Search NASA⌕ Search

SEARCH · Search NASA

Results for “KNOWLEDGE PRESERVATION”

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 91 records · Page 5

Physics-based hybrid machine learning for critical heat flux prediction with uncertainty quantification

Critical heat flux (CHF) is a key quantity in nuclear system modeling due to its impact on heat transfer, safety margins, and reactor performance. This study develops and validates an uncertainty-aware hybrid modeling approach that combines machine learning with physics-based models to predict CHF in cases of dryout. The Biasi and Bowring empirical correlations were paired with three ML uncertainty quantification (UQ) techniques: deep neural network (DNN) ensembles, Bayesian neural networks (BNNs), and deep Gaussian processes (DGPs). A pure ML model without a base model was evaluated for comparison. Model performance was assessed under plentiful (7,350 points) and limited (9 points) training data scenarios using parity, uncertainty distributions, and calibration curves. Results show that the Biasi hybrid DNN ensemble achieved the best overall performance, with a mean absolute relative error of 1.846%, and well-calibrated uncertainty estimates. The BNN-based hybrids showed slightly higher error (2.14%) but superior uncertainty calibration. DGP models underperformed, with over 6% error and poor uncertainty calibration. All hybrid models outperformed pure machine learning configurations, demonstrating resistance against data scarcity. These findings indicate that hybrid modeling significantly improves predictive accuracy, interpretability, and resilience to data scarcity. The integration of uncertainty awareness provides actionable confidence in CHF predictions, which is vital for safety-critical decisions in nuclear applications. This hybrid approach offers a viable pathway for deploying ML models in reactor analysis tools while preserving domain knowledge and physical consistency.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Density Functional Tight-Binding Models for Band Structures of Transition-Metal Alloys and Surfaces across the d -Block

First-principles electronic structure simulations are an invaluable tool for understanding chemical bonding and reactions. While machine-learning models such as interatomic potentials significantly accelerate the exploration of potential energy surfaces, electronic structure information is generally lost. Particularly in the field of heterogeneous catalysis, simulated electron band structures provide fundamental insights into catalytic reactivity. This ab initio knowledge is preserved in semiempirical methods such as density functional tight binding (DFTB), which extend the accessible computational length and time scales beyond first-principles approaches. In this paper here we present Shell-Optimized Atomic Confinement (SOAC) DFTB electronic-part-only parametrizations for bulk and surface band structures of all d-block transition metals that enable efficient predictions of electronic descriptors for large structures or high-throughput studies on complex systems outside the computational reach of density functional theory.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Single-Net Continual Learning with Progressive Segmented Training

There is an increasing need of continual learning in dynamic systems, such as the self-driving vehicle, the surveillance drone, and the robotic system. Such a system requires learning from the data stream, training the model to preserve previous information and adapt to a new task, and generating a single-headed vector for future inference. Different from previous approaches with dynamic structures, this work focuses on a single network and model segmentation to prevent catastrophic forgetting. Leveraging the redundant capacity of a single network, model parameters for each task are separated into two groups: one important group which is frozen to preserve current knowledge, and secondary group to be saved (not pruned) for a future learning. A fixed-size memory containing a small amount of previously seen data is further adopted to assist the training. Without additional regularization, the simple yet effective approach of Progressive Segmented Training (PST) successfully incorporates multiple tasks and achieves the state-of-the-art accuracy in the single-head evaluation on CIFAR-10 and CIFAR-100 datasets. Moreover, the segmented training significantly improves computation efficiency in continual learning at the edge.

Du, Xiaocong↗

Embedded Hardware Solution for Cybersecurity in Industrial Control Systems

Industrial Control Systems are the backbone of manufacturing, power, and critical infrastructure. These systems are made up of electronic and mechanical devices that operate on complex feedback systems that allow the production of consumer products, electricity, clean water, and nuclear industries. This broad reach of cybersecurity has earned it a place in the strategic roadmaps for many government agencies, including the Department of Energy. Their efforts stem from two general initiatives which are to improve cybersecurity in the energy sector and to strengthen effectiveness of Department of Energy incident management capabilities. Particular emphasis is in improving learning and preservations of knowledge at the labs, and to encourage collaboration between the labs. This project aims to develop Savannah River National Laboratory’s knowledge of industrial control systems cybersecurity by providing our engineers the opportunity to learn and execute advanced concepts in cybersecurity to real problems in industrial control systems. Successful execution of this project will enhance Savannah River National Laboratory’s knowledge and capability in the realm of Industrial Control System and Operational Technology cybersecurity as well as advance understanding of defensive needs in these environments. This will directly impact Savannah River National Laboratory’s capability to contribute and collaborate with Department of Energy, other Labs, and commercial partners in meeting Department of Energy national industrial control systems/Operations Technology cybersecurity goals.

97 MATHEMATICS AND COMPUTING↗

gridPULSE: Catalog of National Laboratory Testing Resources for Grid Devices

This catalog was created through the Grid Modernization Laboratory Consortium's (GMLC) Testing Network project. The GMLC is a strategic partnership between the Department of Energy (DOE) and its National Laboratories to bring together leading experts, technologies, and resources to collaborate on the goal of modernizing the nation's electric grid. The benefits of the GMLC include more efficient use of resources; shared networks; improving learning and preservation of knowledge; enhanced lab coordination and collaboration; and regional perspective and relationships with local stakeholders and industry.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Synchronic Web Digital Identity: Speculations on the Art of the Possible

As search, social media, and artificial intelligence continue to reshape collective knowledge, the preservation of trust on the public infosphere has become a defining challenge of our time. Given the breadth and versatility of adversarial threats, the best—and perhaps only—defense is an equally broad and versatile infrastructure for digital identity. This document discusses the opportunities and implications of building such an infrastructure from the perspective of a national laboratory. The technical foundation for this discussion is the emergence of the Synchronic Web, a Sandia-developed infrastructure for asserting cryptographic provenance at Internet scale. As of the writing of this document, there is ongoing work to develop the underlying technology and apply it to multiple mission-specific domains within Sandia. The primary objective of this document to extend the body of existing work toward the more public-facing domain of digital identity. Our approach depends on a non-standard, but philosophically defensible notion of identity: digital identity is an unbroken sequence of states in a well-defined digital space. From this foundation, we abstractly describe the infrastructural foundations and applied configurations that we expect to underpin future notions of digital identity.

97 MATHEMATICS AND COMPUTING↗

Archean sedimentary systems and crustal evolution

Current knowledge of preserved Archean sedimentary rocks suggests that they accumulated in at least three major depositional settings. These are represented generally by sedimentary units: (1) in early Archean, pre-3.0 Ga old greenstone belts, (2) on late Archean sialic cratons, and (3) in late Archean, post-3.0 Ga old greenstone belts. Research suggests that the Archean was characterized by at least two distinctive and largely diachronous styles of crustal evolution. Thick, stable early Archean simatic platforms, perhaps analogous to modern oceanic islands formed over hot spots, underwent a single cycle of cratonization to form stable continental blocks in the early Archean. Later formed Archean continents show a two stage evolution. The initial stage is reflected in the existence of older sialic material, perhaps representing incompletely cratonized areas or microcontinents of as yet unknown origin. During the second stage, late Archean greenstone belts, perhaps analogous to modern magmatic arcs or back arc basins, developed upon or adjacent to these older sialic blocks. The formation of this generation of Archean continents was largely complete by the end of the Archean. These results suggest that Archean greenstone belts may represent a considerable range of sedimentological and tectonic settings.

Lowe, D. R.↗

MSFC Propulsion Systems Department Knowledge Management Project

This slide presentation reviews the Knowledge Management (KM) project of the Propulsion Systems Department at Marshall Space Flight Center. KM is needed to support knowledge capture, preservation and to support an information sharing culture. The presentation includes the strategic plan for the KM initiative, the system requirements, the technology description, the User Interface and custom features, and a search demonstration.

Caraccioli, Paul A.↗

Shielding Integral Benchmark Archive and Database (SINBAD)

The Shielding Integral Benchmark Archive and Database (SINBAD) collection of benchmarks was initiated in the early 1990 s. SINBAD is an international collaboration between the Organization for Economic Cooperation and Development s Nuclear Energy Agency Data Bank (OECD/NEADB) and the Radiation Safety Information Computational Center (RSICC) at Oak Ridge National Laboratory (ORNL). SINBAD is a major attempt to compile experiments and corresponding computational models with the goal of preserving institutional knowledge and expertise that need to be handed down to future scientists. SINBAD is also a learning tool for university students and scientists who need to design experiments or gain expertise in modeling and simulation. The SINBAD database is currently divided into three categories fission, fusion, and accelerator benchmarks. Where possible, each experiment is described and analyzed using deterministic or probabilistic (Monte Carlo) radiation transport software.

43 PARTICLE ACCELERATORS↗

An Unattended Verification Station for UF6 Cylinders: Implementation Concepts and Development Status

In recent years, the International Atomic Energy Agency (IAEA) has pursued innovative techniques and an integrated suite of safeguards measures to address the verification challenges posed by advanced centrifuge technologies and the growth in separative work unit capacity at modern centrifuge enrichment plants. These measures would include permanently installed, unattended instruments capable of performing the routine and repetitive measurements previously performed by inspectors. Among the unattended instruments currently being explored by the IAEA is an Unattended Cylinder Verification Stations (UCVS) that could provide independent verification of the declared relative enrichment, 235U mass and total uranium mass of 100% of the declared cylinders moving through the plant, as well as the application and verification of an ‘Non-destructive Assay Fingerprint’ to preserve verification knowledge on the contents of each cylinder throughout its life in the facility. As the IAEA’s vision for a UCVS has evolved, Pacific Northwest National Laboratory (PNNL) and Los Alamos National Laboratory (LANL) have been developing and testing candidate nondestructive assay (NDA) methods for inclusion in a UCVS. Modeling and multiple field campaigns have indicated that these methods are capable of assaying relative cylinder enrichment with a precision comparable to or perhaps better than today’s high-resolution handheld devices, without the need for manual wall-thickness corrections. In addition, the methods interrogate the full volume of the cylinder, thereby offering the IAEA a new capability to assay the absolute 235U mass in the cylinder and much-improved sensitivity to substituted or removed material. Building on this prior work a UCVS field prototype is being developed and tested. This paper describes potential implementation concepts, a preliminary UCVS prototype design and functional description, and preparations for upcoming field trials.

Smith, Leon E.↗

Shielding integral benchmark archive and database (SINBAD)

The shielding integral benchmark archive and database (SINBAD) collection of experiments descriptions was initiated in the early 1990s. SINBAD is an international collaboration between the Organization for Economic Cooperation and Development's Nuclear Energy Agency Data Bank (OECD/NEADB) and the Radiation Safety Information Computational Center (RSICC) at Oak Ridge National Laboratory (ORNL). SINBAD was designed to compile experiments and corresponding computational models with the goal of preserving institutional knowledge and expertise that need to be handed down to future scientists. SINBAD can serve as a learning tool for university students and scientists who need to design experiments or gain expertise in modeling and simulation. The SINBAD database is currently divided into three categories - fission, fusion, and accelerator experiments. Many experiments are described and analyzed using deterministic or stochastic (Monte Carlo) radiation transport software. The nuclear cross sections also play an important role as they are necessary in performing computational analysis. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Millennial Time Capsules as a Promising Means for Preserving Records for Future Generations - 16542

When we deal with preservation of Records, Knowledge and Memory (RK and M) for any long-term, non-inspected facility such a deep geological repository, there is no single technical or cultural provision that can be relied upon to do alone the preservation job 100%. Rather we should increase our chances by implementing a combination of approaches based on different components that provide redundancy and/or pointers to one another. Time capsules are no exception. However, they seem well suited to support national archives and other preservation elements, in order to preserve RK and M as long as possible or interesting. Time capsules are a ready to go, workable concept, with many examples of implementation of large-size, millennial time capsules at small depths based on science and lessons to be learned. A proposal is also made for considering and developing small-size time capsules placed strategically deep underground at repository level. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

FUSRAP Data Management and Information Transitions - 19459

The Formerly Utilized Sites Remedial Action Program (FUSRAP) was created in the mid-1970's to clean up radiological contamination resulting from early development of nuclear weapons. The US Department of Energy (US DOE) was responsible for FUSRAP until October 1997, when Congress transferred administration and execution of FUSRAP site cleanup to the US Army Corps of Engineers (USACE). Currently, the US DOE Office of Legacy Management (LM) determines the eligibility of a site for referral to USACE. USACE determines whether the site is included in FUSRAP for cleanup, conducts remedial actions, and transfers sites back to LM for long-term surveillance and maintenance (LTS and M). LM and USACE established the Data Management Working Group (DMWG) to (1) establish a common understanding of US DOE and USACE district-specific data, records practices, and requirements as they apply to FUSRAP programmatic documents for site transition and transfer; (2) establish methods to ensure accurate data transfer; and (3) develop a timeline for data and records transfer. These methods will ensure continuity of stewardship while transitioning from cleanup to LTS and M and beyond as more FUSRAP sites transfer from USACE to LM. The DMWG has met monthly and the meetings are attended by a wide range of LM and USACE project managers and subject matter experts. Discussion topics have addressed site records management, including data needs and uses; transfer and posting of Administrative Record and Permanent Record files for public accessibility; and management of legal and classified files. There have been a number of group successes, including the following: - Establishment of a common understanding of data and records practices and requirements as the sites transition and transfer from USACE to LM; - Development of effective lines of communication between USACE and LM concerning data management for transition; - Establishment of a cohesive approach to data management between the two agencies. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

An international perspective on nuclear criticality safety

Full text of publication follows: As one of the standing technical committees within the OECD Nuclear Energy Agency (NEA) the Nuclear Science Committee overseas a programme of work aimed at delivering the following key objectives: - the advancement of scientific knowledge needed to enhance the performance and safety of current nuclear systems; - building a solid scientific and technical basis for the development of future-generation nuclear systems; - the preservation of essential knowledge in the field of nuclear science. Criticality safety is an important aspect of all three of these objectives, and the NEA Nuclear Science Working Party on Nuclear Criticality (WPNCS) coordinates many international activities under its programme of work. An overview of the WPNCS is given in the following presentation, and many of the papers presented during the ICNC 2015 technical sessions will be giving details of the outputs from the various projects and activities which make up the WPNCS programme of work. This presentation focuses on issues associated with knowledge management, particularly on how an international framework for nuclear science and criticality safety activities can provide a very effective means of transferring knowledge across technical disciplines, and between different generations of engineers and scientists. To ensure that it is the essential knowledge that is preserved and transferred, we need to understand current trends and future requirements in criticality safety. In this respect, an international framework provides both the broad scope as well as the long-term perspective needed to make an informed judgement on which information/knowledge will be of most relevance to future practitioners of criticality safety. (author)

42 ENGINEERING↗

Institutional Memory Preservation at NASA Glenn Research Center

In this era of downsizing and deficit reduction, the preservation of institutional memory is a widespread concern for U.S. companies and governmental agencies. The National Aeronautical and Space Administration faces the pending retirement of many of the agency's long-term, senior engineers. NASA has a marvelous long-term history of success, but the agency faces a recurring problem caused by the loss of these engineers' unique knowledge and perspectives on NASA's role in aeronautics and space exploration. The current work describes a knowledge elicitation effort aimed at demonstrating the feasibility of preserving the more personal, heuristic knowledge accumulated over the years by NASA engineers, as contrasted with the "textbook" knowledge of launch vehicles. Work on this project was performed at NASA Glenn Research Center and elsewhere, and focused on launch vehicle systems integration. The initial effort was directed toward an historic view of the Centaur upper stage which is powered by two RL-10 engines. Various experts were consulted, employing a variety of knowledge elicitation techniques, regarding the Centaur and RL-10. Their knowledge is represented in searchable Web-based multimedia presentations. This paper discusses the various approaches to knowledge elicitation and knowledge representation employed, and assesses successes and challenges in trying to perform large-scale knowledge preservation of institutional memory. It is anticipated that strategies for knowledge elicitation and representation that have been developed in this grant will be utilized to elicit knowledge in a variety of domains including the complex heuristics that underly use of simulation software packages such as that being explored in the Expert System Architecture for Rocket Engine Numerical Simulators.

Coffey, J.↗

Noise-Based Selection of Robust Inherited Model for Accurate Continual Learning

There is a growing demand for an intelligent system to continually learn knowledge from a data stream. Continual learning requires both the preservation of previous knowledge (i.e., avoiding catastrophic forgetting) and the acquisition of new knowledge. Different from previous works that focus only on model adaptation (e.g., regularization, network expansion, memory rehearsal, etc.), we propose a novel training scheme named acquisitive learning (AL), which emphasizes both the knowledge inheritance and knowledge acquisition. AL starts from an elaborately selected model with pre-trained knowledge (the inherited model) and then adapts it to new data using segmented training. The selection is achieved by injecting random noise to various inherited models for better model robustness, which promises higher accuracy in further knowledge acquisition. The approach is validated by the visualization of the loss landscape and quantitative roughness measurement. The combination of the selective inherited model and knowledge acquisition reduces catastrophic forgetting by 10X on the CIFAR-100 dataset.

Du, Xiaocong↗

Results of the OECD-NEA Initiative 'Preservation of RK and M Across Generations' - 18084

The OECD Nuclear Energy Agency (NEA) has co-ordinated an international initiative to develop understanding and guidelines related to the long-term preservation of records, knowledge and memory (RK and M) for radioactive waste repositories. The initiative will come to a close in April 2018. Currently, a number of deliverables are in the process of publication and a preview and overview is presented in this paper. Central is the 'Final Report of the RK and M Initiative', which sets out the RK and M fundamentals, including basic principles of RK and M preservation, the concept of oversight and the relevant time frames, the mediated and non-mediated transmission of messages, and ultimately the need for a systemic strategy. Furthermore, it presents a menu of approaches and mechanisms of RK and M preservation, such as markers, time capsules, archives, the Key Information File (KIF), education, art, land use control, and monitoring. Finally, the main report provides suggestions on how to develop a systemic strategy. A bibliography and a regulatory catalogue on RK and M preservation, and concept reports describing the KIF and the Set of Essential Records (SER), will complement the set of deliverables of the RK and M Initiative. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗