通常工程の状態は当初管理状態にあり,それが何らかの要因により管理状態から逸脱し,管理外れ状態に移行するものと考えられる.現工程の状態を判定する場合,現工程から得られたサンプル・データと併せて,これ以前に得られていたサンプル・データを十分に活用して現工程の状態を判定することは有意義な方法であるといえる.ここに,連続するサンプリングにおけるサンプル・データを活用し,工程状態を解析する方法として,CUSUM管理図やEWMA管理図といった時系列管理図法が存在する.ただし,従来の時系列管理図の検出特性の評価においては,時系列管理図の適用にあたって,適用当初から工程が管理外れ状態であった場合の検出特性が取り上げられている.これに対して,既述のように初期状態において管理状態にあった工程が何らかの要因によりある時点から管理外れ状態に移行する状況が一般的であると考えられる.そこで本研究では,まず工程の品質特性の管理状態の分布からの乖離の程度をKullback-Leibler情報量に基づき定量化し,これを打点統計量としそ工程の状態判定を行う(x,s)同時管理図をもとに,CUSUM管理図およびEWMA管理図といったいくつかの時系列管理図を設計する.さらに,管理図の適用当初から管理外れ状態にある状況のもとでの検出特性を評価する従来の研究とは別に,管理状態のままでいくらかの時間を経過した後に管理外れ状態に移行する状況を想定し,このもとでの管理外れ状態に対する検出特性を評価することを考える.くわえて,管理外れ状態を検出するばかりでなく,工程の管理状態からの逸脱期を特定することは品質管理・工程管理において重要な課題である.このことに鑑み,工程を時系列的に解析する上述の管理図をもとに,情報量規準に関する概念を応用し,どの時点において管理状態からの逸脱が生じたのかを解析的に識別する方法を提案する.

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers.

‘Measurement is not an end in itself …’ might seem to be a paradoxical way of introducing a book about measurement. But measurement is important to the majority since it gives objective evidence on which to base decisions. The need for quality-assured measurement has evolved and widened over the centuries. Quality-assured measurement remains a topic of burgeoning and increasingly multidisciplinary interest.

task; therefore, experimental methods based on statistical algorithms of data processing are being developed now. Specific features of application of these algorithms are considered.

p-Charts and np-Charts are commonly used in monitoring variables of the fraction type and these charts assume that the monitored variables are binomially distributed. In this paper we propose a new control chart called Beta Charts, for monitoring fraction data (p). The Beta Chart presents the control limits based on the Beta probability distribution. It was applied for monitoring the variables in three real studies, and it was compared to the control limits with three schemes. The comparative analysis showed that: (i) Beta approximation to the Binomial distribution was more appropriate with values confined in the [0,1]-interval; and (ii) the charts proposed were more sensitive to the average run length (ARL), in both in-control and out-of-control processes monitoring. The Beta Charts outperform the control charts analyzed for monitoring fraction data.

of the Dissertation .................................................................................................ii Acknowledgements ............................................................................................................. iv Table of

ii Dedication v Acknowledgements vi List of Tables ix List of Figures xi Chapter One Assembly Time Estimation Methods: A Review 1 1.1 Process Based Assembly Time Estimation 2 1.2 Product Based Assembly Time Estimation 11 Chapter Two Defining the Research Motivation 21 2.1 Research Questions 27 2.2 Research Roadmap 30 2.3 Exploratory Study 35 Chapter Three Interference Detection Method – Graph Generation 46 3.1 Demonstration of IDM 48 3.2 Interference Detection Method Graph Generation Test Cases 50 3.3 Comparison of Graph Generation Methods 53 Chapter Four Application of IDM During The Conceptual Design Stage 67 4.1 Set of Models 68 4.2 Reducing Model Fidelity 70 4.3 Artificial Neural Network Generation 71 4.4 Experimental Sets 72 4.5 Conceptual Model Time Estimate Results 74 4.6 Conclusions and Future Work 81 Chapter Five Testing of Interference Detection Method 83 5.1 Internal Testing – CEDAR Products 83

i Acknowledgements ii Table of contents iii List of figures vii List of tables xi Terminology xii Chapter 1: Introduction 1 1.1 Manufacturing SMEs: current challenges and issues 3 1.1.1 Packaging machinery manufacturers 4 1.2 Research aims and objectives 6 1.2.1 Aim of the research 6 1.2.2 Objectives 6 1.3 Overview of the thesis 8 Chapter 2: Product development: current state and issues in SMEs 12 2.1 Product development 13 2.1.1 Benefits of following a systematic product development process model 17 2.1.2 Types of design activities 18 2.2 Product development in SMEs 22 2.2.1 Important findings from the literature 24 2.3 Relevant techniques to improve product development 26 2.3.1 Standardisation 26 2.3.2 Modular design 26 2.3.3 Product platform development 26 2.3.4 Development of a product architecture 27 2.4 Application in SMEs 28 2.4.1 Challenges for SMEs 31

doi:10.1016/j.jtcvs.2004.03.022 I f surgical performance—often measured by postoperative outcome of initial hospital stay—is monitored at all, the most common means is by risk-adjusted annual or semiannual audit. Observed occurrence of outcome measures (eg, in-hospital death and complications) as a proportion of cases performed is compared with expected performance using, for example, the Society of Thoracic Surgeons’ regression equations or EuroSCORE, which account for many aspects of case mix. Sometimes observed (O) and expected (E) proportions are subtracted, sometimes divided (O/E ratio); sometimes confidence limits of these comparisons are provided, and occasionally P values are given. Is this periodic, widespread, but rather coarse monitoring of surgical performance sufficient?

a b s t r a c t This study examines whether a fish stock can be managed using cumulative sum (CUSUM) control charts if limited historical information is available for the fish stock. We use the term 'limited' in the sense of having a minimal number of historical observations of relevant stock indicators and their respective control means (or reference points) that are aligned with the objectives of fisheries management. In the present study, a Decision Interval Cumulative Sum (DI-CUSUM) control chart was used to monitor two indices from a simulated fishery; the recruitment indicator and the large fish indicator (LFI). The fishery was subsequently managed using a harvest control rule (HCR) that triggered only when a significant deviation in the indicator trend was detected by the DI-CUSUM. The HCR was constructed using methods adopted from engineering process control theory where the adjustment in total allowable catch was determined by estimating the size of the shift in the indicator time series. We found that monitoring a combined indicator of both recruitment and LFI was more successful in controlling the fishery irrespective of the initial state of the fish stock. We discuss how DI-CUSUM could be incorporated into the management process for data poor fisheries.

[This abstract is based on the author's abstract.] An application in real-time multivariate statistical process control for the automated power calibration of a family of high-energy power supplies is described. After a variance-stabilizing transfor..

[This abstract is based on the author's abstract.] A rank-based process control chart that is distribution-free and that does not require any parameter estimations is proposed. The nonparametric control chart is based on rank distribution of the c..

[This abstract is based on the author's abstract.] A method is presented for evaluating the performance of a manufacturing process that produces parts described by two critical characteristics. Process performance is described most often either throug..

Within the quality community, quality management is often pictured as and referred to as rapidly evolving and continuously learning by interaction with the surrounding world. In general, an ability ...

With the recent increase in the threat of biological terrorism as well as the continual risk of other diseases, the research in public health surveillance and disease monitoring has grown tremendously. There is an abundance of data available in all sorts of forms. Hospitals, federal and local governments, and industries are all collecting data and developing new methods to be used in the detection of anomalies. Many of these methods are developed, applied to a real data set, and incorporated into software. This research, however, takes a different view of the evaluation of these methods. We feel that there needs to be solid statistical evaluation of proposed methods no matter the intended area of application. Using proof-by-example does not seem reasonable as the sole evaluation criteria especially concerning methods that have the potential to have a great impact

With the rapid rise in the ubiquity and sophistication of Internet technology and the accompanying growth in the number of network attacks, network intrusion detection has become increasingly important. Anomaly-based network intrusion detection refers to finding exceptional or nonconforming patterns in network traffic data compared to normal behavior. Finding these anomalies has extensive applications in areas such as cyber security, credit card and insurance fraud detection, and military surveillance for enemy activities. Network Anomaly Detection: A Machine Learning Perspective presents machine learning techniques in depth to help you more effectively detect and counter network intrusion. In this book, youll learn about: Network anomalies and vulnerabilities at various layers The pros and cons of various machine learning techniques and algorithms A taxonomy of attacks based on their characteristics and behavior Feature selection algorithms How to assess the accuracy, performance, completeness, timeliness, stability, interoperability, reliability, and other dynamic aspects of a network anomaly detection system Practical tools for launching attacks, capturing packet or flow traffic, extracting features, detecting attacks, and evaluating detection performance Important unresolved issues and research challenges that need to be overcome to provide better protection for networks Examining numerous attacks in detail, the authors look at the tools that intruders use and show how to use this knowledge to protect networks. The book also provides material for hands-on development, so that you can code on a testbed to implement detection methods toward the development of your own intrusion detection system. It offers a thorough introduction to the state of the art in network anomaly detection using machine learning approaches and systems.

With the rapid development of sensor techniques, we often face the challenges of monitoring big data streams in modern quality control, which consist of massive series of real-time, continuously and sequentially ordered observations. For example, in manufacturing industries, hundreds or thousands of variables are observed during online production for quality insurance. Also, smart grid infrastructure needs to simultaneously monitor massive access points for intrusion and threat detection. As another example, an image sensing device continuously collects high-resolution images at high frequency for video surveillance and object movement tracking. Ideally, in those applications, it is preferable to detect assignable causes as early as possible, while maintaining a prespecified in-control Average Run Length (ARL).

With the growing of automation in manufacturing, process quality characteristics are being measured at higher rates and data are more likely to be autocorrelated. A widely used approach for statistical process monitoring in the case of autocorrelated data is the residual chart. This chart requires that a suitable model has been identified for the time series of process observations before residuals can be obtained. In this work, a new neural-based procedure, which is alleviated from the need for building a time series model, is introduced for quality control in the case of serially correlated data. In particular, the Elman's recurrent neural network is proposed for manufacturing process quality control. Performance comparisons between the neural-based algorithm and several control charts are also presented in the paper in order to validate the approach. Different magnitudes of the process mean shift, under the presence of various levels of autocorrelation, are considered. The simulation results indicate that the neural-based procedure may perform better than other control charting schemes in several instances for both small and large shifts. Given the simplicity of the proposed neural network and its adaptability, this approach is proved from simulation experiments to be a feasible alternative for quality monitoring in the case of autocorrelated process data.

With the development of the global economy and the ease of air transportation, extra emphasis has been placed on flight safety. There are precise specifications and procedures in the operation and maintenance of aircraft. Human errors and mechanical disorders are two key factors of flight safety. The maintenance personnel need to follow an outlined procedure to avoid human errors and ensure flight safety. Readability of aircraft maintenance technical orders can affect the quality and reliability of aircraft maintenance. To ensure the editing quality of technical orders, controlling/monitoring the number of unreadable sentences is important and necessary. In this study, the number of unreadable sentences found in a technical order was used as the measure of readability performance (RP) as well as a readability performance index was provided to evaluate whether the RP of individual readability characteristics of technical orders was adequate. Different readability characteristics make different grade of RP loss. Based on fuzzy multiple criteria decision-making (Fuzzy MCDM) approaches, we investigated the expert opinions to rank and calculate the weights of all readability characteristics. At the same time, we proposed the upper limits of unreadable sentences according the weights of individual readability characteristics. In this paper, the technical orders issued by Taiwan Aerospace Industrial Development Corporation was used as an example to evaluate the readability of the technical orders and total RP losses for individual readability characteristics. Finally, an improved way of editing quality for technical orders was recommended.

With the development of information technology, there is a need for computational works everywhere and every time. Thus, people should be able to carry out their heavy computations without having the burden of purchasing expensive hardware and software. Cloud computing is an attractive solution to such needs, but the high energy consumption of physical machines in a Cloud data center is a matter of great concern. Therefore, some of the low-loaded machines can be turned off or switched into low energy mode using server consolidation approaches. In this paper, a Stochastic Process-Based Dynamic Server Consolidation (SB-DSC) policy is developed to reduce the total cost of data centers while satisfying the required quality of service. A novel algorithm, which we call it Stochastic Process-Based BFD (SBBFD), is employed in SB-DSC policy to perform virtual machine placements over time. SBBFD overcomes most drawbacks of other algorithms proposed in the literature. The simulation results on real workload data show that SB-DSC leads to a noticeable reduction in total cost in terms of power consumption, SLA violations, number of mode switching and number of migrations.