SEARCH · Engineering Papers
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.
The CRADA objective is to 1) Generalize web tools for data analytics and machine learning, and 2) Develop and test a smart-alarm system at Jefferson Lab.
EDX ClaiMM is a centralized data & analytical platform designed to revolutionize U.S. critical minerals and materials (CMM) activities. By providing a robust digital infrastructure, ClaiMM will accelerate the combination, leveraging, and rapid utilization of vital data, advanced tools, and cutting-edge research advancements in CMM. This adaptive digital research hub connects the CMM community to essential knowledge products and offers access to interoperable datasets, databases, models, software, and tools from the National Energy Technology’s (NETL’s) Energy Data eXchange (EDX) and other authoritative sources, serving both public and private sectors. EDX ClaiMM delivers AI-informed solutions to address fundamental knowledge gaps and fosters the innovation of new techniques for enhanced characterization and recovery of CMMs within the U.S. By leveraging cloud-hosted, scalable digital infrastructure, ClaiMM meets public–private applied energy needs. It equips the CMM community with priority digital resources that harness on-site and cloud compute capabilities, enabling big data storage, advanced processing, analytics, and visualization.
Smart manufacturing is bound to play a crucial role in reducing global energy consumption while accelerating economic development. This phenomenon is evident in advanced sensing technology developments, data analytics and machine learning/AI, automated controls, cloud computing, etc. The immediate opportunity for smart manufacturing is to improve the energy efficiency of manufacturing operations through innovations in processes and controls. The manufacturing industry still consumes about 30% of total global energy production, which is significant. This program focused on the recommendation of heterogeneous sensors to monitor the Chemical Vapor Infiltration (CVI) process parameters, states, and key performance indices and utilize cloud computing to sort and analyze the real-time data. Learning through data analytics helps guide the control parameters affecting energy usage. The target process in the program is the CVI process at the Honeywell South Bend facility, which is one of the most energy-intensive manufacturing industries.
We present the current status of a large-scale computing framework to address the need of the multidisciplinary effort to study chemical dynamics. Specifically, we are enabling scientists to process and store experimental data, run large-scale computationally expensive high-fidelity physical simulation, and analyze these results using the state-of-the-art data analytics tools, machine learning, and uncertainty quantification methods using heterogeneous computing resources, such as CPU and GPU cluster. The framework can integrate or abstract out multiple domains based on roles. In order to develop this framework, we have leveraged an existing framework coupled with in-house heterogeneous computing resources. We present the results of using this framework on a single metadata triggered workflow to accelerate an additive manufacturing use case.
AI and simulation workloads consume and generate large amounts of data that need to be searched, transformed and merged with other data. With the goal of treating data as a first-class citizen inside a traditionally compute-centric HPC environment, we explore how the use of accelerators and high-speed interconnects can speed up tasks which otherwise constitute bottlenecks in computational discovery workflows. BlazingSQL is SQL engine that runs natively on NVIDIA GPUs and supports internode communication for fast analytics on terabyte-scale tabular data sets. We show how a fast interconnect improves query performance if leveraged through the Unified Communication X (UCX) middleware. We envision that future computing platforms will integrate accelerated database query capabilities for immediate and interactive analysis of large simulation data.
Emerging modern data analytics attracts much attention in materials research and shows great potential for enabling data-driven design. Data populated from the high-throughput CALPHAD approach enables researchers to better understand underlying mechanisms and to facilitate novel hypotheses generation, but the increasing volume of data makes the analysis extremely challenging. Here in this paper, we introduce an easy-to-use, versatile, and open-source data analytics frontend, ASCENDS (Advanced data SCiENce toolkit for Non-Data Scientists), designed with the intent of accelerating data-driven materials research and development. The toolkit is also of value beyond materials science as it can analyze the correlation between input features and target values, train machine learning models, and make predictions from the trained surrogate models of any scientific dataset. Various algorithms implemented in ASCENDS allow users performing quantified correlation analyses and supervised machine learning to explore any datasets of interest without extensive computing and data science background. The detailed usage of ASCENDS is introduced with an example of experimental high-temperature alloy data.
EEMS Program activities during FY 2020 focused on analytical research and large-scale modeling and simulation to understand the impacts that new mobility technologies and services will have at the vehicle-, traveler-, and overall transportation system-level. This research included the development of a multi-fidelity, end-to-end transportation system models and tools to evaluate the complex interactions among the various actors within the mobility landscape, analysis of empirical data to characterize which solutions may provide the largest benefits, and development of new control systems and algorithms that use vehicle connectivity and automation to improve the performance and efficiency of individual vehicles as well as the overall traffic system. This document presents a brief overview of the EEMS Program and documents progress and results from projects within each of the EEMS activity areas. The Computational Modeling and Simulation key activity area summarizes work within the sub-areas of (1) the SMART (Systems and Modeling for Accelerated Research in Transportation) Mobility Lab Consortium, (2) Artificial Intelligence, High-Performance Computing, and Data Analytics, and (3) Core Simulation and Evaluation Tools. Additionally, the program’s advanced R&D projects are summarized within (4) the Connectivity and Automation Technology key activity area. Each of the individual progress reports provide a project overview and highlights of the technical results.
<span style="font-family: Calibri, sans-serif; font-size: 12pt;">Accelerated materials design should match the recent trends in the product development cycles. Materials data analytics can be used to significantly shorten development time of specialized alloys needed for next generation energy applications. However, it faces a challenge of scarce data available for training ML models. Incorporation of the domain knowledge into deep-learning graph structure via fuzzy pre-training and causal process imitation presents a viable approach to developing accurate data-driven models and reliable alloy design tools, with limited datasets. Artificial Intelligence (AI) was used in this study to incorporate such knowledge in the domain-specific computational tool, pyroMind. The tool provides not only novel design ideas but also their interpretation via physics and engineering concepts.</span>
Abstract Artificial intelligence (AI) and machine learning (ML) have been increasingly used in materials science to build predictive models and accelerate discovery. For selected properties, availability of large databases has also facilitated application of deep learning (DL) and transfer learning (TL). However, unavailability of large datasets for a majority of properties prohibits widespread application of DL/TL. We present a cross-property deep-transfer-learning framework that leverages models trained on large datasets to build models on small datasets of different properties. We test the proposed framework on 39 computational and two experimental datasets and find that the TL models with only elemental fractions as input outperform ML/DL models trained from scratch even when they are allowed to use physical attributes as input, for 27/39 (≈ 69%) computational and both the experimental datasets. We believe that the proposed framework can be widely useful to tackle the small data challenge in applying AI/ML in materials science.
In the face of a changing climate, the understanding, predictions, and projections of natural and human systems are increasingly crucial to prepare and cope with extremes and cascading hazards, determine unexpected feedbacks and potential tipping points, inform long-term adaptation strategies, and guide mitigation approaches. Increasingly complex socio-economic systems require enhanced predictive information to support advanced practices. Such new predictive challenges drive the need to fully capitalize on ambitious scientific and technological opportunities. These include the unrealized potential for very high-resolution modeling of global-to-local Earth system processes across timescales, reduction of model biases, enhanced integration of human systems and the Earth Systems, better quantification of predictability and uncertainties; expedited science-to-service pathways, and co-production of actionable information with stakeholders. Enabling technological opportunities include exascale computing, advanced data storage, novel observations and powerful data analytics, including artificial intelligence and machine learning. Looking to generate community discussions on how to accelerate progress on U.S. climate predictions and projections, representatives of Federally-funded U.S. modeling groups outline here perspectives on a six-pillar national approach grounded in climate science that builds on the strengths of the U.S. modeling community and agency goals. This calls for an unprecedented level of coordination to capitalize on transformative opportunities, augmenting and complementing current modeling center capabilities and plans to support agency missions. Tangible outcomes include projections with horizontal spatial resolutions finer than 10 km, representing extremes and associated risks in greater detail, reduced model errors, better predictability estimates, and more customized projections to support next generation climate services.
As computing systems approach the limits of traditional silicon technology, the diminishing returns in performance per watt present a significant barrier to sustaining growth in HPC. From a large-scale scientific supercomputing facility point of view, we propose a multifaceted strategy toward specialized hardware and architectures that are optimized for energy efficiency in specific applications. We also emphasize the need for integrating energy-aware practices across all levels of HPC, from system design and software development to operational policies. We discuss strategic opportunities such as the adoption of application-specific accelerators, the development of energy-efficient algorithms, and the implementation of data-driven operational analytics. Our goal is to develop a comprehensive roadmap ensuring that future leadership systems at OLCF can meet scientific demands while operating within stringent energy budgets, thereby supporting sustainable computing growth.
We present the timing of the first five millisecond pulsars discovered in the globular cluster Omega Centauri and the discovery of a pulsar with a spin period of 3.68 ms. With a timing baseline of ~3.5 yr we are able to measure the derivative of the spin frequency ($\dot{\nu }$) for the first five pulsars. Upper limits on the pulsar line-of-sight acceleration are estimated and compared with predictions based on analytical models of the cluster. We find that PSRs J1326-4728B and D show large negative accelerations, which are in tension with the minimum acceleration predicted by analytical models. We searched for pulsed γ-ray signals using 14.3 yr of data from the Fermi Large Area Telescope. Although we found no evidence for γ-ray pulsations, PSRs J1326-4728A, B, C, and E are associated with X-ray sources. This suggests that the observed γ-ray emission from Omega Centauri is likely caused by the emission of the ensemble of MSPs. Finally, the linearly polarized emission from PSR J1326-4728A yields a rotation measure of -18 ± 8 rad m -2 .
Abstract Materials acceleration platforms (MAPs) operate on the paradigm of integrating combinatorial synthesis, high‐throughput characterization, automatic analysis, and machine learning. Within a MAP, one or multiple autonomous feedback loops may aim to optimize materials for certain functional properties or to generate new insights. The scope of a given experiment campaign is defined by the range of experiment and analysis actions that are integrated into the experiment framework. Herein, the authors present a method for integrating many actions within a hierarchical experimental laboratory automation and orchestration (HELAO) framework. They demonstrate the capability of orchestrating distributed research instruments that can incorporate data from experiments, simulations, and databases. HELAO interfaces laboratory hardware and software distributed across several computers and operating systems for executing experiments, data analysis, provenance tracking, and autonomous planning. Parallelization is an effective approach for accelerating knowledge generation provided that multiple instruments can be effectively coordinated, which the authors demonstrate with parallel electrochemistry experiments orchestrated by HELAO. Efficient implementation of autonomous research strategies requires device sharing, asynchronous multithreading, and full integration of data management in experimental orchestration, which to the best of the authors’ knowledge, is demonstrated for the first time herein.
Computer vision algorithms are increasingly leveraged to accelerate geospatial analysis for disaster response and recovery. As the diversity of remote sensing imagery grows with optical, SAR, and other modalities, a perquisite for analytics is cross-modal image registration. There is a high potential to harness computer vision for this pre-processing requirement toward enabling downstream analytics such as heterogeneous change detection, automated feature extraction, and data fusion. Advancement in these areas has the potential to simplify data wrangling tasks and further accelerate disaster response timelines. The SpaceNet 9 challenge (launching in mid-2024) focuses on addressing the cross-modal image registration problem and demonstrating the utility of such modules on earthquake impacted scenarios. This paper describes the motivation for the SpaceNet 9 and provides a first overview of the dataset, the baseline algorithm, and implications for seeking cross-modal image registration in Earth observation. Code is available at https://github.com/SpaceNetChallenge/SpaceNet9.
This LANSCE Facility Longevity Assessment examined 34 total systems critical to the operation of the LANSCE accelerator complex. These systems are primarily owned, operated, and maintained by AOT Division. Based on the criteria used in this assessment, there are 12 AOT Division systems that are potential high risk as related to longterm LANSCE operations. These systems may also impact the success of LAMP. Also summarized in this report are the facility activities planned over the next decade. These include general facility maintenance and infrastructure upgrades, as well as planned electrical system upgrades. The LANSCE Facility Operations (LFO) Division is responsible for this activity scope and budget planning. Most of the planned activities are general in nature and should have little impact on accelerator operations or LAMP activities as currently envisioned. However, close coordination with between AOT, LFO, and LAMP would be prudent. A next step should be to develop an actionable plan to address the high-risk systems. Potential criteria for developing such a plan could include: (1) Evaluating the global impact of the system (operations, program, safety, etc.), (2) Ranking by impact (ranked now by risk), (3) Estimating cost to implement the corrective actions (Is there already a plan or not?), and (4) Estimating the time required for corrective actions. An alternative could include integration of the assessment data into both the Asset Management System and the Maintenance Analytics Risk Register.
Today’s data-driven analytics and machine learning workload have been largely driven by the General-PurposeGraphics Processing Units (GPGPUs). To accelerate dense matrix multiplications on the GPUs, Tensor Core Units (TCUs) have been introduced in recent years. In this paper, we study linear-algebra-based and vertex-centric algorithms for various graph kernels on the GPUs with an objective of applying this new hardware feature to graph applications. We identify the potential stages in these graph kernels that can be executed on the Tensor Core Units. In particular, we leverage the reformulation of the reduction and scan operations in terms of matrix multiplication [1]on the TCUs. We demonstrate that executing these operations on the TCUs, available inside different graph kernels, can assist in establishing an end-to-end pipeline on the GPGPUs without depending on hand-tuned external libraries and still can deliver comparable performance for various graph analytics.
Batch-to-batch variation in powder compositions for binder jet additive manufacturing (BJAM) can significantly deter defining an “ideal” sintering window for a given alloy. One way to overcome the problem is by running sintering experiments at various temperatures for each batch of the powder. However, such an approach increases the time required to achieve large-scale production of parts. The predictive capabilities of computational thermodynamic tools like CALPHAD can be leveraged to overcome the challenge, especially for binder jet additive manufacturing, since the process occurs under near-equilibrium conditions. However, calculating the sintering window using CALPHAD can be computationally expensive, considering many possible feedstock compositions within “specification”. Here, we generate high throughput CALPHAD data for nickel-based superalloys to develop machine learning models to predict the sintering window rapidly. The predictive capability of the models has been validated using published results on BJAM of Inconel 718 and 625. Further, validated models are lightweight and can be deployed in an industrial setting to get sintering window in an accelerated manner.
In the rapidly evolving landscape of scientific computing, Graphics Processing Units (GPUs) have become indispensable for their unparalleled ability to handle parallel tasks in complex calculations, simulations, and data analysis. Their utility is further magnified in machine learning and AI applications, where they significantly accelerate model training and predictive analytics. Within this context, the Triton Inference Server emerges as a pivotal open-source tool, specializing in AI inferencing and optimizing GPU utilization across various platforms and frameworks. This paper presents an in-depth study on distributed High Throughput Computing (HTC), specifically focusing on the HTCondor framework and its resource provisioning tools, GlideinWMS and HEPCloud. These systems enable large-scale scientific experiments like CMS and DUNE to efficiently access and utilize vast computational resources. The paper explores the core architectural components of GlideinWMS, including jobs, user pools, and worker nodes, and discusses their integration with GPUs and the Triton server. The primary aim of this research is to develop a solution that optimizes GPU utilization by leveraging Glideins and containers. This approach allows computational jobs, particularly those involving AI models, to use GPUs only when essential, thereby facilitating efficient sharing of limited GPU resources. To validate this architecture, the study conducted three key tests involving custom scripts, container-based servers, and Triton server deployments. However, the study faces challenges, notably in locating the Triton server and ensuring secure remote access. To address these issues, future work will focus on developing a proxy mechanism and enhancing security protocols. In conclusion, this study offers a comprehensive roadmap for effective and efficient GPU utilization in distributed High Throughput Computing. It aims to contribute significantly to the scientific community by solving pressing problems and implementing robust solutions in collaboration with the GlideinWMS and HEPCloud teams. The research sets the stage for a more efficient, scalable, and cost-effective paradigm in scientific computing.