Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “shared mooring”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
NASA NTRS1987

Dome craters on Ganymede

Voyager observations reveal impact craters on Ganymede that are characterized by the presence of broad, high albedo, topographic domes situated within a central pit. Fifty-seven craters with central domes were identified in images covering approx. 50% of the surface. Owing to limitations in resolution, and viewing and illumination angles, the features identified are most likely a subset of dome craters. The sample appears to be sufficiently large to infer statistically meaningful trends. Dome craters appear to fall into two distinct populations on plots of the ratio of dome diameter to crater rim diameter, large-dome craters and small-dome craters. The two classes are morphologically distinct from one another. In general, large dome craters show little relief and their constituent landforms appear subdued with respect to fresh craters. The physical attributes of small-dome craters are more sharply defined, a characteristic they share with young impact craters of comparable size observed elsewhere in the solar system. Both types of dome craters exhibit central pits in which the dome is located. As it is difficult to produce domes by impact and/or erosional processes, an endogenic origin for the domes is reasonably inferred. Several hypotheses for their origin are proposed. These hypotheses are briefly reviewed.

Moore, J. M.↗
NASA NTRS2004

Evidence for Possible Exposed Water Ice Deposits in Martian Low Latitude Chasms and Chaos

A light-toned interior layer deposit (ILD) on the floor of the deep martian depression Juventae Chasma is found to have a relatively high thermal inertia approx. 500 J m(exp -2) s(exp -1/2) K(exp -1). This could imply rock, but is also similar to the average value of thermal inertia found for north polar layered deposits. Furthermore, ILD-B is found to exhibit a bluff and terrace structure . A terrace structure arises naturally in model simulations of the sublimation of large ice deposits. Such a staircase terrain, of course, is a further characteristic of north polar layered terrain. Morphological similarity, thermal inertia in the range of thermal inertias of the north polar cap layered terrain, and relatively high albedo lead us to propose that the ILD-B may consist of residual water ice partially covered by, and perhaps mixed with, varying amounts of dust or sand. Other ILDs (A-C) are also found in Juventae Chasma. While these ILDs lack the close morphological resemblance to the north polar cap, they share many other common features and appear to be part of the same formation. Similar ILDs are found in chaotic terrain elsewhere in the martian tropics. This leads us to propose that water ice may exist in the martian tropics today and may be implicit in the formation of chaotic terrain.

Leovy, C.↗
NASA NTRSDate not supplied

Buto Facula, Ganymede: Palimpsest Exemplar

The Galilean satellites with icy surfaces (Europa, Ganymede, Callisto) are host, among other things, to a variety of large impact features that are, if not unique to these bodies, rarely encountered on planetary and satellite surfaces in the Solar System. These features include impact basins with central pits, domed floors, and so-called "pene-palimpsests" and "palimpsests". The authors have conducted a multidisciplinary approach to study how these features formed. It is likely that the particular combination of geophysical factors and impactor characteristics that is shared by these satellites is responsible for these features. One component of this study has been to derive the topography of a number of these features, and then use these Digital Terrain Models (DTM) as a fundamental tool in producing facies maps of a range of these features for which adequate data exist. Here we present the analysis of Buto Facula, Ganymede.

Jeffrey Morgan Moore↗
NASA NTRS2006

Full-Duplex Digital Communication on a Single Laser Beam

A proposed free-space optical communication system would operate in a full-duplex mode, using a single constant-power laser beam for transmission and reception of binary signals at both ends of the free-space optical path. The system was conceived for two-way data communication between a ground station and a spacecraft in a low orbit around the Earth. It has been estimated that in this application, a data rate of 10 kb/s could be achieved at a ground-station-to-spacecraft distance of 320 km, using a laser power of only 100 mW. The basic system concept is also applicable to terrestrial free-space optical communications. The system (see figure) would include a diode laser at one end of the link (originally, the ground station) and a liquid-crystal- based retroreflecting modulator at the other end of the link (originally, the spacecraft). At the laser end, the beam to be transmitted would be made to pass through a quarter-wave plate, which would convert its linear polarization to right circular polarization. For transmission of data from the laser end to the retroreflector end, the laser beam would be modulated with subcarrier phase-shift keying (SC-PSK). The transmitted beam would then pass through an aperture- sharing element (ASE) - basically, a mirror with a hole in it, used to separate the paths of the transmitted and received light beams. The transmitted beam would continue outward through a telescope (which, in the original application, would be equipped with a spacecraft-tracking system) that would launch the transmitted beam along the free-space optical path to the retroreflector end.

Hazzard, D. A.↗
NASA NTRS2004

Data Grid Management Systems

The "Grid" is an emerging infrastructure for coordinating access across autonomous organizations to distributed, heterogeneous computation and data resources. Data grids are being built around the world as the next generation data handling systems for sharing, publishing, and preserving data residing on storage systems located in multiple administrative domains. A data grid provides logical namespaces for users, digital entities and storage resources to create persistent identifiers for controlling access, enabling discovery, and managing wide area latencies. This paper introduces data grids and describes data grid use cases. The relevance of data grids to digital libraries and persistent archives is demonstrated, and research issues in data grids and grid dataflow management systems are discussed.

Moore, Reagan W.↗
NASA NTRS1993

Workshop on the Martian Northern Plains: Sedimentological, periglacial, and paleoclimatic evolution

The penultimate meeting in the Mars Surface and Atmosphere Through Time (MSATT) series of workshops was held on the campus of the University of Alaska in Fairbanks, Alaska, 12-13 Aug. 1993. This meeting, entitled 'The Martian Northern Plains: Sedimentological, Periglacial, and Paleoclimatic Evolution,' hosted by the Geophysical Institute at the University of Alaska, was designed to help foster an exchange of ideas among researchers of the Mars science community and the terrestrial glacial and periglacial science community. The technical sessions of the workshop were complemented by field trips to the Alaska Range and to the Fairbanks area and a low-altitude chartered overflight to the Arctic Costal Plain, so that, including these trips, the meeting lasted from 9-14 Aug. 1993. The meeting, field trips, and overflight were organized and partially funded by the Lunar and Planetary Institute and the MSATT Study Group. The major share of logistical support was provided by the Publications and Program Services Department of the Lunar and Planetary Institute. The workshop site was selected to allow easy access to field exposures of active glaciers and glacial and periglacial landforms. In all, 25 scientists attended the workshop, 24 scientists (plus 4 guests and the meeting coordinator) participated in the field trips, and 18 took part in the overflight. This meeting reaffirmed the value of expertly led geologic field trips conducted in association with topical workshops.

Kargel, J. S.↗
NASA NTRS1994

The Virtual Mission Operations Center

Spacecraft management is becoming more human intensive as spacecraft become more complex and as operations costs are growing accordingly. Several automation approaches have been proposed to lower these costs. However, most of these approaches are not flexible enough in the operations processes and levels of automation that they support. This paper presents a concept called the Virtual Mission Operations Center (VMOC) that provides highly flexible support for dynamic spacecraft management processes and automation. In a VMOC, operations personnel can be shared among missions, the operations team can change personnel and their locations, and automation can be added and removed as appropriate. The VMOC employs a form of on-demand supervisory control called management by exception to free operators from having to actively monitor their system. The VMOC extends management by exception, however, so that distributed, dynamic teams can work together. The VMOC uses work-group computing concepts and groupware tools to provide a team infrastructure, and it employs user agents to allow operators to define and control system automation.

Moore, Mike↗
DOE OSTI2025

Differential Expression of Core Metabolic Functions in Candidatus Altiarchaeum Inhabiting Distinct Subsurface Ecosystems

Candidatus Altiarchaea are widespread across aquatic subsurface ecosystems and possess a highly conserved core genome, yet adaptations of this core genome to different biotic and abiotic factors based on gene expression remain unknown. Here, we investigated the metatranscriptome of two Ca. Altiarchaeum populations that thrive in two substantially different subsurface ecosystems. In Crystal Geyser, a high-CO2 groundwater system in the USA, Ca. Altiarchaeum crystalense co-occurs with the symbiont Ca. Huberiarchaeum crystalense, while in the Muehlbacher sulfidic spring in Germany, an artesian spring high in sulfide concentration, Ca. A. hamiconexum is heavily infected with viruses. We here mapped metatranscriptome reads against their genomes to analyse the in situ expression profile of their core genomes. Out of 537 shared gene clusters, 331 were functionally annotated and 130 differed significantly in expression between the two sites. Main differences were related to genes involved in cell defence like CRISPR-Cas, virus defence, replication, transcription and energy and carbon metabolism. Our results demonstrate that altiarchaeal populations in the subsurface are likely adapted to their environment while influenced by other biological entities that tamper with their core metabolism. We consequently posit that viruses and symbiotic interactions can be major energy sinks for organisms in the deep biosphere.

archaea↗
DOE OSTI2022

Thermal properties of the metallic delafossite PdCo⁢O 2 : A combined experimental and first-principles study

Metallic delafossite materials (e.g., PdCo⁢O 2 , PtCo⁢O 2 ), have attracted much recent attention due to record-high oxide conductivities, the origins of which remain unclear. Relatively little attention has been paid to their related thermal properties, however. Here, we address this via wide temperature range experimental studies of the crystal structure, thermal expansion, and specific heat of single-crystal PdCo⁢O 2 , combined with density-functional theory (DFT) calculations of the electronic and phononic densities of states, and thus thermal properties. PdCo⁢O 2 is shown to retain the $R\bar{3}m$ space group from 12 to 1000 K, exhibiting a- and c-axis thermal expansion in good quantitative agreement with DFT-based lattice dynamics calculations. The Co-O bond lengths additionally elucidate the stability of the low-spin state of the nominally Co 3+ ions, which is a notable difference between the edge-shared Co-O octahedra in PdCo⁢O 2 and the corner-shared octahedra in Co-based perovskites. Measurements of specific heat from 1.9 to 400 K provide accurate values for the Debye temperature and Sommerfeld coefficient, the phononic part being describable via a combined Debye-Einstein approach (accounting for high-frequency oxygen-related optical phonons), with excess intermediate-temperature specific heat due to a prominent low-energy peak in the phonon density of states. Most significantly, all electronic and phononic contributions to the specific heat are shown to be remarkably closely reproduced by DFT-based calculations, establishing quantitative understanding of key thermal properties of the metallic delafossite PdCo⁢O 2 .

36 MATERIALS SCIENCE↗
NASA NTRS2013

High Speed Mobility Through On-Demand Aviation

Game changing advances come about by the introduction of new technologies at a time when societal needs create the opportunity for new market solutions. A unique opportunity exists for NASA to bring about such a mobility revolution in General Aviation, extendable to other aviation markets, to maintain leadership in aviation by the United States. This report outlines the research carried out so far under NASA's leadership towards developing a new mobility choice, called Zip Aviation1,2,3. The feasibility, technology and system gaps that need to be addressed, and pathways for successful implementation have been investigated to guide future investment. The past decade indicates exciting trends in transportation technologies, which are quickly evolving. Automobiles are embracing automation to ease driver tasks as well as to completely control the vehicle with added safety (Figure 1). Electric propulsion is providing zero tail-pipe emission vehicles with dramatically lower energy and maintenance costs. These technologies have not yet been applied to aviation, yet offer compelling potential benefits across all aviation markets, and in particular to General Aviation (GA) as an early adopter market. The benefits of such an adoption are applicable in the following areas: 􀁸 Safety: The GA market experiences accident rates that are substantially higher than automobiles or commercial airlines, with 7.5 fatal accidents per 100 million vehicle miles compared to 1.3 for automobiles and.068 for airlines. Approximately 80% of these accidents are caused by some form of pilot error, with another 13% caused by single point propulsion system failure. 􀁸 Emissions: Environmental constraints are pushing for the elimination of 100Low Lead (LL) fuel used in most GA aircraft, with aviation fuel the #1 source of lead emissions into the environment. Aircraft also have no emission control systems (i.e. no catalytic converters etc.), so they are gross hydrocarbon polluters compared to automobiles. 􀁸 Community Noise: Hub and smaller GA airports are facing increasing noise restrictions, and while commercial airliners have dramatically decreased their community noise footprint over the past 30 years, GA aircraft noise has essentially remained same, and moreover, is located in closer proximity to neighborhoods and businesses. 􀁸 Operating Costs: GA operating costs have risen dramatically due to average fuel costs of over $6 per gallon, which has constrained the market over the past decade and resulted in more than 50% lower sales and 35% less yearly operations. Infusion of autonomy and electric propulsion technologies can accomplish not only a transformation of the GA market, but also provide a technology enablement bridge for both larger aircraft and the emerging civil Unmanned Aerial Systems (UAS) markets. The NASA Advanced General Aviation Transport Experiments (AGATE) project successfully used a similar approach to enable the introduction of primary composite structures and flat panel displays in the 1990s, establishing both the technology and certification standardization to permit quick adoption through partnerships with industry, academia, and the Federal Aviation Administration (FAA). Regional and airliner markets are experiencing constant pressure to achieve decreasing levels of community emissions and noise, while lowering operating costs and improving safety. But to what degree can these new technology frontiers impact aircraft safety, the environment, operations, cost, and performance? Are the benefits transformational enough to fundamentally alter aircraft competiveness and productivity to permit much greater aviation use for high speed and On-Demand Mobility (ODM)? These questions were asked in a Zip aviation system study named after the Zip Car, an emerging car-sharing business model. Zip Aviation investigates the potential to enable new emergent markets for aviation that offer "more flexibility than the existing transportation solutions." These studies indicate that autonomy and electric propulsion technology infusions offer a unique opportunity to provide breakthrough capabilities for new high speed, on-demand travel alternatives that can leapfrog the need for future expensive ground-based infrastructure investment. At the same time, such investments offer a method of laying the foundation for these technologies to be incubated for commercial aviation at lower cost, and with lower initial certification thresholds due to the relatively poor capabilities of GA aircraft to permit early adoption and private market capitalization by rapid technology accelerations, as depicted in Figure 2.

Moore, Mark D.↗
NASA NTRS2018

Hawkeye Ocean Color Instrument – Performance Summary

Hawkeye is an ocean color instrument that is part of the SeaHawk satellite developed for SOCON, the Sustained Ocean Color Observations using Nanosatellites program funded by the Gordon and Betty Moore Foundation and managed by the University of North Carolina – Wilmington (UNC-W). HawkEye has spectral characteristics similar to SeaWiFS, but with 8 times finer resolution and a smaller field of view more appropriate for lakes, rivers, and near-shore terrestrial environments. With a volume of only 10 X 10 X 10 cm (a CubeSat 1U), it can produce 8 bands of image data in a single pass, each with 1800 x 6000 pixels, with a resolution of 120 meters per pixel. This paper will present a short summary of instrument design, the spacecraft interface, and “lessons learned” during this effort. Scientists considering using linear arrays in a pushbroom mode for remote sensing will find this useful. Much of the discussion will center on optical performance, such as flat field calibration, polarization effects, stray light, out-of-band response, and exposure linearity. Images from field tests will be shown.The Hawkeye instrument is an ocean color measuring instrument designed to fly on the SeaHawk satellite developed for SOCON, the Sustained Ocean Color Observations using Nanosatellites program funded by the Gordon and Betty Moore Foundation and managed by the University of North Carolina – Wilmington (UNC-W). The Hawkeye instrument measures ocean color in 8 spectral bands, similar to SeaWiFS, except Band 7, which is shifted to a slightly lower wavelength to avoid the oxygen absorption feature that a wider band overlapped on SeaWiFS. The instrument is approximately 1/3rd the volume of the entire satellite, which is a 3U Cubesat manufactured by Clydespace in Glasgow, Scotland. The purpose of this instrument is to ascertain the quality of ocean color data possible with such a small, inexpensive instrument and bus. The nominal orbit is 540 km, and the nominal pixel geometric instantaneous field of view (GIFOV) 120 meters on a side. Each band will produce an image 1800 x 6000 pixels in size, for a total field of view of 216 X 720 km.2) DESIGN CONCEPTThe Hawkeye instrument uses linear arrays in pushbroom mode to collect data over a two dimensional area. The instrument has 4 linear CCD arrays, the Onsemi KLI-4104, to collect the 8 bands of data. Figure 1 illustrates the optical design for two bands, sharing a single array.

Hawkeye↗
NASA NTRS2018

High Power Advanced Solar Electric Propulsion Development Under NASA's NextSTEP Project

Human missions to Mars will require next generation solar electric propulsion (SEP) systems that can operate at power levels of 300 kW or higher to efficiently transport crew and cargo. In the mid-2020s, NASA is planning to launch the Deep Space Gateway (DSG), which is a crew-tended habitat that will be placed in orbit around the Moon. NASA is developing 12.5 kW Hall thrusters for the DSG. In the late 2020s, the DSG will be followed by a second vehicle, called the Deep Space Transport (DST), which will simulate a crewed Mars mission in cislunar space. In parallel with the 12.5 kW thrusters for the DSG, NASA is initiating development of 100 kW thrusters for potential demonstration on the DST. In 2016, NASA issued a Next Space Technologies for Exploration Partnerships (NextSTEP) Broad Agency Announcement to solicit proposals for development of high-power electric thrusters. The partnerships require at least 50 percent cost sharing by the companies. Three companies were selected for development activities lasting up to three years: Ad Astra, Aerojet Rocketdyne, and MSNW. Ad Astra is developing a Variable Specific Impulse Magnetoplasma Rocket (VASIMR), Aerojet Rocketdyne is developing a Nested Hall Thruster (NHT), and MSNW is developing an Electrodeless Lorentz Force (ELF) plasma thruster. The primary goal of these activities is to operate the thrusters at 100 kW for 100 continuous hours in a vacuum chamber. In addition to the thrusters, the companies are developing Power Processing Units (PPUs), propellant feed systems, and thermal management systems. This presentation will describe the three activities, the progress they have made in achieving the primary testing goal, the technical challenges they have encountered, and discuss the importance of extensibility.

Moore, Chris↗
NASA NTRS2008

CME-Producing Precursors to the 2006 December 13 X-Flare

We revisit one of the largest explosions observed during the Hinode era, the X4.3 class event of 2006 Dec 13. We gain insight into the main eruption through study of two sub-C-class precursor eruptions, occurring within 12 hours of and originating from the same (or nearby) neutral line as the X-flare. The precursors share some features in common with the main eruption, and their lower energy and consequent slower development renders interpretation of these features easier to decipher than in the rapidly explosive main eruption. In addition, because the weak precursors occurred in a magnetically strong region, magnetic connections indicated by soft X-ray loops are readily visible in these cases, while such connections can be much less apparent in weaker-region eruptions. Hinode/SOT magnetograms indicate that photospheric magnetic dynamic activity in the "magnetic core" is the likely ultimate source of the eruptions. All the eruptions, however, produce Coronal Mass Ejections (CMEs) that have wider spatial extent than the localized source region; this is a long-observed but puzzling phenomena, which can address directly here using the high-quality Hinode data. For the precursor eruptions, Hinode/XRT images show that the initial eruptions occur inside larger-scale magnetic structures that encompass the core. The exploding core field blows out this larger-scale structure, resulting in the CME having angular extent far exceeding that of the source-region core alone; this is the arch-arch-blowout scenario for CMEs of Moore & Sterling (2007). Similar processes occur in the main eruption, except that the much larger energy release in that eruption compared to the precursors results in much faster and larger-scale phenomena.

Sterling, Alphonse C.↗
DOE OSTI2026

Accelerating Structure–Property Relationship Discovery with Multimodal Machine Learning and Self-Driving Microscopy

Microscopy combined with local spectroscopy is widely used to correlate nanoscale structure with functional properties in materials, but conventional measurements rely heavily on human-selected sampling locations and predefined targets, limiting data set diversity and the potential for discovery. Here, we present a framework that integrates autonomous microscopy with dual-novelty deep kernel learning (DN-DKL) for adaptive data acquisition and a dual variational autoencoder (VAE) for representation learning. DN-DKL actively guides the microscopy toward structurally and spectroscopically novel regions, enabling efficient collection of large spectral data sets. Dual-VAE embeds local structures and spectroscopic responses into a shared latent manifold that serves as a structure–property relationship map. We applied this framework for the investigation of halide perovskite films by using conductive atomic force microscopy. The results reveal distinct hysteresis behaviors that are linked to specific nanoscale structural motifs, including grain boundary junction points that show hysteresis under different bias conditions and asymmetric grain boundaries that suppress the charge transport. This framework establishes a general strategy that leverages the complementary strengths of self-driving microscopy, machine learning, and human expertise to accelerate scientific discovery in functional materials.

atomic force microscopy↗
NASA NTRS2004

Main-Sequence CMEs as Magnetic Explosions: Compatibility with Observed Kinematics

We examine the kinematics of 26 CMEs of the morphological main sequence of CMEs, those having the classic three-part bubble structure of (1) a bright front eveloping (2) a dark cavity within which rides (3) a bright blob/filamentary feature. Each CME is observed in Yohkoh/SXT images to originate from near the limb (> or equal to 0.7 R(sub Sun) from disk center). The basic data (from the SOHO LASCO CME Catalog) for the kinematics of each CME are the sequence of LASCO images of the CME, the time of each image, the measured radial distance of the front edge of the CME in each image, and the measured angular extent of the CME. About half of our CMEs (12) occur with a flare, and the rest (14) occur without a flare. While the average linear-fit speed of the flare CMEs (1000 km/s) is twice that of the non-flare CMEs (510 km/s), the flare CMEs and the non-flare CMEs are similar in that some have nearly flat velocity-height (radial extent) profiles (little acceleration), some have noticeably falling velocity profiles (noticeable deceleration), and the rest have velocity profiles that rise considerably through the outer corona (blatant acceleration). This suggests that in addition to sharing similar morphology, main-sequence CMEs all have basically the same driving mechanism. The observed radial progression of each of our 26 CMEs is fit by a simple model magnetic plasmoid that is in pressure balance with the radial magnetic field in the outer corona and that propels itself outward by magnetic expansion, doing no net work on its surroundings. On average over the 26 CMEs, this model fits the observations as well as the assumption of constant acceleration. This is compatible with main-sequence CMEs being magnetically driven, basically magnetic explosions, with the velocity profile in the outer corona being largely dictated by the initial Alfien speed in the CME (when the front is at approx. 3 (sub Sun), analogous to the mass of a main-sequence star dictating the luminosity.

Moore, Ron↗
NASA NTRS2011

High Precision Thermal, Structural and Optical Analysis of an External Occulter Using a Common Model and the General Purpose Multi-Physics Analysis Tool Cielo

The efficient simulation of multidisciplinary thermo-opto-mechanical effects in precision deployable systems has for years been limited by numerical toolsets that do not necessarily share the same finite element basis, level of mesh discretization, data formats, or compute platforms. Cielo, a general purpose integrated modeling tool funded by the Jet Propulsion Laboratory and the Exoplanet Exploration Program, addresses shortcomings in the current state of the art via features that enable the use of a single, common model for thermal, structural and optical aberration analysis, producing results of greater accuracy, without the need for results interpolation or mapping. This paper will highlight some of these advances, and will demonstrate them within the context of detailed external occulter analyses, focusing on in-plane deformations of the petal edges for both steady-state and transient conditions, with subsequent optical performance metrics including intensity distributions at the pupil and image plane.

integrated modeling↗
DOE OSTI2023

Suspending democratic (dis)belief: Nonliberal energy polities of solar power in Morocco and Tanzania

This paper proposes the concept of technodemocratic imaginaries (TDIMs) to supplement the widely-used framework of sociotechnical imaginaries (STIMs) for analyzing nonliberal political contexts in energy social science research. Taking the power of coproductionist analysis of STIMs seriously, TDIMs highlight the inherent instabilities within overarching STIMs by arguing that not only are particular groups marginalized and excluded but that they continue to practice and mobilize their own imaginaries of collective governance and justice vis-à-vis energy systems. In contrast, the existing idea of contested STIMs, in which a shared mode of knowledge politics undergirds different imaginaries, casts aside subaltern modes of political engagement and knowledge-making. TDIMs expand conceptions of democracy to include shared practices of credibility in nonliberal political orders. The intent is not to promote democratic relativism but rather to ask scholars and international energy-access practitioners to suspend their democratic disbelief when studying energy matters in so-called nonliberal contexts. We develop the concept of TDIMs by comparing two African nation-states–Morocco and Tanzania–to show how the states and subaltern groups do (or do not) develop TDIMs related to solar power. While international governance organizations often portray Morocco as authoritarian and Tanzania corrupt, each state differently experienced colonization and decolonization and practices different relationships with domestic subaltern groups. Whereas low-income citizens and indigenous groups seek integration into the Moroccan state’s STIM, the Maasai in Tanzania chart their own TDIM separate from the state and international development groups.

14 SOLAR ENERGY↗
DOE OSTI2024

Decode the Workload: Training Deep Learning Models for Efficient Compute Cluster Representation

Monitoring the status of a high throughput computing cluster running computationally intensive production jobs is a crucial yet challenging system administration task due to the complexity of such systems. To this end, we train autoencoders using the Linux kernel CPU metrics of the cluster. Additionally, we explore assisting these models with graph neural networks to share information across threads within a compute node. The models are compared in terms of their ability to: 1) Produce a compressed latent representation that captures the salient features of the input, 2) Detect anomalous activity, and 3) Make distinction between different kinds of jobs run at Jefferson Lab. The goal is to have a robust encoder whose compressed embeddings are used for several downstream tasks. We extend this study further by deploying these models in a human-in-the-loop production-based setting for the anomaly detection task and discuss the associated implementation aspects such as continual learning and the criterion to generate alarms. This study represents a first step in the endeavor towards building self-supervised large-scale foundation models for computing centers.

Mohammed, Ahmed↗