Search NASA⌕ Search

SEARCH · Search NASA

Results for “resilient algorithms”

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 19 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

Optical storage media data integrity studies

Optical disk-based information systems are being used in private industry and many Federal Government agencies for on-line and long-term storage of large quantities of data. The storage devices that are part of these systems are designed with powerful, but not unlimited, media error correction capacities. The integrity of data stored on optical disks does not only depend on the life expectancy specifications for the medium. Different factors, including handling and storage conditions, may result in an increase of medium errors in size and frequency. Monitoring the potential data degradation is crucial, especially for long term applications. Efforts are being made by the Association for Information and Image Management Technical Committee C21, Storage Devices and Applications, to specify methods for monitoring and reporting to the user medium errors detected by the storage device while writing, reading or verifying the data stored in that medium. The Computer Systems Laboratory (CSL) of the National Institute of Standard and Technology (NIST) has a leadership role in the development of these standard techniques. In addition, CSL is researching other data integrity issues, including the investigation of error-resilient compression algorithms. NIST has conducted care and handling experiments on optical disk media with the objective of identifying possible causes of degradation. NIST work in data integrity and related standards activities is described.

Podio, Fernando L.↗
NASA NTRS1991

A Scheduling Algorithm for Replicated Real-Time Tasks

We present an algorithm for scheduling real-time periodic tasks on a multiprocessor system under fault-tolerant requirement. Our approach incorporates both the redundancy and masking technique and the imprecise computation model. Since the tasks in hard real-time systems have stringent timing constraints, the redundancy and masking technique are more appropriate than the rollback techniques which usually require extra time for error recovery. The imprecise computation model provides flexible functionality by trading off the quality of the result produced by a task with the amount of processing time required to produce it. It therefore permits the performance of a real-time system to degrade gracefully. We evaluate the algorithm by stochastic analysis and Monte Carlo simulations. The results show that the algorithm is resilient under hardware failures.

Yu, Albert C.↗
NASA NTRS1998

Resiliency of the Multiscale Retinex Image Enhancement Algorithm

The multiscale retinex with color restoration (MSRCR) continues to prove itself in extensive testing to be very versatile automatic image enhancement algorithm that simultaneously provides dynamic range compression, color constancy, and color rendition, However, issues remain with regard to the resiliency of the MSRCR to different image sources and arbitrary image manipulations which may have been applied prior to retinex processing. In this paper we define these areas of concern, provide experimental results, and, examine the effects of commonly occurring image manipulation on retinex performance. In virtually all cases the MSRCR is highly resilient to the effects of both the image source variations and commonly encountered prior image-processing. Significant artifacts are primarily observed for the case of selective color channel clipping in large dark zones in a image. These issues are of concerning the processing of digital image archives and other applications where there is neither control over the image acquisition process, nor knowledge about any processing done on th data beforehand.

Rahman, Zia-Ur↗
NASA NTRSDate not supplied

Extracting Lessons of Resilience Using Machine Mining of the ASRS Database

NASA’s Aviation Safety Reporting System (ASRS) database is the world's largest repository of voluntary, confidential safety information provided by aviation's frontline personnel, including pilots, air traffic controllers, mechanics, flight attendants, dispatchers, and other members of the aviation community and the public. The database contains close to 2 million narratives, many of which describe everyday situations in which people saved the day. In these situations, people’s resilient behavior solved a problem, dealt with a malfunction, and maintained a safe operation despite a serious perturbation. To be able to extract lessons of such resilience from this large database, the use of machine learning algorithms is being explored. In this report, we describe a comparison between two such algorithms: Perilog and Word2Vec. An identical search using both programs was done on a database containing approximately 470,000 ASRS reports submitted between 1988 and 2022. The comparison reveals some of the strength and weaknesses of each algorithm as well as the challenges inherent in using such algorithms to extract lessons of resilience from the ASRS database.

resilience↗
NASA NTRS2022

Radiation Tolerance and Mitigation for Neuromorphic Processors

Neuromorphic processors are designed to execute Deep Neural Networks (DNNs) at very high speed using only a fraction of the electrical power needed to run a DNN on a traditional CPU or GPU. This unique capability makes Neuromorphic processors a prime candidate for space systems, where advanced computational tasks like image analysis, depth map reconstruction, or rover control need to be executed in a power-starved environment. In contrast to the growing number of applications of Neuromorphic processors in smart phones, the automotive and robotics domain, the space environment is unforgiving because of extreme temperatures and high levels of radiation. Any space system, operating beyond LEO requires computing hardware that is resilient against radiation effects. However, Neuromorphic processors have not yet been designed or tested for their radiation tolerance. In this report, we consider traditional methods of detection of radiation events and mitigation via redundancy and gauge their effectiveness on DNNs. In contrast to traditional flight software, however, neural networks represent a statistical algorithm, which might affect its resilience against radiation events. We will focus on the analysis of the tolerance of DNNs with respect to radiation events and discuss techniques to detect radiation hits using on-chip triple modular redundancy (TMR) on an Intel Loihi neuromorphic processor and to mitigate radiation damage. We describe an architecture for on-chip TMR for the Intel Loihi and present results of initial experiments.

Neural Networks↗
NASA NTRSDate not supplied

Understanding Resilience Optimization Architectures With an Optimization Problem Repository

Optimizing a system’s resilience can be challenging, especially when it involves considering both the inherent resilience of a robust design and the active resilience of a health management system to a set of computationally-expensive hazard simulations. While prior work has developed specialized architectures to effectively and efficiently solve combined design and resilience optimization problems, the comparison of these architectures has been limited to a single case study. To further study resilience optimization formulations, this work develops a problem repository which includes previously-developed resilience optimization problems and additional problems presented in this work: a notional system resilience model, a pandemic response model, and a cooling tank hazard prevention model. This work then uses models in the repository at large to understand the characteristics of resilience optimization problems and study the applicability of optimization architectures and decomposition strategies. Based on the comparisons in the repository, applying an optimization architecture effectively requires understanding the alignment and coupling relationships between the design and resilience models, as well as the efficiency characteristics of the algorithms. While alignment determines the necessity of a surrogate of resilience cost in the upper-level design problem, coupling determines the overall applicability of a sequential, alternating, or bilevel structure. Additionally, the application of decomposition strategies is dependent on there being limited interactions between variable sets, which often does not hold when a resilience policy is parameterized in terms of actions to take in hazardous model states rather than specific given scenarios.

Resilience↗
NASA NTRS2007

Automated Target Planning for FUSE Using the SOVA Algorithm

The SOVA algorithm was originally developed under the Resilient Systems and Operations Project of the Engineering for Complex Systems Program from NASA s Aerospace Technology Enterprise as a conceptual framework to support real-time autonomous system mission and contingency management. The algorithm and its software implementation were formulated for generic application to autonomous flight vehicle systems, and its efficacy was demonstrated by simulation within the problem domain of Unmanned Aerial Vehicle autonomous flight management. The approach itself is based upon the precept that autonomous decision making for a very complex system can be made tractable by distillation of the system state to a manageable set of strategic objectives (e.g. maintain power margin, maintain mission timeline, and et cetera), which if attended to, will result in a favorable outcome. From any given starting point, the attainability of the end-states resulting from a set of candidate decisions is assessed by propagating a system model forward in time while qualitatively mapping simulated states into margins on strategic objectives using fuzzy inference systems. The expected return value of each candidate decision is evaluated as the product of the assigned value of the end-state with the assessed attainability of the end-state. The candidate decision yielding the highest expected return value is selected for implementation; thus, the approach provides a software framework for intelligent autonomous risk management. The name adopted for the technique incorporates its essential elements: Strategic Objective Valuation and Attainability (SOVA). Maximum value of the approach is realized for systems where human intervention is unavailable in the timeframe within which critical control decisions must be made. The Far Ultraviolet Spectroscopic Explorer (FUSE) satellite, launched in 1999, has been collecting science data for eight years.[1] At its beginning of life, FUSE had six gyros in two IRUs and four reaction wheels. Over time through various failures, the satellite has been left with one reaction wheel on the vehicle skew axis and two gyros. To remain operational, a control scheme has been implemented using the magnetic torque rods and the remaining momentum wheel.[2] As a consequence, there are attitude regions where there is insufficient torque authority to overcome environmental disturbances (e.g. gravity gradient torques). The situation is further complicated by the fact that these attitude regions shift inertially with time as the spacecraft moves through earth s magnetic field during the course of its orbit. Under these conditions, the burden of planning targets and target-to-target slew maneuvers has increased significantly since the beginning of the mission.[3] Individual targets must be selected so that the magnetic field remains roughly aligned with the skew wheel axis to provide enough control authority to the other two orthogonal axes. If the field moves too far away from the skew axis, the lack of control authority allows environmental torques to pull the satellite away from the target and can potentially cause it to tumble. Slew maneuver planning must factor the stability of targets at the beginning and end, and the torque authority at all points along the slew. Due to the time varying magnetic field geometry relative to any two inertial targets, small modifications in slew maneuver timing can make large differences in the achievability of a maneuver.

Heatwole, Scott↗
NASA NTRS2019

Towards Resilient Autonomous Navigation of Drones

Robots and particularly drones are especially useful in exploring extreme environments that pose hazards to humans. To ensure safe operations in these situations, usually perceptually degraded and without good GNSS, it is critical to have a reliable and robust state estimation solution. The main body of literature in robot state estimation focuses on developing complex algorithms favoring accuracy. Typically, these approaches rely on a strong underlying assumption: the main estimation engine will not fail during operation. In contrast, we propose an architecture that pursues robustness in state estimation by considering redundancy and heterogeneity in both sensing and estimation algorithms. The architecture is designed to expect and detect failures and adapt the behavior of the system to ensure safety. To this end, we present HeRO (Heterogeneous Redundant Odometry): a stack of estimation algorithms running in parallel supervised by a resiliency logic. This logic carries out three main functions: a) perform confidence tests both in data quality and algorithm health; b) re-initialize those algorithms that might be malfunctioning; c) generate a smooth state estimate by multiplexing the inputs based on their quality. The state and quality estimates are used by the guidance and control modules to adapt the mobility behaviors of the system. The validation and utility of the approach are shown with real experiments on a ying robot for the use case of autonomous exploration of subterranean environments, with particular results from the STIX event of the DARPA Subterranean Challenge.

Agha-mohammadi, Ali-akbar↗
NASA NTRSDate not supplied

NASA and Blue Origin Collaborative Assessment of Precision Landing Algorithms and Computing

NASA’s Safe and Precise Landing Integrated Capabilities Evolution (SPLICE) project is developing sensor, algorithm, and compute technologies for precision landing and hazard avoidance. These technologies are being tested as an integrated Precision Landing and Hazard Avoidance (PL&HA) system on Blue Origin’s New Shephard suborbital vehicle. A key goal for the computing element of this technology development is to characterize the performance of the SPLICE software workloads on the project’s Descent and Landing Computer (DLC). The DLC is a multi-core processor designed as a surrogate for NASA’s High-Performance Space Computer (HPSC). Measurements of the SPLICE workload performance on the DLC provides NASA insight on how PL&HA capabilities will perform on the HPSC, and guidance on how the SPLICE algorithms can be implemented to best utilize the DLC platform. This insight can also be used to derive requirements to guide trade studies on candidate computing architectures, for use on platforms like Blue Moon. NASA and Blue Origin are collaborating under an agreement to pursue this mutual benefit. Performance metrics collected are based on measurement of common compute resources such as percentage used of memory bandwidth, I/O utilization, interrupt latency, and kernel vs. user space code residency. Where possible existing performance counters and metrics that are part of the operating system kernel are used. As the design has a significant FPGA component, performance counters are identified and instantiated in the fabric to measure DMA performance and interface metrics. Collection of metrics is performed on the DLC with a representative workload that simulates a full landing cycle of the Blue Origin New Shepard vehicle. Consideration is given to the other compute implementations and whether they can run SPLICE algorithms at the same rate and with the same latency as the DLC. One option being considered is the use of a RISC-V soft core instantiated in a radiation resilient FPGA fabric such as the Xilinx KU60. Select algorithms from the SPLICE code will be run for comparison with the DLC. This paper describes how the DLC is instrumented to collect performance measurements of the SPLICE workloads, preliminary results from these measurements, and their implications on SPLICE algorithm implementation. The results of experimentation to derive candidate requirements for architecture trades on a PL&HA computing system are also presented.

computer performance↗
NASA NTRS2018

A Prognostics Framework Development for Swarm Satellite Formations

Prognostics is the science of predicting the failure(s) of a component or a system and understanding how the performance will change in the event of a failure or degradation mechanism. With accurate predictions of possible failures, autonomous mitigative actions can be taken to correct/repair any issues or alert human operators of a failure threshold exceedance requiring condition-based maintenance. Although there is extensive research on failure predictions for a component or a system, there are significantly more opportunities to foray into failure predictions and prognostics for a system of systems such as an airspace consisting of multiple aircraft, a fleet of unmanned aerial vehicles, and a swarm of intelligent satellite systems. Failure prediction and mitigation are particularly important in autonomous systems such as satellite swarm systems that need effective resource management and minimal human interactions. Based on NASA's decadal survey, there is a clear need to prioritize the development of satellite swarm technology for studies of space physics and Earth science. The science community will propose future missions that return in-situ measurements from a 3-D (three-dimensional) volume of space, with relative spacecraft motion and inter-satellite baselines controlled according to the mission objectives. For such multi-spacecraft missions, it is required that ground operations resources do not scale with the number of satellites, thus compromising the swarm or leading to inefficiencies in resource allocation. Swarms of tens or hundreds of small satellites will require autonomy in attitude control, navigation and failure. Although significant research has been conducted in the areas of autonomous formation flying algorithms, less attention has been given to the development of resilient systems robust to failures.The focus of this research paper is the integration of model-based prognostics into the swarm dynamics control and decision-making algorithms. We simulate swarm management strategies for a subsystem failure to demonstrate the importance of failure predictions by comparing two cases: (i) no health information is provided to the system and utilized in the decision-making process and (2) system health information is obtained using prognostics and employed by the control system. One example scenario presented is for the GPS (Global Positioning System) system of an individual satellite to perform off-nominally due to increasing estimated error. In this scenario, the keep-out zone for that satellite would become more conservative, thereby decreasing the risk of collision. This is achieved via tuning the individual artificial repulsive functions assigned to each satellite.This paper is structured as follows. First we provide an overview of current swarm technology development, where we specifically use the term swarm to define multiple satellites flying in formation in similar orbits, with cross-link communication and station-keeping capabilities. Second, we give an introduction to the Swarm Orbital Dynamics Advisor (SODA), a tool that accepts high-level configuration commands and provides the orbital maneuvers required to achieve the prescribed formation configuration. Third, we provide the details of the model-based prognostics algorithm implementation in SODA. Finally, we present different case studies for potential component/subsystem failures and the swarm responses based with and without failure prediction information.

prognostics↗
NASA NTRS2018

Optimization of Second Fault Detection Thresholds to Maximize Mission POS

In order to support manned spaceflight safety requirements, the Space Launch System (SLS) has defined program-level requirements for key systems to ensure successful operation under single fault conditions. To accommodate this with regards to Navigation, the SLS utilizes an internally redundant Inertial Navigation System (INS) with built-in capability to detect, isolate, and recover from first failure conditions and still maintain adherence to performance requirements. The unit utilizes multiple hardware- and software-level techniques to enable detection, isolation, and recovery from these events in terms of its built-in Fault Detection, Isolation, and Recovery (FDIR) algorithms. Successful operation is defined in terms of sufficient navigation accuracy at insertion while operating under worst case single sensor outages (gyroscope and accelerometer faults at launch). In addition to first fault detection and recovery, the SLS program has also levied requirements relating to the capability of the INS to detect a second fault, tracking any unacceptable uncertainty in knowledge of the vehicle's state. This detection functionality is required in order to feed abort analysis and ensure crew safety. Increases in navigation state error and sensor faults can drive the vehicle outside of its operational as-designed environments and outside of its performance envelope causing loss of mission, or worse, loss of crew. The criteria for operation under second faults allows for a larger set of achievable missions in terms of potential fault conditions, due to the INS operating at the edge of its capability. As this performance is defined and controlled at the vehicle level, it allows for the use of system level margins to increase probability of mission success on the operational edges of the design space. Due to the implications of the vehicle response to abort conditions (such as a potentially failed INS), it is important to consider a wide range of failure scenarios in terms of both magnitude and time. As such, the Navigation team is taking advantage of the INS's capability to schedule and change fault detection thresholds in flight. These values are optimized along a nominal trajectory in order to maximize probability of mission success, and reducing the probability of false positives (defined as when the INS would report a second fault condition resulting in loss of mission, but the vehicle would still meet insertion requirements within system-level margins). This paper will describe an optimization approach using Genetic Algorithms to tune the threshold parameters to maximize vehicle resilience to second fault events as a function of potential fault magnitude and time of fault over an ascent mission profile. The analysis approach, and performance assessment of the results will be presented to demonstrate the applicability of this process to second fault detection to maximize mission probability of success.

Anzalone, Evan↗
NASA NTRS2017

Initial Results of Heuristic Guided Orbit Selection for a Low Frequency Radio Interferometric Spacecraft Constellation

A constellation of radio telescope spacecraft can leverage interferometry to accurately image distant objects throughout the universe, but mission design must balance among many interrelated constraints. In particular, the number of craft and the selection of time-varying orbital parameters play a pivotal role in determining what interferometric baselines are feasible with respect to different targets, and thus drives the breadth and quality of data available to the constellation. The large combinatorial orbit configuration space and competing concerns present a challenging problem that is not well addressed by traditional mission design processes. This paper describes application of automated optimization methods to help direct mission design effort to the most promising dynamic constellation geometries: those that achieve broad interferometric coverage but remain cost-effective and resilient to failures. Several automatic heuristic-driven optimization algorithms representing complementary search strategies were created to explore among concrete constellation configuration plans. Evaluation of each candidate constellation plan was accelerated by efficiently combining precomputed caches of orbital and interferometric data. Results indicate that leveraging automated optimization for constellation mission design is both practical and illuminating: generated solutions provided both evidence for existing design intuitions as well as fresh insights into novel configurations.

Hernandez, Sonia↗
NASA NTRSDate not supplied

Extracting Lessons of Human Performance and Resilience Management from a Study of Weather-Related Safety Risks and Incidents Reported to NASA's Aviation Safety Reporting System (ASRS)

We present a study of weather-related incident reports submitted to NASA’s Aviation Safety Reporting System (ASRS) by air carrier pilots and air traffic controllers in the US. Using specific examples, we examine the weather-related risks reported and the relevant aspects of human performance and resilience management exhibited during these incidents. We describe our methods of lesson extraction to maximize learning, including the use of advance algorithms, and we conclude with a review of the weather-related risks encountered and the resilient practices employed to mitigate them.

resilience↗
NASA NTRS1987

Reliable communication in the presence of failures

The design and correctness of a communication facility for a distributed computer system are reported on. The facility provides support for fault-tolerant process groups in the form of a family of reliable multicast protocols that can be used in both local- and wide-area networks. These protocols attain high levels of concurrency, while respecting application-specific delivery ordering constraints, and have varying cost and performance that depend on the degree of ordering desired. In particular, a protocol that enforces causal delivery orderings is introduced and shown to be a valuable alternative to conventional asynchronous communication protocols. The facility also ensures that the processes belonging to a fault-tolerant process group will observe consistant orderings of events affecting the group as a whole, including process failures, recoveries, migration, and dynamic changes to group properties like member rankings. A review of several uses for the protocols is the ISIS system, which supports fault-tolerant resilient objects and bulletin boards, illustrates the significant simplification of higher level algorithms made possible by our approach.

Birman, Kenneth P.↗
NASA NTRS2000

A Real-Time High Performance Data Compression Technique For Space Applications

A high performance lossy data compression technique is currently being developed for space science applications under the requirement of high-speed push-broom scanning. The technique is also error-resilient in that error propagation is contained within a few scan lines. The algorithm is based on block-transform combined with bit-plane encoding; this combination results in an embedded bit string with exactly the desirable compression rate. The lossy coder is described. The compression scheme performs well on a suite of test images typical of images from spacecraft instruments. Hardware implementations are in development; a functional chip set is expected by the end of 2001.

Yeh, Pen-Shu↗
NASA NTRS2000

Visually Lossless Data Compression for Real-Time Frame/Pushbroom Space Science Imagers

A visually lossless data compression technique is currently being developed for space science applications under the requirement of high-speed push-broom scanning. The technique is also applicable to frame based imaging and is error-resilient in that error propagation is contained within a few scan lines. The algorithm is based on a block transform of a hybrid of modulated lapped transform (MLT) and discrete cosine transform (DCT), or a 2-dimensional lapped transform, followed by bit-plane encoding; this combination results in an embedded bit string with exactly the desirable compression rate as desired by the user. The approach requires no unique table to maximize its performance. The compression scheme performs well on a suite of test images typical of images from spacecraft instruments. Flight qualified hardware implementations are in development; a functional chip set is expected by the end of 2001. The chip set is being designed to compress data in excess of 20 Msamples/sec and support quantizations from 2 to 16 bits.

Yeh, Pen-Shu↗
NASA NTRS2018

Optimization of Second Fault Detection Thresholds to Maximize Mission Probability of Success

In order to support manned spaceflight safety requirements, the Space Launch System (SLS) has defined program-level requirements for key systems to ensure successful operation under single fault conditions. The SLS program has also levied requirements relating to the capability of the Inertial Navigation System to detect a second fault. This detection functionality is required in order to feed abort analysis and ensure crew safety. Increases in navigation state error due to sensor faults in a purely inertial system can drive the vehicle outside of its operational as-designed environmental and performance envelope. As this performance outside of first fault detections is defined and controlled at the vehicle level, it allows for the use of system level margins to increase probability of mission success on the operational edges of the design. A top-down approach is utilized to assess vehicle sensitivity to second sensor faults. A wide range of failure scenarios in terms of both fault magnitude and time is used for assessment. The approach also utilizes a schedule to change fault detection thresholds autonomously. These individual values are optimized along a nominal trajectory in order to maximize probability of mission success in terms of system-level insertion requirements while minimizing the probability of false positives. This paper will describe an approach integrating Genetic Algorithms and Monte Carlo analysis to tune the threshold parameters to maximize vehicle resilience to second fault events over an ascent mission profile. The analysis approach and performance assessment and verification will be presented to demonstrate the applicability of this approach to second fault detection optimization to maximize mission probability of success through taking advantage of existing margin.

Anzalone, Evan J.↗
NASA NTRS2021

Autonomous Off-road Navigation over Extreme Terrains with Perceptually-challenging Conditions

We propose a framework for resilient autonomous navigation in perceptuallychallenging unknown environments with mobility-stressing elements such asuneven surfaces with rocks and boulders, steep slopes, negative obstacles like cliffsand holes, and narrow passages. Environments are GPS-denied and perceptuallydegradedwith variable lighting from dark to lit and obscurants (dust, fog, smoke).Lack of prior maps and degraded communication eliminates the possibility of prioror off-board computation or operator intervention. This necessitates real-time onboardcomputation using noisy sensor data. To address these challenges, we proposea resilient architecture that exploits redundancy and heterogeneity in sensing modalities.Further resilience is achieved by triggering recovery behaviors upon failure.We propose a fast settling algorithm to generate robust multi-fidelity traversabilityestimates in real-time. The proposed approach was deployed on multiple physicalsystems including skid-steer and tracked robots, high-speed RC car and legged robotsand as a part of Team CoSTAR’s effort to theDARPASubterranean Challenge, wherethe team won 2nd and 1st place in the Tunnel and Urban Circuit, respectively.

Agha-mohammadi, Ali-akbar↗