Search NASA⌕ Search

SEARCH · Search NASA

Results for “fault localization”

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 469 records · Page 26
All sourcesNASA NTRSDOE OSTI
NASA NTRS2002

Fail-Safe Operation of a High-Temperature Magnetic Bearing Investigated for Gas Turbine Engine Applications

The Structural Mechanics and Dynamics Branch at the NASA Glenn Research Center has developed a three-axis high-temperature magnetic bearing suspension rig to enhance the safety of the bearing system up to 1000 F. This test rig can accommodate thrust and radial bearings up to a 22.84 cm (9 in.) diameter with a maximum axial loading of 22.25 kN (5000 lb) and a maximum radial loading up to 4.45 kN (1000 lb). The test facility was set up to test magnetic bearings under high-temperature (1100 F) and high-speed (20,000 rpm) conditions. The magnetic bearing is located at the center of gravity of the rotor between two high-temperature grease-packed mechanical ball bearings. The drive-end duplex angular contact ball bearing, which is in full contact, acts as a moment release and provides axial stability. The outboard end ball bearing has a 0.015-in. radial clearance between the rotor to act as a backup bearing and to compensate for axial thermal expansion. There is a 0.020-in. radial air gap between the stator pole and the rotor. The stator was wrapped with three 1-kW band heaters to create a localized hot section; the mechanical ball bearings were outside this section. Eight threaded rods supported the stator. These incorporated a plunger and Bellville washers to compensate for radial thermal expansion and provide rotor-to-stator alignment. The stator was instrumented with thermocouples and a current sensor for each coil. Eight air-cooled position sensors were mounted outside the hot section to monitor the rotor. Another sensor monitored this rotation of the outboard backup bearing. Ground fault circuit interrupts were incorporated into all power amplifier loops for personnel safety. All instrumentation was monitored and recorded on a LabView-based data acquisition system. Currently, this 12-pole heteropolar magnetic bearing has 13 thermal cycles and over 26 hr of operation at 1000 F.

Choi, Benjamin B.↗
NASA NTRS1995

Autonomous control system reconfiguration for spacecraft with non-redundant actuators

The Small Satellite Technology Initiative (SSTI) 'CLARK' spacecraft is required to be single-failure tolerant, i.e., no failure of any single component or subsystem shall result in complete mission loss. Fault tolerance is usually achieved by implementing redundant subsystems. Fault tolerant systems are therefore heavier and cost more to build and launch than non-redundent, non fault-tolerant spacecraft. The SSTI CLARK satellite Attitude Determination and Control System (ADACS) achieves single-fault tolerance without redundancy. The attitude determination system system uses a Kalman Filter which is inherently robust to loss of any single attitude sensor. The attitude control system uses three orthogonal reaction wheels for attitude control and three magnetic dipoles for momentum control. The nominal six-actuator control system functions by projecting the attitude correction torque onto the reaction wheels while a slower momentum management outer loop removes the excess momentum in the direction normal to the local B field. The actuators are not redundant so the nominal control law cannot be implemented in the event of a loss of a single actuator (dipole or reaction wheel). The spacecraft dynamical state (attitude, angular rate, and momentum) is controllable from any five-element subset of the six actuators. With loss of an actuator the instantaneous control authority may not span R(3) but the controllability gramian integral(limits between t,0) Phi(t, tau)B(tau )B(prime)(tau) Phi(prime)(t, tau)d tau retains full rank. Upon detection of an actuator failure the control torque is decomposed onto the remaining active axes. The attitude control torque is effected and the over-orbit momentum is controlled. The resulting control system performance approaches that of the nominal system.

Grossman, Walter↗
NASA NTRSDate not supplied

Using Satellite Soil Moisture and Rainfall in the Landslide Hazard Assessment for Situational Awareness System

The Landslide Hazard Assessment for Situational Awareness system(LHASA)gives a global view of landslide hazard in nearly real time. Currently, it is being upgraded from version 1 to version 2, which entails improvements along several dimensions. These include the incorporation of new predictors, machine learning, and new event-based landslide inventories. As a result, LHASA version 2 substantially improves on the prior performanceand introduces a probabilistic element to the global landslide nowcast. Data from the soil moisture active-passive (SMAP) satellite has been assimilated into a globally consistent data product with a latency less than 3 days, known as SMAP Level 4. In LHASA, thesedata representthe antecedent conditions prior to landslide-triggering rainfall. In some cases, soil moisture may have accumulated over aperiod of many months. The model behind SMAP Level 4 also estimates the amount of snow on the ground, which is an important factor in some landslide events. LHASA also incorporates this information as an antecedent condition that modulates the response torainfall. Slope, lithology, and active faults were also used as predictor variables. These factors can have a strong influence on where landslides initiate.LHASA relies on precipitation estimates from the Global Precipitation Measurement mission to identify the locations where landslides are most probable. The low latency and consistent global coverage of these data make them ideal for real-time applications at continental to global scales. LHASA relies primarily on rainfall from the last 24 hours to spothazardous sites, which is rescaled by the local 99thpercentile rainfall.However, the multi-day latency of SMAP requires the use of a 2-day antecedent rainfall variable to represent the accumulation of rain between the antecedent soil moisture and current rainfall. LHASA merges these predictors with XGBoost, a commonly used machine-learning tool, relying on historical landslide inventories to develop the relationship between landslide occurrence and various risk factors. The resulting model relies heavily on current daily rainfall, but other factors also play an important role. LHASA outputsthe probability oflandslide occurrence ona grid of roughly one kilometer over all continents from 60 North to 60 South latitude. Evaluation over the period 2019-2020 showsthat LHASA version 2 doubles the accuracy of the global landslide nowcast without increasing the global false alarm rate. LHASA also identifies the areas where the human exposure to landslide hazard is most intense. Landslide hazard is divided into 4 levels: minimal, low, moderate, and high. Next, the number of persons and the length of major roads (primary and secondary roads)within each of these areas is calculated for every second-level administrative district (county). These results can be viewedthrough a web portal hosted at the Goddard Space Flight Center. In addition, users can download daily hazard and exposure data.LHASAversion 2uses machine learning and satellite data to identify areas of probable landslide hazard within hours of heavy rainfall. Itsglobal maps are significantly more accurate, and it now includes rapid estimates of exposed populations and infrastructure. In addition, a forecast mode will be implemented soon.

Thomas Stanley↗
NASA NTRS2013

Damage Detection Response Characteristics of Open Circuit Resonant (SansEC) Sensors

The capability to assess the current or future state of the health of an aircraft to improve safety, availability, and reliability while reducing maintenance costs has been a continuous goal for decades. Many companies, commercial entities, and academic institutions have become interested in Integrated Vehicle Health Management (IVHM) and a growing effort of research into "smart" vehicle sensing systems has emerged. Methods to detect damage to aircraft materials and structures have historically relied on visual inspection during pre-flight or post-flight operations by flight and ground crews. More quantitative non-destructive investigations with various instruments and sensors have traditionally been performed when the aircraft is out of operational service during major scheduled maintenance. Through the use of reliable sensors coupled with data monitoring, data mining, and data analysis techniques, the health state of a vehicle can be detected in-situ. NASA Langley Research Center (LaRC) is developing a composite aircraft skin damage detection method and system based on open circuit SansEC (Sans Electric Connection) sensor technology. Composite materials are increasingly used in modern aircraft for reducing weight, improving fuel efficiency, and enhancing the overall design, performance, and manufacturability of airborne vehicles. Materials such as fiberglass reinforced composites (FRC) and carbon-fiber-reinforced polymers (CFRP) are being used to great advantage in airframes, wings, engine nacelles, turbine blades, fairings, fuselage structures, empennage structures, control surfaces and aircraft skins. SansEC sensor technology is a new technical framework for designing, powering, and interrogating sensors to detect various types of damage in composite materials. The source cause of the in-service damage (lightning strike, impact damage, material fatigue, etc.) to the aircraft composite is not relevant. The sensor will detect damage independent of the cause. Damage in composite material is generally associated with a localized change in material permittivity and/or conductivity. These changes are sensed using SansEC. The unique electrical signatures (amplitude, frequency, bandwidth, and phase) are used for damage detection and diagnosis. An operational system and method would incorporate a SansEC sensor array on select areas of the aircraft exterior surfaces to form a "Smart skin" sensing surface. In this paper a new method and system for aircraft in-situ damage detection and diagnosis is presented. Experimental test results on seeded fault damage coupons and computational modeling simulation results are presented. NASA LaRC has demonstrated with individual sensors that SansEC sensors can be effectively used for in-situ composite damage detection of delamination, voids, fractures, and rips. Keywords: Damage Detection, Composites, Integrated Vehicle Health Monitoring (IVHM), Aviation Safety, SansEC Sensors

Dudley, Kenneth L.↗
DOE OSTI2021

Impact Analysis of Data Integrity Attacks on FACTS-based Wide-Area Voltage Control System

Energy management system (EMS) consists of several wide-area control applications that serve as a backbone for security, stability, and reliability of the power system. Wide-area voltage control system (WAVCS), one of the critical wide-area applications, operates in coordination with local Flexible AC Transmission System (FACTS) devices to provide voltage security and optimal management of active and reactive power resources. Since the WAVCS relies on wide-area communication and data sharing devices, possible cybersecurity vulnerabilities have to be addressed to ensure the closed-loop operation of WAVCS. In this paper, we present a methodology for performing an impact analysis of cyber-attacks in WAVCS cybersecurity. In particular, different types of data integrity attacks, such as malicious tripping, fault replay, and signal altering attacks, are considered, and detailed impact analysis is conducted in a testbed environment using the Kundur's four machine two-area system. For performing an impact analysis, the transient voltage stability of the sensitive bus voltage is studied, followed by the quantitative assessment and severity ranking using the voltage profile index. Our experimental evaluation reveals that the data integrity attacks on control signals exhibit a higher attack severity than on the measurement signals. Further, the severity of these attacks varies with nature (static or dynamic), location, and types of attacks.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2020

Utah FORGE: Phase 3 Magnetotelluric (MT) Data

New high-quality tensor MT data at 122 sites, including the vertical magnetic field and utilizing ultra-remote referencing, have been acquired over the Utah FORGE project area. The results will be used to delineate the densities of faults and fractures in crystalline basement rocks, to define the heat sources, and to derive baseline 3D resistivity structure for later MT monitoring of temporal changes in resistivity structure following well stimulation in the EGS reservoir. There are three files here related to Utah FORGE magnetotelluric (MT) data acquisition and processing. The FORGE MT EDIs zip file contains the observed MT responses in industry-standard EDI format. For each site, there is an EDI response file that utilized a local independent reference for noise cancellation, and a file that utilized a distant reference cancelling noise associated with the DC transmission line of the Delta IPP passing down the west side of Milford Valley. These two site files could be merged as appropriate. The FORGE Model Cell Center file contains the model volume of the 3D Forge MT inversion for characterizing the resistivity structure in the project area. It was derived using finite element inversion methodology described in Wannamaker et al, in the attached FORGE Phase3 Geophysics paper, from the MT observation EDI files. It is ASCII format (.dat) and entries are defined at top of the file in a simple x-y-z-Rho listing in UTM coordinates. The element layers drape the topography so the Rho value layers are not purely horizontal slices. This greatly simplifies the listing.

15 GEOTHERMAL ENERGY↗
DOE OSTI2022

Polytypism in semi-disordered lizardite and amesite by low-dose HAADF-STEM

Serpentine minerals exert important controls on the physical properties of ultramafic rocks and have the potential to influence deformation phenomena in fault zones and to control the release of water in subducted slabs. Sheet serpentine generally, and lizardite and amesite specifically, can adopt alternative crystallographic stacking arrangements called polytypes. Polytypism has been extensively studied in fully ordered crystals, but it remains largely enigmatic in the more common semi-disordered crystals that in long-range analyses such as X-ray diffraction only exhibit random combinations of 0b and ±1/3b interlayer shifts. To date, atomic-resolution imaging to identify locally ordered polytypes has been precluded by the beam-sensitive nature of this hydrous magnesium silicate mineral. Here, in this study, we employed low-dose high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) to study the polytypic structure of semi-disordered lizardite and amesite. Because the electron dose was as low as ~6000 e–/Å 2 , it was possible to directly resolve oxygen atomic columns and all the cations with a resolution of ~1 Å and reveal the short-range order. For lizardite, we identified long-period non-standard polytypes, including examples with 3, 4, 8, and 9 layers stemming from the ordering of the octahedral tilt along the a-axis. For amesite, we found short-range ordered polytypes with periodicities of up to 42 Å stemming from the ordering of interlayer shifts along the b-axis. The resolution was sufficient to determine the relative abundance of 6R 2 , 6R 1 , 2H 1 , and 2H 2 polytypes in amesite to be 46.1, 29.6, 7.7, and 1.9%, respectively. This is contrary to the expectation that the most common form of amesite is the 2H 2 polytype, which may be more likely to form macroscopic crystals suitable for conventional X-ray diffraction-based studies. We conclude that HAADF-STEM methods open the way for the characterization of beam-sensitive minerals and to resolve the structural details of less well-ordered (but possibly more abundant) minerals at a unit-cell scale.

58 GEOSCIENCES↗
NASA NTRS1993

Plains tectonism on Venus: The deformation belts of Lavinia Planitia

High-resolution radar images from the Magellan spacecraft have revealed the first details of the morphology of the Lavinia Planitia region of Venus. A number of geologic units can be distinguished, including volcanic plains units with a range of ages. Transecting these plains over much of the Lavinia region are two types of generally orthogonal features that we interpret to be compressional wrinkle ridges and extensional grooves. The dominant tectonic features of Lavinia are broad elevated belts of intense deformation that transect the plains with complex geometry. They are many tens to a few hundred kilometers wide, as much as 1000 km long, and elevated hundreds of meters above the surrounding plains. Two classes of deformation belts are seen in the Lavinia region. 'Ridge belts' are composed of parallel ridges, each a few hundred meters in elevation, that we interpret to be folds. Typical fold spacings are 5-10 km. 'Fracture belts' are dominated instead by intense faulting, with faults in some instances paired to form narrow grabens. There is also some evidence for modest amounts of horizontal shear distributed across both ridge and fracture belts. Crosscutting relationships among the belts show there to be a range in belt ages. In western Lavinia, in particular, many ridge and fracture belts appear to bear a relationship to the much smaller wrinkle ridges and grooves on the surrounding plains: ridge morphology tends to dominate belts that lie more nearly parallel to local plains wrinkle ridges, and fracture morphology tends to dominate belts that lie more nearly parallel to local plains grooves. We use simple models to explore the formation of ridge and fracture belts. We show that convective motions in the mantle can couple to the crust to cause horizontal stresses of a magnitude sufficient to induce the formation of deformation belts like those observed in Lavinia. We also use the small-scale wavelengths of deformation observed within individual ridge belts to place an approximate lower limit on the venusian thermal gradient in the Lavinia region at the time of deformation.

Squyres, Steven W.↗
DOE OSTI2025

Federated Learning for Efficient Condition Monitoring and Anomaly Detection in Industrial Cyber-Physical Systems

Detecting and localizing anomalies in cyber-physical systems (CPS) has become increasingly challenging as systems grow in complexity, particularly due to varying sensor reliability and node failures in distributed environments. While federated learning (FL) offers a foundation for distributed model training, existing approaches lack mechanisms to handle these CPS-specific challenges. This paper presents an enhanced FL framework that introduces three key innovations: adaptive model aggregation based on sensor reliability, dynamic node selection for resource optimization, and Weibull-based checkpointing for fault tolerance. Our framework enables reliable condition monitoring while addressing the computational and reliability challenges of industrial CPS deployments. Experiments on NASA Bearing and Hydraulic System Datasets demonstrate superior performance over state-of-the-art FL methods, achieving 99.5% AUC-ROC in anomaly detection and maintaining accuracy under node failures. Statistical validation using Mann-Whitney (U) test confirms significant improvements (p < 0.05) in both detection accuracy and computational efficiency across diverse operational scenarios.1

Marfo, William [University of Texas at El Paso,Dep↗
NASA NTRS1995

Global Dynamic Modeling of Space-Geodetic Data

The proposal had outlined a year for program conversion, a year for testing and debugging, and two years for numerical experiments. We kept to that schedule. In first (partial) year, author designed a finite element for isostatic thin-shell deformation on a sphere, derived all of its algebraic and stiffness properties, and embedded it in a new finite element code which derives its basic solution strategy (and some critical subroutines) from earlier flat-Earth codes. Also designed and programmed a new fault element to represent faults along plate boundaries. Wrote a preliminary version of a spherical graphics program for the display of output. Tested this new code for accuracy on individual model plates. Made estimates of the computer-time/cost efficiency of the code for whole-earth grids, which were reasonable. Finally, converted an interactive graphical grid-designer program from Cartesian to spherical geometry to permit the beginning of serious modeling. For reasons of cost efficiency, models are isostatic, and do not consider the local effects of unsupported loads or bending stresses. The requirements are: (1) ability to represent rigid rotation on a sphere; (2) ability to represent a spatially uniform strain-rate tensor in the limit of small elements; and (3) continuity of velocity across all element boundaries. Author designed a 3-node triangle shell element which has two different sets of basis functions to represent (vector) velocity and all other (scalar) variables. Such elements can be shown to converge to the formulas for plane triangles in the limit of small size, but can also applied to cover any area smaller than a hemisphere. The difficult volume integrals involved in computing the stiffness of such elements are performed numerically using 7 Gauss integration points on the surface of the sphere, beneath each of which a vertical integral is performed using about 100 points.

Bird, Peter↗
NASA NTRS1988

The Clouds distributed operating system - Functional description, implementation details and related work

Clouds is an operating system in a novel class of distributed operating systems providing the integration, reliability, and structure that makes a distributed system usable. Clouds is designed to run on a set of general purpose computers that are connected via a medium-of-high speed local area network. The system structuring paradigm chosen for the Clouds operating system, after substantial research, is an object/thread model. All instances of services, programs and data in Clouds are encapsulated in objects. The concept of persistent objects does away with the need for file systems, and replaces it with a more powerful concept, namely the object system. The facilities in Clouds include integration of resources through location transparency; support for various types of atomic operations, including conventional transactions; advanced support for achieving fault tolerance; and provisions for dynamic reconfiguration.

Dasgupta, Partha↗
NASA NTRS2008

Mars Global Geologic Mapping: Amazonian Results

We are in the second year of a five-year effort to map the geology of Mars using mainly Mars Global Surveyor, Mars Express, and Mars Odyssey imaging and altimetry datasets. Previously, we have reported on details of project management, mapping datasets (local and regional), initial and anticipated mapping approaches, and tactics of map unit delineation and description [1-2]. For example, we have seen how the multiple types and huge quantity of image data as well as more accurate and detailed altimetry data now available allow for broader and deeper geologic perspectives, based largely on improved landform perception, characterization, and analysis. Here, we describe early mapping results, which include updating of previous northern plains mapping [3], including delineation of mainly Amazonian units and regional fault mapping, as well as other advances.

Tanaka, K. L.↗
NASA NTRS1995

Detecting Structural Failures Via Acoustic Impulse Responses

Advanced method of acoustic pulse reflectivity testing developed for use in determining sizes and locations of failures within structures. Used to detect breaks in electrical transmission lines, detect faults in optical fibers, and determine mechanical properties of materials. In method, structure vibrationally excited with acoustic pulse (a "ping") at one location and acoustic response measured at same or different location. Measured acoustic response digitized, then processed by finite-impulse-response (FIR) filtering algorithm unique to method and based on acoustic-wave-propagation and -reflection properties of structure. Offers several advantages: does not require training, does not require prior knowledge of mathematical model of acoustic response of structure, enables detection and localization of multiple failures, and yields data on extent of damage at each location.

Bayard, David S.↗
DOE OSTI2025

Machine Learning for Anomaly Detection in Neural Network Security and SRF Cavities

This dissertation explores the development and deployment of machine learning approaches to address critical challenges in anomaly detection across two distinct domains: neural network security in federated learning settings and cavity behavior analysis in particle accelerator operations at Jefferson Lab in Newport News, Virginia. Anomaly detection identifies deviations from expected patterns, safeguarding systems in cybersecurity, industry, and research against malicious activities and failures. This dissertation demonstrates how our machine learning approaches enhance detection accuracy and efficiency in both neural network security and industrial applications. First, we investigate vulnerabilities in deep neural networks deployed in federated learning. Although federated learning preserves user privacy by training models locally, it remains vulnerable to backdoor attacks, in which malicious participants embed hidden triggers that induce targeted misbehavior. We propose a self-supervised contrastive learning framework to detect and mitigate such backdoor attacks. In our experiments, this method achieves higher detection accuracy and lower false positive rates than existing defenses, while operating without access to local model updates or original training data and thus preserving the privacy guarantees of the federated setting. Second, we address the operational reliability of superconducting radio-frequency (SRF) cavities at the Continuous Electron Beam Accelerator Facility (CEBAF). Our research leverages an unsupervised learning approach, combined with Principal Component Analysis (PCA) and k-means clustering, to identify anomalous behaviors in SRF cavities. Our method detects subtle anomalous behavior by analyzing SRF signal data. This knowledge allows for the early detection and resolution of potential faults, significantly improving the efficiency and reliability of operations. Third, we extend these insights to time-series anomaly detection more broadly. We design a contrastive-learning based model tailored to increasingly dynamic environments and academic research. This model improves detection accuracy in settings that require real-time monitoring and predictive maintenance. Our research underscores the broader applicability and impact of advanced machine learning techniques in anomaly detection. By extracting meaningful patterns from complex data, machine learning can significantly enhance security in distributed neural networks and improve the efficiency of particle accelerator operations. This dissertation serves as a stepping stone for future investigations into the vast possibilities of anomaly detection, inspiring further exploration and development of machine learning techniques in this field.

Ferguson, Hal [Old Dominion University]↗
DOE OSTI2026

Explaining System-Level Prognostics with Established Machine Learning Methods

System-level prognostics is crucial for ensuring reliability and enabling predictive maintenance in complex systems with interconnected components. This study presents a framework that integrates data-driven methods to predict the remaining useful life (RUL) of a subsystem under multiple and concurrent faults within a nuclear power plant system with explainable artificial intelligence (XAI). A nuclear power plant (NPP) operation was simulated to model the degradation behavior of NPP components, and four machine learning models—Gradient Boosting Regressor (GBR), Support Vector Regressor (SVR), Fully Connected Neural Network (FCNN), and Long Short-Term Memory (LSTM)—were evaluated for prognostics with a novel system RUL parameter. The LSTM model demonstrated potential superior repeatability, while SHAP (SHapley Additive exPlanations) for explainability provided consistent and trustworthy global explanations. In contrast, LIME (Local Interpretable Model-agnostic Explanations) offered localized interpretability but showed reduced stability for sequential data. Key findings include the interplay between component-level degradation and system-wide performance, with LSTM effectively capturing these dynamics through sequence-level predictions. The XAI techniques enhanced transparency by identifying critical features influencing model predictions and aligning with domain knowledge. Furthermore, this framework has significant implications for improving trust and understanding in predictive maintenance, particularly in safety-critical industries like nuclear energy.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
NASA NTRS2003

Numerical Simulations of Silverpit Crater Collapse

The Silverpit crater is a recently discovered, 60-65 Myr old complex crater, which lies buried beneath the North Sea, about 150 km east of Britain. High-resolution images of Silverpit's subsurface structure, provided by three-dimensional seismic reflection data, reveal an inner-crater morphology similar to that expected for a 5-8 km diameter terrestrial crater. The crater walls show evidence of terrace-style slumping and there is a distinct central uplift, which may have produced a central peak in the pristine crater morphology. However, Silverpit is not a typical 5-km diameter terrestrial crater, because it exhibits multiple, concentric rings outside the main cavity. External concentric rings are normally associated with much larger impact structures, for example Chicxulub on Earth, or Orientale on the Moon. Furthermore, external rings associated with large impacts on the terrestrial planets and moons are widely-spaced, predominantly inwardly-facing, asymmetric scarps. However, the seismic data show that the external rings at Silverpit represent closely-spaced, concentric faultbound graben, with both inwardly and outwardly facing fault-scarps. This type of multi-ring structure directly analogous to the Valhalla-type multi-ring basins found on the icy satellites. Thus, the presence and style of the multiple rings at Silverpit is surprising given both the size of the crater and its planetary setting. A further curiosity of the Silverpit structure is that the external concentric rings appear to be extensional features on the West side of the crater and compressional features on the East side. The crater also lies in a local depression, thought to be created by postimpact movement of a salt layer buried beneath the crater.

Collins, G. S.↗
NASA NTRS2002

Gilgamesh: A Multithreaded Processor-In-Memory Architecture for Petaflops Computing

Processor-in-Memory (PIM) architectures avoid the von Neumann bottleneck in conventional machines by integrating high-density DRAM and CMOS logic on the same chip. Parallel systems based on this new technology are expected to provide higher scalability, adaptability, robustness, fault tolerance and lower power consumption than current MPPs or commodity clusters. In this paper we describe the design of Gilgamesh, a PIM-based massively parallel architecture, and elements of its execution model. Gilgamesh extends existing PIM capabilities by incorporating advanced mechanisms for virtualizing tasks and data and providing adaptive resource management for load balancing and latency tolerance. The Gilgamesh execution model is based on macroservers, a middleware layer which supports object-based runtime management of data and threads allowing explicit and dynamic control of locality and load balancing. The paper concludes with a discussion of related research activities and an outlook to future work.

management locality load balance↗
NASA NTRS2013

Memory-Efficient Onboard Rock Segmentation

Rockster-MER is an autonomous perception capability that was uploaded to the Mars Exploration Rover Opportunity in December 2009. This software provides the vision front end for a larger software system known as AEGIS (Autonomous Exploration for Gathering Increased Science), which was recently named 2011 NASA Software of the Year. As the first step in AEGIS, Rockster-MER analyzes an image captured by the rover, and detects and automatically identifies the boundary contours of rocks and regions of outcrop present in the scene. This initial segmentation step reduces the data volume from millions of pixels into hundreds (or fewer) of rock contours. Subsequent stages of AEGIS then prioritize the best rocks according to scientist- defined preferences and take high-resolution, follow-up observations. Rockster-MER has performed robustly from the outset on the Mars surface under challenging conditions. Rockster-MER is a specially adapted, embedded version of the original Rockster algorithm ("Rock Segmentation Through Edge Regrouping," (NPO- 44417) Software Tech Briefs, September 2008, p. 25). Although the new version performs the same basic task as the original code, the software has been (1) significantly upgraded to overcome the severe onboard re source limitations (CPU, memory, power, time) and (2) "bulletproofed" through code reviews and extensive testing and profiling to avoid the occurrence of faults. Because of the limited computational power of the RAD6000 flight processor on Opportunity (roughly two orders of magnitude slower than a modern workstation), the algorithm was heavily tuned to improve its speed. Several functional elements of the original algorithm were removed as a result of an extensive cost/benefit analysis conducted on a large set of archived rover images. The algorithm was also required to operate below a stringent 4MB high-water memory ceiling; hence, numerous tricks and strategies were introduced to reduce the memory footprint. Local filtering operations were re-coded to operate on horizontal data stripes across the image. Data types were reduced to smaller sizes where possible. Binary- valued intermediate results were squeezed into a more compact, one-bit-per-pixel representation through bit packing and bit manipulation macros. An estimated 16-fold reduction in memory footprint relative to the original Rockster algorithm was achieved. The resulting memory footprint is less than four times the base image size. Also, memory allocation calls were modified to draw from a static pool and consolidated to reduce memory management overhead and fragmentation. Rockster-MER has now been run onboard Opportunity numerous times as part of AEGIS with exceptional performance. Sample results are available on the AEGIS website at http://aegis.jpl.nasa.gov.

Burl, Michael C.↗