John C. Sloan

From Web Service Artifact to a Readable and Verifiable Model

Models of Web service compositions that are both readable and verifiable will benefit organizations that integrate purportedly reusable Web services. Colored Petri nets (CPNs) are at once verifiable and visually expressive, capable of presenting subtle flaws in service composition. Constructing CPN models from business process execution language (BPEL) artifacts had been a manual process requiring human judgment. Building on results from the workflow community, we automate the mapping of artifacts written in BPEL to models used by CPN Tools - a formal verification environment for development, simulation, and model checking of colored Petri nets. We extend related work that already converts BPEL to Petri nets, to reflect hierarchy and data type (color in CPN terminology), while improving model layout. We present a prototype implementation that mines both a BPEL artifact and the Petri net generated from it by an existing tool. The prototype partitions the Petri net into subnets, lays them out, colors them, and generates their XML file for import into CPN tools. Our results include depictions of subnets produced and initial simulation results for a well-known case study.

OCEAN TURBINES — A RELIABILITY ASSESSMENT

This paper identifies factors that impact reliability and safety of ocean turbines. We describe how physical and environmental factors will impact the design of its machine condition monitoring (MCM) system. Environmental factors like fouling, corrosion, and inaccessibility of equipment sets this MCM problem apart from those encountered by wind turbines, hydroelectric plants, or even ship hulls and propellers. Fouling constitutes the primary and most persistent source of failure. In addition to compromising turbine efficiency and reliability, fouling reduces sensor data quality — masking faults that will ultimately lead to failure. Unmitigated fouling triggers a form of biological succession known as flocculation that may eventually attract threatened species of tortoises and cetaceans to this rotating machinery. We review and suggest refinements to a class of non-toxic biologically-inspired anti-fouling techniques known as engineered topographies. Advances in this area will enable turbines to operate in portions of the water column that maximize momentum flux while minimizing retrieval cost.

Fourier transforms for vibration analysis: A review and case study

An important part of machine condition monitoring and prognostic health monitoring (MCM/PHM) using waveform sensors (such as vibration sensors) is data transformation, where the output from accelerometers is transformed into the time-frequency domain. Although Fourier analysis is a respected time-frequency transform in other application domains, many works in the field of MCM/PHM recommend against it and suggest more elaborate transforms such as wavelet-based approaches. These recommendations are not often justified by experimental data, however. In the present work, we present a discussion of the current state of research on Fourier transforms for MCM/PHM of vibration signals, as well as a case study demonstrating that for some experiments, the Fourier transform can produce very good results.

Assuring Timeliness in an e-Science Service-Oriented Architecture

An improvement to public-resource e-science portals shows promise in solving a well-known dilemma: how to dynamically discover a provider PC that is ready to deliver computing power when the scientific community requires it.

FEATURE SELECTION FOR OPTIMIZATION OF WAVELET PACKET DECOMPOSITION IN RELIABILITY ANALYSIS OF SYSTEMS

One of the most important types of signal found in the area of machine condition monitoring/prognostic health monitoring (MCM/PHM) is the vibration signal, a type of waveform. Many time-frequency domain techniques have been proposed to interpret such signals, including wavelet packet decomposition (WPD). Previous work has shown how to extend the WPD algorithm to operate on streaming signals, but the number of output variables becomes exponential in the number of levels of decomposition, hindering data mining in limited-memory environments. Feature selection techniques, well understood in other areas of data mining, can be used to greatly reduce the number of output variables and speed up the machine learning algorithms. This paper presents a case study comparing two versions of WPD both with and without feature selection, demonstrating that removing most of the features produced by the WPD does not impair its performance within the context of MCM/PHM.

Using Feature Selection to Determine Optimal Depth for Wavelet Packet Decomposition of Vibration Signals for Ocean System Reliability

Vibration signals are an important source of information for machine condition monitoring/prognostic health monitoring to ensure the reliability of ocean systems. Because they are waveforms, vibration data must be transformed into the frequency domain before they can be used to build classification and prediction models. One popular transformation is wavelet packet decomposition, a higher resolution variant of wavelet transformation. For wavelet packet decomposition, depth is an important parameter to control the maximum level of detail while minimizing the computational time when constructing and using the decomposition tree. Little guidance exists in the literature to assist researchers in choosing a depth, however. In this paper, we present a feature selection-based approach to determining the optimum depth for wavelet packet decomposition. First, the data is transformed using a very high depth, and all of the features are ordered based on their importance for predicting the class. Then, a depth which captures the most important features is chosen. Finally, a model is built using that depth. We show that a classification model built according to this procedure retains almost all of the accuracy of models built using a much deeper transform, while allowing for smaller depths and vastly fewer features.

Feature Selection for Vibration Sensor Data Transformed by a Streaming Wavelet Packet Decomposition

Vibration signals play a valuable role in the remote monitoring of high-assurance machinery such as ocean turbines. Because they are waveforms, vibration data must be transformed prior to being incorporated into a machine condition monitoring/prognostic health monitoring (MCM/PHM) solution to detect which frequencies of oscillation are most prevalent. One downside of these transformations, especially the streaming version of the wavelet packet decomposition (denoted SWPD), is that they can produce a large number of features, hindering the model building and evaluation process. In this paper we demonstrate how feature selection techniques may be applied to the output of the SWPD transformation, vastly reducing the total number of features used to build models. The resulting data can be used to build more accurate models for use in MCM/PHM while minimizing computation time.

A Dynamometer for an Ocean Turbine Prototype: Reliability through Automated Monitoring

An ocean turbine extracts the kinetic energy from ocean currents to generate electricity. Machine Condition Monitoring(MCM) / Prognostic Health Monitoring (PHM) systems allow for self-checking and automated fault detection, and are integral in the construction of a highly reliable ocean turbine. This paper presents an onshore test platform for an ocean turbine as well as a case study showing how machine learning can be used to detect changes in the operational state of this plant based on its vibration signals. In the case study, seven widely used machine learners a retrained on experimental data gathered from the test platform, a dynamometer, to detect changes in the machinesstate. The classification models generated by these classifiers are being considered as possible components of the state detection module of an MCM/PHM system for ocean turbines, and would be used for fault prediction. Experimental results presented here show the effectiveness of decision tree and random forest learners on distinguishing between faulty and normal states based on vibration data preprocessed by a wavelet transform.

TESTING AND FORMAL VERIFICATION OF SERVICE ORIENTED ARCHITECTURES

We examine two open engineering problems in the area of testing and formal verification of internet-enabled service oriented architectures (SOA). The first involves deciding when to formally and exhaustively verify versus when to informally and non-exhaustively test. The second concerns scalability limitations associated with formal verification, to which we propose a semi-formal technique that uses software agents. Finally, we assess how these findings can improve current software quality assurance practices. Addressing the first problem, we present and explain two classes of tradeoffs. External tradeoffs between assurance, performance, and flexibility are determined by the business needs of each application, whether it be in engineering, commerce, or entertainment. Internal tradeoffs between assurance, scale, and level of detail involve the technical challenges of feasibly verifying or testing an SOA. To help decide whether to exhaustively verify or non-exhaustively test, we present and explain these two classes of tradeoffs. Identifying a middle ground between testing and verification, we propose using software agents to simulate services in a composition. Technologically, this approach has the advantage of assuring the quality of compositions that are too large to exhaustively verify. Operationally, it supports testing these compositions in the laboratory without access to source code or use of network resources of third-party services. We identify and exploit the structural similarities between agents and services, examining how doing so can assure the quality of service compositions.

Ensemble Coordination for Discrete Event Control

An ensemble is a collection of independent processes, each tasked with drawing potentially differing conclusions about the same data. Using Petri nets, this paper formally describes how ensembles are organized and their behavior coordinated to effect distributed discrete event control of an ocean turbine prototype. Compositions, duals, reverses, and cliques formed over known Petri net graphs comprise the building blocks of the proposed ensemble coordination strategy. The behavior of an ensemble of controllers tasked with fault triage are subject to constraints formulated herein. The controller tasked with prognosis and health management (PHM) itself uses an ensemble of classifiers to detect faults. This ensemble is subject to constraints imposed by stream processing, which require a non-blocking form of rendezvous synchronization. Furthermore, results from each classifier must be fused in a manner that rewards that classifiers ability to predict faults. We identify two competing merit schemes -- one based on individual classifier performance and the other on performance of the sub-ensembles to which that classifier participates. Finally, we model check these Petri nets and report their results.