General Purpose Data-Driven System Monitoring for Space Operations

Iverson, David; Martin, Roscoe C.; Schwabacher, Mark; Spirkovska, Liljana; Taylor, William M.; Mackey, Ryan; Castle, J. P.

doi:10.2514/6.2009-1909

Cited by 44 publications

(32 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some details on SequenceMiner will be provided in Section 3.3. Other anomaly detection methods, such as Orca [3] and the Inductive Monitoring System (IMS) [12], find anomalies in multivariate continuous data. Both Orca and IMS are distance-based anomaly detection methods, in that they use a metric related to distance, such as the average Euclidean distance to its k-nearest neighbors, to assess the anomalousness of a point.…”

Section: Current State-of-the-art Algorithms On Anomaly Detectionmentioning

confidence: 99%

Multiple kernel learning for heterogeneous anomaly detection

Das

Matthews

Srivastava

et al. 2010

Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

158

118

View full text Add to dashboard Cite

The world-wide aviation system is one of the most complex dynamical systems ever developed and is generating data at an extremely rapid rate. Most modern commercial aircraft record several hundred flight parameters including information from the guidance, navigation, and control systems, the avionics and propulsion systems, and the pilot inputs into the aircraft. These parameters may be continuous measurements or binary or categorical measurements recorded in one second intervals for the duration of the flight. Currently, most approaches to aviation safety are reactive, meaning that they are designed to react to an aviation safety incident or accident. In this paper, we discuss a novel approach based on the theory of multiple kernel learning to detect potential safety anomalies in very large data bases of discrete and continuous data from world-wide operations of commercial fleets. We pose a general anomaly detection problem which includes both discrete and continuous data streams, where we assume that the discrete streams have a causal influence on the continuous streams. We also assume that atypical sequences of events in the discrete streams can lead to off-nominal system performance. We discuss the application domain, novel algorithms, and also discuss results on real-world data sets. Our algorithm uncovers operationally significant events in high dimensional data streams in the aviation industry which are not detectable using state of the art methods.

show abstract

Section: Current State-of-the-art Algorithms On Anomaly Detectionmentioning

confidence: 99%

Multiple kernel learning for heterogeneous anomaly detection

Das

Matthews

Srivastava

et al. 2010

Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

158

118

View full text Add to dashboard Cite

show abstract

“…The IMS [19,20] is similar to Orca in that it is distance-based. Like Orca, it uses Euclidean distance as its distance metric.…”

Section: Inductive Monitoring Systemmentioning

confidence: 99%

“…We therefore decided to use unsupervised anomaly detection algorithms, since they do not require labeled examples of anomalies. This article presents nine anomalies detected by four unsupervised anomaly detection algorithms: Orca [17], GritBot [18] (a commercial product), the Inductive Monitoring System (IMS) [19,20] and a one-class support vector machine (SVM) [21]. The next four subsections describe these four algorithms.…”

Section: Anomaly Detectionmentioning

confidence: 99%

Unsupervised Anomaly Detection for Liquid-Fueled Rocket Propulsion Health Monitoring

Schwabacher

Oza

Matthews

2009

Journal of Aerospace Computing, Information, and Communication

Self Cite

View full text Add to dashboard Cite

This article describes the results of applying four unsupervised anomaly detection algorithms to data from two rocket propulsion testbeds. The first testbed uses historical data from the Space Shuttle Main Engine. The second testbed uses data from an experimental rocket engine test stand located at NASA Stennis Space Center. The article describes nine anomalies detected by the four algorithms. The four algorithms use four different definitions of anomalousness. Orca uses a nearest-neighbor approach, defining a point to be an anomaly if its nearest neighbors in the data space are far away from it. The Inductive Monitoring System clusters the training data, and then uses the distance to the nearest cluster as its measure of anomalousness. GritBot learns rules from the training data, and then classifies points as anomalous if they violate these rules. One-class support vector machines map the data into a high-dimensional space in which most of the normal points are on one side of a hyperplane, and then classify points on the other side of the hyperplane as anomalous. Because of these different definitions of anomalousness, different algorithms detect different anomalies. We therefore conclude that it is useful to use multiple algorithms.

show abstract

“…The screening can be slow and labor intensive. Experimentation using general-purpose data-mining tools (clustering, outlier method, decision tree) [28,29] was conducted in flights STS-115 and 116. An expert systems approach was later employed to build an automatic detection tool by searching for data characteristics based on impact simulation and test experience (e.g., while an elevated G rms response can be caused by either a debris impact or mission event, the former shows up on multiple adjacent sensors and tapers off rapidly from the impacted panel, whereas the latter shows up globally across a large number of panels on both wings.…”

Section: Ascent Monitoringmentioning

confidence: 99%

Structural Health Monitoring and Risk Management of a Reusable Launch Vehicle

Yap

Macias

Kaouk³

et al. 2012

Journal of Spacecraft and Rockets

View full text Add to dashboard Cite

A structural-health-monitoring system can contribute to the risk management of a structure operating under hazardous conditions. An example is the wing leading-edge impact-detection system that monitors the debris hazards to the Space Shuttle Orbiter's reinforced carbon-carbon panels. Since return to flight after the Columbia accident, the system was developed and subsequently deployed on board the orbiter to detect ascent and on-orbit debris impacts, so as to support the assessment of wing leading edge structural integrity before orbiter reentry. As structural health monitoring is inherently an inverse problem, the analyses involved, including those performed for this system, tend to be associated with significant uncertainty. The use of probabilistic approaches to handle the uncertainty has resulted in the successful implementation of many development and application milestones. Nomenclature d = damage size, in. G max = maximum acceleration response G rms = rms of acceleration response P = damage probability P C = joint probability of critical damage P C=I = conditional probability of critical damage given an impact has occurred P F = probability of multisensor corroboration being weaker than flight experience P I = probability of impact P M = probability of multisensor corroboration being stronger than model experience r diag = diagonal corroboration ratio of multisensor response r horz = horizontal corroboration ratio of multisensor response = logarithmic decay dam = damage-G max correlation factor mea = test measurement scaling factor pan = reinforced carbon-carbon panel-to-panel test scaling factor sen = sensor configuration test scaling factor tst = test-article boundary condition scaling factor

show abstract

General Purpose Data-Driven System Monitoring for Space Operations

Cited by 44 publications

References 4 publications

Multiple kernel learning for heterogeneous anomaly detection

Multiple kernel learning for heterogeneous anomaly detection

Unsupervised Anomaly Detection for Liquid-Fueled Rocket Propulsion Health Monitoring

Structural Health Monitoring and Risk Management of a Reusable Launch Vehicle

Contact Info

Product

Resources

About