ray absorptiometry [DXA]), body composition (DXA), cardiovascular fitness (cycle VO2peak), muscle strength and endurance (isometric/isokinetic testing, power endurance testing), sensorimotor performance (sit-to-stand, obstacle course), postural control (computerized dynamic posturography), and blood and urine biochemical markers of bone metabolism will be assessed before, during, and after spaceflight. RESULTS Fifteen subjects (4 Active [CEVIS + ARED], 11 Control) have been recruited. Data collection is ongoing. CONCLUSIONS This study will assess the efficacy of using exploration exercise modalities, including removing the treadmill exercise capability or exclusively using the E4D, compared to the nominal ISS exercise regimen across an entire mission on bone, muscle, aerobic, and sensorimotor health and performance. Findings from this study will help provide a recommendation on whether these exploration exercise modalities can sufficiently protect against physiological deconditioning during spaceflight or whether a treadmill may be required to maintain current levels of protection during future exploration class spaceflight missions. Supported by the NASA Human Research Program and NASA Exploration Capabilities

A N Varanoske↗

Evaluating the potential of disaggregated memory systems for HPC applications

Summary Disaggregated memory is a promising approach that addresses the limitations of traditional memory architectures by enabling memory to be decoupled from compute nodes and shared across a data center. Cloud platforms have deployed such systems to improve overall system memory utilization, but performance can vary across workloads. High‐performance computing (HPC) is crucial in scientific and engineering applications, where HPC machines also face the issue of underutilized memory. As a result, improving system memory utilization while understanding workload performance is essential for HPC operators. Therefore, learning the potential of a disaggregated memory system before deployment is a critical step. This paper proposes a methodology for exploring the design space of a disaggregated memory system. It incorporates key metrics that affect performance on disaggregated memory systems: memory capacity, local and remote memory access ratio, injection bandwidth, and bisection bandwidth, providing an intuitive approach to guide machine configurations based on technology trends and workload characteristics. We apply our methodology to analyze thirteen diverse workloads, including AI training, data analysis, genomics, protein, fusion, atomic nuclei, and traditional HPC bookends. Our methodology demonstrates the ability to comprehend the potential and pitfalls of a disaggregated memory system and provides motivation for machine configurations. Our results show that eleven of our thirteen applications can leverage injection bandwidth disaggregated memory without affecting performance, while one pays a rack bisection bandwidth penalty and two pay the system‐wide bisection bandwidth penalty. In addition, we also show that intra‐rack memory disaggregation would meet the application's memory requirement and provide enough remote memory bandwidth.

Ding, Nan↗

Hardware-Based Emulator with Deep Learning Model for Building Energy Control and Prediction Based on Occupancy Sensors’ Data

Heating, ventilation, and air conditioning (HVAC) is the largest source of residential energy consumption. Occupancy sensors’ data can be used for HVAC control since it indicates the number of people in the building. HVAC and sensors form a typical cyber-physical system (CPS). In this paper, we aim to build a hardware-based emulation platform to study the occupancy data’s features, which can be further extracted by using machine learning models. In particular, we propose two hardware-based emulators to investigate the use of wired/wireless communication interfaces for occupancy sensor-based building CPS control, and the use of deep learning to predict the building energy consumption with the sensor data. We hypothesize is that the building energy consumption may be predicted by using the occupancy data collected by the sensors, and question what type of prediction model should be used to accurately predict the energy load. Another hypothesis is that an in-lab hardware/software platform could be built to emulate the occupancy sensing process. The machine learning algorithms can then be used to analyze the energy load based on the sensing data. To test the emulator, the occupancy data from the sensors is used to predict energy consumption. The synchronization scheme between sensors and the HVAC server will be discussed. We have built two hardware/software emulation platforms to investigate the sensor/HVAC integration strategies, and used an enhanced deep learning model—which has sequence-to-sequence long short-term memory (Seq2Seq LSTM)—with an attention model to predict the building energy consumption with the preservation of the intrinsic patterns. Because the long-range temporal dependencies are captured, the Seq2Seq models may provide a higher accuracy by using LSTM architectures with encoder and decoder. Meanwhile, LSTMs can capture the temporal and spatial patterns of time series data. The attention model can highlight the most relevant input information in the energy prediction by allocating the attention weights. The communication overhead between the sensors and the HVAC control server can also be alleviated via the attention mechanism, which can automatically ignore the irrelevant information and amplify the relevant information during CNN training. Our experiments and performance analysis show that, compared with the traditional LSTM neural network, the performance of the proposed method has a 30% higher prediction accuracy.

Ye, Zhijing↗

KG-Hub—building and exchanging biological knowledge graphs

Knowledge graphs (KGs) are a powerful approach for integrating heterogeneous data and making inferences in biology and many other domains, but a coherent solution for constructing, exchanging, and facilitating the downstream use of KGs is lacking. Here we present KG-Hub, a platform that enables standardized construction, exchange, and reuse of KGs. Features include a simple, modular extract–transform–load pattern for producing graphs compliant with Biolink Model (a high-level data model for standardizing biological data), easy integration of any OBO (Open Biological and Biomedical Ontologies) ontology, cached downloads of upstream data sources, versioned and automatically updated builds with stable URLs, web-browsable storage of KG artifacts on cloud infrastructure, and easy reuse of transformed subgraphs across projects. Current KG-Hub projects span use cases including COVID-19 research, drug repurposing, microbial–environmental interactions, and rare disease research. KG-Hub is equipped with tooling to easily analyze and manipulate KGs. KG-Hub is also tightly integrated with graph machine learning (ML) tools which allow automated graph ML, including node embeddings and training of models for link prediction and node classification.

59 BASIC BIOLOGICAL SCIENCES↗

Predictive Data-driven Platform for Subsurface Energy Production

Subsurface energy activities such as unconventional resource recovery, enhanced geothermal energy systems, and geologic carbon storage require fast and reliable methods to account for complex, multiphysical processes in heterogeneous fractured and porous media. Although reservoir simulation is considered the industry standard for simulating these subsurface systems with injection and/or extraction operations, reservoir simulation requires spatio-temporal “Big Data” into the simulation model, which is typically a major challenge during model development and computational phase. In this work, we developed and applied various deep neural network-based approaches to (1) process multiscale image segmentation, (2) generate ensemble members of drainage networks, flow channels, and porous media using deep convolutional generative adversarial network, (3) construct multiple hybrid neural networks such as convolutional LSTM and convolutional neural network-LSTM to develop fast and accurate reduced order models for shale gas extraction, and (4) physics-informed neural network and deep Q-learning for flow and energy production. We hypothesized that physicsbased machine learning/deep learning can overcome the shortcomings of traditional machine learning methods where data-driven models have faltered beyond the data and physical conditions used for training and validation. We improved and developed novel approaches to demonstrate that physics-based ML can allow us to incorporate physical constraints (e.g., scientific domain knowledge) into ML framework. Outcomes of this project will be readily applicable for many energy and national security problems that are particularly defined by multiscale features and network systems.

58 GEOSCIENCES↗

Knowledge management information tool Web analytics - 15182

The Knowledge Management Information Tool (KM-IT) is a web-based [01] system developed to maintain and preserve the Department of Energy's (DOE's) D and D knowledge base. The system was developed by Florida International University's Applied Research Center (FIU-ARC) with the support of the D and D community including the DOE Office of Environmental Management, and with the active collaboration and support of the DOE's Energy Facility Contractors Group (EFCOG). The KM-IT system is a D and D community driven system tailored to serve the technical issues faced by the D and D workforce across the DOE Complex. The D and D KM-IT is an interactive web-based system operational and available for free access at www.dndkm.org. [02] The D and D KM-IT is currently composed of the following tailored D and D modules: Web Crawler, Hotline, Technology, Document Library (including Hanford's and Savannah River Sites' ALARA Center Reports), Specialist Directory, Lessons Learned, Best Practices, Video/Picture Library, Vendors, Collaboration Tools, Training, and Mobile Systems. The Web Crawler dynamically searches through the D and D KM-IT repository as well as the web and displays search results based on the search criteria. The Hotline allows registered users to post questions/problems related to D and D and to receive solutions from a subject-matter specialist. The Technology module serves as repository of D and D technologies and contains technology/ demonstration fact sheets and vendor information. D and D community members can add Lessons Learned, Best Practices, Documents, and Pictures/Videos to the D and D KM-IT repository, after a formal approval process. The Vendor module provides a directory of commercial vendors who provide D and D related technologies, supplies, and services. D and D KM-IT makes excellent use of the knowledge that exists within the D and D community by allowing D and D project managers around the DOE complex to collaborate by sharing innovative ideas, past experiences, and practices and by maintaining a directory of subject matter specialists. The mobile version of the KM-IT can be accessed from m.dndkm.org [03] and currently has six active modules: Vendor, Technology, Picture/Video Library, Hotline, Lessons Learned and Specialist Directory. Web analytics is a key tool being implemented by the D and D KM-IT project team in an effort to enhance the system to better meet users' needs. The D and D KM-IT project team incorporates web analytics as part of its operational elements to analyze information and key data points on how the system is being used. Web analytics is the measurement, collection, analysis and reporting of internet data for the purposes of understanding and optimizing web usage. The use of web analytics retains or attracts D and D community members to access and share knowledge through virtual collaboration on the KM-IT platform. D and D KM-IT web analytics measures details on the overall usage of the system without tracking personal information or specific user behavior: how many users visited the application, how many of those users were unique visitors, how visitors come to the system (i.e., by following a link to the system or by using the web site URL via bookmark or directly typing the address), what keywords were used to perform searches with the search engine, how long the average user stayed on a given screen or on the entire system and what links were clicked on. This information is used to identify which areas of the system are more popular than the others and provides information that can be used to streamline the application for a better user experience. (authors)

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Convergent Protocols for Computing Protein–Ligand Interaction Energies Using Fragment-Based Quantum Chemistry

Fragment-based quantum chemistry methods offer a way to sidestep the steep nonlinear scaling of electronic structure calculations so that large molecular systems can be investigated using high-level methods. Here, we use fragmentation to compute protein–ligand interaction energies in systems with several thousand atoms, using a new software platform for managing fragment-based calculations that implements a screened many-body expansion. Convergence tests using a minimal-basis semiempirical method (HF-3c) indicate that two-body calculations, with single-residue fragments and simple hydrogen caps, are sufficient to reproduce interaction energies obtained using conventional supramolecular electronic structure calculations, to within 1 kcal/mol at about 1% of the computational cost. We also demonstrate that the HF-3c results are illustrative of trends obtained with density functional theory in basis sets up to augmented quadruple-ζ quality. Strategic deployment of fragmentation facilitates the use of converged biomolecular model systems alongside high-quality electronic structure methods and basis sets, bringing ab initio quantum chemistry to systems of hitherto unimaginable size. This will be useful for generation of high-quality training data for machine learning applications.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Probing Coherent States of Light and Matter in Two-Dimensional Semiconductors. Final Report

The ability to enhance light-matter interactions in engineered optical environments is well-established in micro- and nano-photonics. New classes of materials bring new rich correlations between spin, momentum, and light polarization that can be exploited in these photonic systems for information processing and low-power electronics. Interfacing these novel material features with strong optical interactions suggests the compelling capability to create hybrid light-matter systems harnessing low-energy and protected material properties such as spin. The long-term objective of this program was to investigate the interplay between pseudospin in 2D materials and light, revealing new coherent phenomena. The materials studied are the monolayer transition metal dichalcogenides (TMDs), which support polarization-sensitive optical transitions when isolated to a single sub-nanometer crystal layer in thickness. In particular, monolayer TMD semiconductors such as MoS 2 exhibit degenerate valleys in momentum space with distinct spin character. These valleys can be separately addressed by circularly polarized light, providing a tool for manipulation. Integrating this valley pseudospin of excitons in monolayer TMDs with engineered strong optical excitations to create novel coherent phenomena was a key long-term goal of this research. The scope of this research primarily focused on studying light-matter interactions of valley-sensitive excitations in monolayer TMDs. The scope also included the integration of monolayer materials that support valley-polarized emission into photonic devices. This approach extends prior work on exciton-polaritons in 2D materials toward a platform with far more control and improved light input and output characteristics to aid scalability of these polarization-sensitive phenomena to many-body, long-range optical networks. During the course of the five-year Early Career program, the approaches led to successful progress on each proposed aim. Initial successful work seeded additional research that developed from the main themes, while this program focus evolved toward coherent manipulation of valley polaritons and integration with photonics. Both of these advanced topics set the stage for new capabilities in future innovative research. The results of this project provide advances in materials and methods relevant for new visions of opto-electronics and photonics that can improve information processing in an era of increased focus on quantum and coherent phenomena. The research has also contributed to the training of two postdoctoral scholars and five graduate students.

2D Materials↗

Satellite Data Simulator Unit: A Multisensor, Multispectral Satellite Simulator Package

Several multisensor simulator packages are being developed by different research groups across the world. Such simulator packages [e.g., COSP , CRTM, ECSIM, RTTO, ISSARS (under development), and SDSU (this article), among others] share overall aims, although some are targeted more on particular satellite programs or specific applications (for research purposes or for operational use) than others. The SDSU or Satellite Data Simulator Unit is a general-purpose simulator composed of Fortran 90 codes and applicable to spaceborne microwave radiometer, radar, and visible/infrared imagers including, but not limited to, the sensors listed in a table. That shows satellite programs particularly suitable for multisensor data analysis: some are single satellite missions carrying two or more instruments, while others are constellations of satellites flying in formation. The TRMM and A-Train are ongoing satellite missions carrying diverse sensors that observe clouds and precipitation, and will be continued or augmented within the decade to come by future multisensor missions such as the GPM and Earth-CARE. The ultimate goals of these present and proposed satellite programs are not restricted to clouds and precipitation but are to better understand their interactions with atmospheric dynamics/chemistry and feedback to climate. The SDSU's applicability is not technically limited to hydrometeor measurements either, but may be extended to air temperature and humidity observations by tuning the SDSU to sounding channels. As such, the SDSU and other multisensor simulators would potentially contribute to a broad area of climate and atmospheric sciences. The SDSU is not optimized to any particular orbital geometry of satellites. The SDSU is applicable not only to low-Earth orbiting platforms as listed in Table 1, but also to geostationary meteorological satellites. Although no geosynchronous satellite carries microwave instruments at present or in the near future, the SDSU would be useful for future geostationary satellites with a microwave radiometer and/or a radar aboard, which could become more feasible as engineering challenges are met. In this short article, the SDSU algorithm architecture and potential applications are reviewed in brief.

Masunaga, Hirohiko↗

A Methodology to Evaluate the Grid Reliability Impact of Oscillations Induced by Large Loads

The rapid growth of hyperscale AI data centers is bringing renewed attention to the reliability risk that sustained forced oscillations pose to bulk power systems, with cyclic computational workloads emerging as a new forcing source. Unlike the broadband, stochastic disturbances from traditional industrial loads such as arc furnaces, AI training and inference facilities can inject large active power swings concentrated at specific frequencies over extended durations - characteristics that existing grid planning practices do not account for. While the North American Electric Reliability Corporation (NERC) has recognized this gap and called for system-level studies of large load interconnections, no standardized methodology exists to screen, simulate, and quantify these risks at the planning stage. This report presents the Risk Assessment Tool for Large Load-induced Events (RATLLE), a Python-based, publicly available script suite developed at the Pacific Northwest National Laboratory to evaluate bulk power system reliability risks from data center-induced oscillations. RATLLE implements a three-module workflow: a screening module that identifies vulnerable interconnection locations and excitable system modes; a simulation module that models cyclic data center load behavior using a commercial positive sequence simulation platform; and an analysis module that computes risk metrics and generates interactive visualization dashboards. The risk metrics, formulated around simulation observables, map oscillation impacts to a three-stage severity scale spanning latent equipment fatigue through imminent cascading failure. The methodology is demonstrated on two Western Electricity Coordinating Council (WECC) system models: a publicly available 240-bus reduced representation and a detailed 2031 Heavy Winter planning case. Case studies illustrate that even modest 50 MW forced oscillations at resonant frequencies can produce wide-area power swings, N-1 security constraint violations, and cascading generator trips through protection actions - outcomes that would not occur under normal operating conditions without oscillations present. The results underscore the need for standardized oscillation impact assessment in large load interconnection studies and provide a reproducible, extensible framework for utilities to adopt or customize within their existing planning workflows.

Biswas, Shuchismita↗

Cyote-attack Chain Estimator

Attack Chain Estimator (ACE) Application Overview The Attack Chain Estimator (ACE) Application is a sophisticated tool designed for the ingestion, classification, sequencing, and enrichment of cybersecurity threat reports. This application leverages advanced machine learning models and extensive historical data to provide comprehensive insights into cyber threats, specifically targeting Industrial Control Systems (ICS). Purpose The primary functions of the ACE Application include: Ingestion of Cybersecurity Threat Reporting: Capable of ingesting text-based threat reports in markdown or text file format. Supports ingestion of structured data from other sources in STIX/JSON format. Classification of Report’s Text-Based Events: Utilizes a DeBERTa classifier, specifically trained on cybersecurity data, to map the events to MITRE ATT&CK for ICS Tactics and Techniques. Classification is performed using multiple Jupyter notebooks and machine learning workflows hosted as FastAPI microservices: regex_data deberta_base_35_train_hft_classifier_mlflow.ipynb hft_regex_classifier_mlflow.ipynb param_train_hft_classifier_mlflow.ipynb regex_tactic_tech.ipynb Ordering of Tactics, Techniques, and Observable Events: Sequences the identified tactics, techniques, and events to form a coherent attack chain. Enrichment with Historical Attack Chain Details: Enhances the attack chain with details from historical attacks using a Markov model developed from CyOTE Precursor Analysis Report data. The Markov model is available as a FastAPI endpoint for seamless integration. Enrichment with Adversary Emulation Capabilities Data: Integrates adversary emulation capabilities data using MITRE Caldera for OT adversary abilities UUIDs. Export of Output Files: Provides options to export the enriched attack chain in JSON or CSV formats. Routing of Output to Other Applications: Facilitates routing of output to various platforms and applications, including: Threat Intelligence Platforms COREII Scout for Threat Intelligence Analysis COREII Modeling and Simulation for Adversary Emulation Technical Description The ACE Application is an advanced cybersecurity tool designed to provide detailed threat analysis and sequence generation. It is built on a robust architecture that integrates natural language processing, machine learning, and historical data modeling. Key Components: Data Ingestion Module: Handles the input of threat reports and data from various formats, ensuring flexibility in data sources. Classification Engine: Employs DeBERTa-based classifiers hosted as FastAPI microservices to analyze and classify threat report events in accordance with the MITRE ATT&CK framework for ICS. Sequence Generator: Orders the classified events into a logical attack chain, providing clear insight into the sequence of tactics and techniques used in the threat. Enrichment Engine: Integrates historical data and adversary emulation capabilities to enhance the attack chain with valuable context and additional details. The historical data enrichment is powered by a Markov model, which is available as a FastAPI endpoint. Export and Routing Module: Facilitates the export of the enriched attack chain in multiple formats and routes the output to designated applications for further analysis or emulation.

Paul, Tony [Idaho National Laboratory (INL), Idaho↗

Combinatorial Evaluation of Physical Feature Engineering, Classical Machine Learning, and Deep Learning Models for Synchrophasor Data at Scale

A major objective of the project was to train and evaluate the effectiveness of multiple event and anomaly detection, identification and classification deep temporal learning models for processing of real-time phasor measurement unit (PMU) data streams. A vast dataset, consisting of two years of phasor measurements from all three U.S. Interconnections, was curated and released by the Department of Energy (DOE) through Pacific Northwest National Laboratory (PNNL). The dataset also included an event log that provided event times and types (e.g. generator trips, line trips, planned service events, transformer operations, etc.). Our analysis of this dataset addressed six (6) of the eleven (11) research priorities identified in Funding Opportunity Announcement (FOA) DE-FOA-0001861 “Big Data Analysis of Synchrophasor Data” (FOA 1861). Rather than being limited to pre-determined specific algorithms, this project relied on the uniquely structured, highly performant underlying time series database capabilities of the PredictiveGrid platform to assess the vast dataset utilizing a wide variety of algorithms.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Distributed-Memory Sparse Deep Neural Network Inference Using Global Arrays

Partitioned Global Address Space (PGAS) models exhibit tremendous promise in developing efficient and productive distributed-memory parallel applications. They have been used extensively in scientific computations due to conveniently offering a ``shared-memory''-like model and convenient interfaces that separate communication with synchronization. Traditionally, PGAS communication models have been applied to dense/contiguously distributed data, but most modern applications depict varied levels of sparsity. Existing PGAS models require certain adaptations to support distributed sparse computations, since associated computations often require matrix arithmetic, in addition to data movement. The Global Arrays toolkit from Pacific Northwest National Laboratory (PNNL) is one of the earliest PGAS models to combine one-sided data communication and distributed matrix operations and is still used in the popular NWChem quantum chemistry suite. Recently, we have expanded the Global Arrays toolkit to support common sparse operations, like sparse matrix-dense matrix multiplies (SpMM), sparse matrix-sparse matrix multiplication (SpGEMM) and Sampled Dense-Dense Matrix Multiplication (SDDMM). As it turns out, these operations are the bedrock of sparse Deep Learning (DL); sparse deep neural networks and Graph Neural Networks (GNNs) have gained increasing attention recently in achieving speedups on training and inference with reduced memory footprints. Unlike scientific applications in High Performance Computing (HPC), modern (distributed-memory capable) DL toolkits often rely on non-standardized and closed-source vendor software optimizations, creating challenges in software-hardware co-design at scale. Our goal is to support a variety of distributed-memory sparse matrix operations and helper functions in the newly created Sparse Global Arrays (SGA), such that it is possible to build portable and productive Machine Learning scenarios for algorithm/software and hardware codesign purposes. Contemporary data-parallel schemes for training/inference are undergoing a major overhaul since model replication limits scalability and causes resource inefficiencies. As such, we have adopted tensor parallelism in decomposing the model and inputs, to mitigate memory issues. Current implementation is built on top of MPI and uses CPUs to maximize the portability across the platforms.

Distributed computing, machine learning↗

3D Reconstruction of Biological Organization and Mineralization in Sediment Attached Biofilms During Uranium Bioremediation

Of central interest in this study were microbial communities attached to sediment particles and in associated groundwater. Are the communities similar? What do the organisms look like and how do they associated with each other? Research was conducted on samples collected from the Department of Energy's (DOE) Rifle Integrated Field Research Challenge (IFRC) site in Rifle, Colorado, USA. This site served first as a test case for in situ bioremediation via biostimulation, and more recently as a location for studying the role of microbial communities in the carbon and other linked biogeochemical cycles. We addressed the question of the nature of planktonic to sediment-attached microbial communities using field and laboratory experimental studies with a range of methods that provide 2- and 3-dimensional topological information coupled to information about chemical speciation, organism type, and activity levels. The research leveraged data from metagenomics and proteomics analyses that obtained through parallel work at the Rifle site in the context of the IFRC. In this project we integrated characterization methods with extensive genomic sequence information and associated environmental data to better understand the processes that occur within groundwater and sediment-attached communities. Notable is the range of characterization methods, including scanning transmission x-ray microscopy (STXM), micro-EXAFS/XANES and microdiffraction and 2D and 3D cryo-electron tomographic analysis, and high-resolutino transmission electron microscope (HRTEM) analysis to characterize these natural microbial communities. A cryo-TEM work was unique because samples for electron microscopic characterization were cryo-plunged directly on site immediately after sampling. This step minimizes post-collection alterations, including cell damages and change of redox state. Among many achievements documented in publications listed at the end of this report, we highlight the following: 1) The development of a platform for routine correlative cryogenic microscopy and spectroscopy with samples prepared on-site. 2) The determination of which organisms dominate planktonic and biofilm communities in the subsurface. 3) Identification of microorganism-mineral associations and discovery of a novel mechanism that sustains activity of iron-reducing bacteria. 4) The detection of bacteria from the OP11-OD1-WWE3 (etc.) radiation and elucidation of their remarkable structural organization by cryog-TEM cryo-electron tomograhpy (cryo-ET). 5) Extensive analysis of biofilms and documentation of the association of cells and Se minerals. 6) The comparison of expressed c-type cytochromes between pure cultures of G. bemidjiensis and related field populations, provided insight into possible molecular mechanisms for U(VI) reduction in the aquifer. At least sixteen publications will result from this project (partial support), which provide both graduate student and post doctoral training.

59 BASIC BIOLOGICAL SCIENCES↗

Expanding Accurate Person Recognition to New Altitudes and Ranges: The BRIAR Dataset

Face recognition technology has advanced significantly in recent years due largely to the availability of large and increasingly complex training datasets for use in deep learning models. These datasets, however, typically comprise images scraped from news sites or social media plat-forms and, therefore, have limited utility in more advanced security, forensics, and military applications. These applications require lower resolution, longer ranges, and ele-vated viewpoints. To meet these critical needs, we collected and curated the first and second subsets of a large multi-modal biometric dataset designed for use in the research and development (R&D) of biometric recognition technolo-gies under extremely challenging conditions. Thus far, the dataset includes more than 350,000 still images and over 1,300 hours of video footage of approximately 1,000 sub-jects. To collect this data, we used Nikon DSLR cameras, a variety of commercial surveillance cameras, specialized long-rage R&D cameras, and Group 1 and Group 2 UAV platforms. The goal is to support the development of algorithms capable of accurately recognizing people at ranges up to 1,000 m and from high angles of elevation. These ad-vances will include improvements to the state of the art in face recognition and will support new research in the area of whole-body recognition using methods based on gait and anthropometry. This paper describes methods used to col-lect and curate the dataset, and the dataset's characteristics at the current stage.

Brogan, Joel↗

Anomaly Detection in Flight Operational Data Using Deep Learning

In this session, we demonstrate two recently developed deep learning models for anomaly detection in flight operational data by the Data Sciences Group at NASA Ames Research Center. The first model is Convolutional Variational Auto-Encoder (CVAE) [1], which is an unsupervised deep encoder-decoder model, designed specifically for finding anomalies in heterogeneous multivariate time series data. We will demonstrate its application to finding anomalies in streaming data from NASA’s Digital Information Platform’s Fuser source. CVAE identifies data instances that are not representative of expected nominal behavior as anomalous. Since it is an unsupervised approach, the flagged anomalies will need to be reviewed by the subject matter experts (SMEs) for validation and labeling and is designed to assist with vulnerability discovery within Safety Monitoring System programs. The second model is Robust and Explainable Semi-supervised Anomaly Detection (RESAD) model [2], which builds on CVAE to allow learning from both minimally labeled data (previously reviewed by the SMEs) as well as majority unlabeled data. RESAD takes advantage of graph theoretic techniques to propagate the labels from the labeled data to the unlabeled data based on a pre-defined similarity metric and structures the learned feature space from flight time-series so that data of the same class would cluster tightly together. This model characteristic is enabled by training with an augmented loss function and allows learning of a more informative feature space for down-stream tasks such as search and active learning. We demonstrate RESAD using data from the NASA DASHlink project [3].

anomaly detection↗