IEEE International Conference on Networking, Sensing and Control, 2004
Based on an engineering model, a statistical process control (SPC) method for processes with mult... more Based on an engineering model, a statistical process control (SPC) method for processes with multiple stages is proposed in this paper. In phase I of the SPC analysis, a maximum likelihood estimation procedure is developed using EM algorithm. In phase II, the complex multistage monitoring problem is transferred to a simple multi-stream monitoring problem by applying group exponential weighted moving average charts to the one-step forecast errors of the model. Run length results show the efficiency of proposed charting method, and a heuristic run-rule-like approach is suggested to improve the traceability along the multiple stages. Two real multistage examples in hood manufacturing and workpiece assembly are presented to illustrate the efficiency of the proposed method.
ACM Transactions on Knowledge Discovery from Data, 2020
This article proposes a novel time-warped sparse non-negative factorization method for functional... more This article proposes a novel time-warped sparse non-negative factorization method for functional data analysis. The proposed method on the one hand guarantees the extracted basis functions and their coefficients to be positive and interpretable, and on the other hand is able to handle weakly correlated functions with different features. Furthermore, the method incorporates time warping into factorization and hence allows the extracted basis functions of different samples to have temporal deformations. An efficient framework of estimation algorithms is proposed based on a greedy variable selection approach. Numerical studies together with case studies on real-world data demonstrate the efficacy and applicability of the proposed methodology.
We would like to thank Dr Laura M. Sangalli for her work as well as many other researchers who co... more We would like to thank Dr Laura M. Sangalli for her work as well as many other researchers who contributed to solving the analysis problem for complexstructural data with complicated spatial or spatiotemporal dependencies. As Dr Laura M. Sangalli mentioned, the complex data dependencies pose a large obstacle to accurate data analysis, including prediction, regression, detection, and especially monitoring. Based on the examples presented by Dr Laura M. Sangalli in Figures 1 and 2 of her paper, the next step after regression is to monitor the complex data. For example, oceanographers might monitor the sea temperature profile at each buoy; and aircraft engineers might monitor the aerodynamic force on the surface of a shuttle winglet. To achieve the monitoring purpose, formulation and regression for the data to understand the model behind is usually the first step. But the anisotropy and non-stationarity of complex data make accurate regression difficult, let alone monitoring it. As an extension to Dr Laura M. Sangalli’s work, our discussion will focus on three areas which we believe are highly-related and complementary to her work. First, profile monitoring will be reviewed, which is regarded as the next step after estimation; second, other types of formulation for spatial and spatiotemporal dependencies will be summarized; last but not least, a possible future direction extending spatial data to a higher dimension, which forms a tensor, will also be explained.
Defect inspection is important in many industries, such as in the manufacturing and pharmaceutica... more Defect inspection is important in many industries, such as in the manufacturing and pharmaceutical industries. Existing methods usually use either low-resolution data, which are obtained from less precise measurements, or high-resolution data, which are obtained from more precise measurements, to estimate the number of defectives in a given amount of goods produced. In this study, a novel approach is proposed that combines the two types of data to construct tolerance intervals with a desired average coverage probability. A simulation study shows that the derived tolerance intervals can lead to better performance than a tolerance interval that is constructed based on only the low-resolution data. In addition, a real-data example shows that the tolerance interval based on only the low-resolution data is more conservative than the tolerance intervals based on both high-resolution and low-resolution data.
The regression control chart is an effective statistical process control (SPC) tool in situations... more The regression control chart is an effective statistical process control (SPC) tool in situations where the output characteristic of interest is affected by an external covariate. It has many applications, including monitoring and diagnosing multistage processes. Its basic purpose is to remove the influence of the covariate using regression adjustment and then apply a regular control scheme to the regression residual. In practice, the regression model relating the output and the covariate is rarely known and needs to be estimated. Little is known about the performance of the regression control chart when the true parameters are replaced with their estimates. In this paper, the run length performance of regression control charts with estimated parameters is studied. Two types of regression control charts are considered: a Shewhart control for regression residuals, and an exponentially weighted moving average (EWMA) control for regression residuals.
To save some space in the paper, some technical details and numerical results are presented in th... more To save some space in the paper, some technical details and numerical results are presented in this supplemental file. This file has four sections. In Section S1, the proofs of the theorems and propositions are provided. In Section S2, some extra simulation results of the four methods discussed in the paper are presented. In Section S3, some supportive simulation results of the real data analysis are presented. In Section S4, some extensions of the proposed diagnostic methods are discussed.
If it is the author's pre-published version, changes introduced as a result of publishing process... more If it is the author's pre-published version, changes introduced as a result of publishing processes such as copy-editing and formatting may not be reflected in this document. For a definitive version of this work, please refer to the published version.
Journal of the Chinese Institute of Industrial Engineers, 2003
ABSTRACT Most products and services today are the results of several process stages and steps. Wi... more ABSTRACT Most products and services today are the results of several process stages and steps. With the current emphasis in industry on improved quality, control charts are widely used in process monitoring. However, conventional statistical process control (SPC) techniques focus mostly on individual stages in a process and do not consider disseminating information throughout the multiple stages of the process. Such techniques are ineffective in analyzing multistage processes. The cause-selecting chart (CSC), based on output adjusted for the effect of the incoming quality is an effective SPC tool for analyzing multistage processes. This paper discusses several important problems associated with conventional CSCs. First, the model relating input and output measures often needs to be estimated in practice before the CSC is implemented. Little is known about the performance of the CSC when the model parameters are estimated. Second, the simple linear regression model widely discussed in the CSC is insufficient to capture the stochastic behavior of the output. In practice, the process disturbance in a multistage process can be autocorrelated. To deal with autocorrelated disturbance, the cumulative score (Cuscore) chart and the triggered Cuscore chart are proposed. Finally, extensions of the CSC to multivariate CSCs are discussed.
2009 6th International Conference on Service Systems and Service Management, 2009
... [14] Mason, RL, Tracy, ND, Young, JC, Decomposition of T2 for multivariate control chart int... more ... [14] Mason, RL, Tracy, ND, Young, JC, Decomposition of T2 for multivariate control chart interpretation, Journal of Quality Technology, V. 27 No.2, 1995, 99-108. ... [23] Acosta-Mejia, CA, Improved p charts to monitor process quality, IIE Transactions, 31(6), 1999, 509-516. ...
2014 IEEE International Conference on Automation Science and Engineering (CASE), 2014
3D Printing or Additive Manufacturing (AM), refers to a new class of technologies of making produ... more 3D Printing or Additive Manufacturing (AM), refers to a new class of technologies of making products directly from any three-dimensional digital models. Broader applications of AM require to lower the cost of AM machines. However, products fabricated by low-end machines suffer from the issue of low dimensional accuracy. In this paper, we intend to address this issue for Fused Deposition Modeling (FDM) process - one of mostly adopted AM technologies. Based on the FDM mechanism, we attribute the dimensional inaccuracy to two significant error sources that affect the shape of the product consecutively: (i) positioning error of the extruder and (ii) shape deformation induced by processing error including phase change and other variations occurred. We first adopt Kriging method to model the extruder positioning error, which is treated as design input error under a novel framework of modeling the product shape deformation. Experimental results and case studies demonstrate the effectiveness of the predictive modeling framework, which can be applied to compensate dimensional error of 3D printed products.
2006 IEEE International Conference on Systems, Man and Cybernetics, 2006
In recent years, there has been an increasing demand for quality control of multivariate dynamic ... more In recent years, there has been an increasing demand for quality control of multivariate dynamic systems. Conventional directionally invariant charts fail to make use of versatile shift patterns in a multivariate process and are sensitive to general failures only. Directionally variant charts, however, are designed for specific and constant shifts and are not suitable for processes with dynamic and unknown failures. This paper proposes an adaptive T2 scheme, which can successfully capture the unknown shift patterns of a multivariate system via an exponentially weighted moving average (EWMA) forecasting procedure. The adaptive scheme preserves the optimality of a directionally variant chart, while provides a scalable extension to Hotelling's T2 procedure. The smoothing parameter of the proposed scheme can be tuned for desired shift sizes. Significant improvement of sensitivity over an intended range is demonstrated by Monte Carlo simulation.
Supplier selection is an important part of supply chain management. Among the numerous methods th... more Supplier selection is an important part of supply chain management. Among the numerous methods that have been proposed, process capability index is considered to be the most effective technique for identifying quality parts. However, supplier selection should be carried out on the basis of quality and cost together. There is no easy tool available to evaluate the price and quality in an integrated manner. In this article, a new approach to supplier selection using capability index and price comparison (CPC) chart is presented. The CPC chart integrates the process capability and price information of multiple suppliers and presents them in a single chart. It provides a simple but effective method to consider quality and price simultaneously in the supplier selection process.
Monitoring the vibration behaviour of large rotating machinery is an effective way to reduce prod... more Monitoring the vibration behaviour of large rotating machinery is an effective way to reduce production losses and enhance safety, efficiency, reliability, availability and longevity in manufacturing processes. As large rotating machinery is increasingly employed in continuous operations at high speeds and with heavy loads, the vibration behaviour of the rotor systems is more and more complex. Focusing on the defects of different joint time-frequency representations, we present an adaptive time-frequency decomposition (ATFD) technique to describe transient vibration. This approach provides a precise interpretation of complex signals that consist of various time-frequency structures through decomposing it into paramedic, redundant and well-localised components in the time-frequency plane. Both computer simulation and an actual case show that this technique has high time-frequency resolution and no interference terms. The analysis results proved this approach could specify critical speed as well as acceleration rate accurately and is very effective in ensuring machinery passing through critical speed field safely.
Journal of the American Statistical Association, 2006
To detect and estimate nonconstant, time-varying mean shifts, statistical process control (SPC) t... more To detect and estimate nonconstant, time-varying mean shifts, statistical process control (SPC) tools, such as the cumulative score (Cuscore) and generalized likelihood ratio test (GLRT) charts, have recently been proposed. However, their efficiency is based on previous and exact knowledge of a reference pattern. In this article a reference-free Cuscore (RFCuscore) chart is proposed that can trace and detect dynamic mean changes quickly without knowing the reference pattern. In addition, a unified framework that contains most of the control charts is presented and applied for a theoretical comparison of the RFCuscore, Cuscore, GLRT, and CUSUM charts in detecting dynamic mean changes. Moreover, numerical simulations and a real example are used to illustrate and verify the results. Both theoretical analysis and numerical results show that the RFCuscore chart performs not only robustly, but also quickly in detecting both small and large dynamic mean changes.
This paper investigates how to adaptively predict the time-varying metrology delay that can reali... more This paper investigates how to adaptively predict the time-varying metrology delay that can realistically occur in the semiconductor manufacturing practice. In the presence of metrology delays, the expected asymptotic double exponentially weighted moving average (dEWMA) control output, by using the EWMA and recursive least squares prediction methods, is derived. It has been found that the relationships between the expected control output and target in both estimation methods are equivalent, and six cases are addressed. Within the context of time-varying metrology delay, a new time update scheme to the recursive least squares-linear trend (RLS-LT) controller, combined with zone tests and the moving average (MA) control chart, is proposed. Simulated single input-single output (SISO) run-to-run processes subject to two time-varying metrology delay scenarios are used to assess the effectiveness of the proposed controller.
Quick detection of unanticipated changes in a financial sequence is a critical problem for practi... more Quick detection of unanticipated changes in a financial sequence is a critical problem for practitioners in the finance industry. Based on refined logarithmic moment estimators for the four parameters of a stable distribution, this article presents a stable-distribution-based ...
International Journal of Six Sigma and Competitive Advantage, 2005
In Hong Kong, more than 500,000 container units per year are handled by small midstream cargo ope... more In Hong Kong, more than 500,000 container units per year are handled by small midstream cargo operators located at the Public Cargo Working Areas (PCWAs). Unfortunately, over 600 fall accidents were reported at PCWAs over the past ten years resulting in more than 45 deaths. This paper reports the results of a Six Sigma study to reduce fall accidents at PCWAs. Work procedures leading to falls of cargo handlers from cargo containers are described and the sigma levels for the rates of such accidents are determined. With help from the Logistics Cargo Supervisors Association and the Hong Kong Occupational Safety and Health Council (OSHC), we conducted focus group interviews and questionnaire surveys at the PCWAs. Data indicated that jumping from one container to another and resting and standing on a lifted container are the most hazardous procedures. The three most critical factors are identified. They are: 1 not concentrating at work 2 disregarding safety regulations 3 using worn-out hooks and slings. Detailed improvement actions to reduce the fall accident rates were identified. Control procedures were documented and passed to OSHC.
International Journal of Production Research, 2013
ABSTRACT To monitor covariance matrices, most of the existing control charts are based on some om... more ABSTRACT To monitor covariance matrices, most of the existing control charts are based on some omnibus test and thus usually are not powerful when one is interested in detecting shifts that occur in a small number of elements of the covariance matrix. A new multivariate exponentially weighted moving average control chart is developed for the monitor of the covariance matrices by integrating the classical [Inline formula]-norm-based test with a maximum-norm-based test. Numerical studies show that the new control chart affords more balanced performance under various shift directions than the existing ones and is thus an effective tool for multivariate SPC applications. The implementation of the proposed control chart is demonstrated with an example from the health care industry.
IEEE International Conference on Networking, Sensing and Control, 2004
Based on an engineering model, a statistical process control (SPC) method for processes with mult... more Based on an engineering model, a statistical process control (SPC) method for processes with multiple stages is proposed in this paper. In phase I of the SPC analysis, a maximum likelihood estimation procedure is developed using EM algorithm. In phase II, the complex multistage monitoring problem is transferred to a simple multi-stream monitoring problem by applying group exponential weighted moving average charts to the one-step forecast errors of the model. Run length results show the efficiency of proposed charting method, and a heuristic run-rule-like approach is suggested to improve the traceability along the multiple stages. Two real multistage examples in hood manufacturing and workpiece assembly are presented to illustrate the efficiency of the proposed method.
ACM Transactions on Knowledge Discovery from Data, 2020
This article proposes a novel time-warped sparse non-negative factorization method for functional... more This article proposes a novel time-warped sparse non-negative factorization method for functional data analysis. The proposed method on the one hand guarantees the extracted basis functions and their coefficients to be positive and interpretable, and on the other hand is able to handle weakly correlated functions with different features. Furthermore, the method incorporates time warping into factorization and hence allows the extracted basis functions of different samples to have temporal deformations. An efficient framework of estimation algorithms is proposed based on a greedy variable selection approach. Numerical studies together with case studies on real-world data demonstrate the efficacy and applicability of the proposed methodology.
We would like to thank Dr Laura M. Sangalli for her work as well as many other researchers who co... more We would like to thank Dr Laura M. Sangalli for her work as well as many other researchers who contributed to solving the analysis problem for complexstructural data with complicated spatial or spatiotemporal dependencies. As Dr Laura M. Sangalli mentioned, the complex data dependencies pose a large obstacle to accurate data analysis, including prediction, regression, detection, and especially monitoring. Based on the examples presented by Dr Laura M. Sangalli in Figures 1 and 2 of her paper, the next step after regression is to monitor the complex data. For example, oceanographers might monitor the sea temperature profile at each buoy; and aircraft engineers might monitor the aerodynamic force on the surface of a shuttle winglet. To achieve the monitoring purpose, formulation and regression for the data to understand the model behind is usually the first step. But the anisotropy and non-stationarity of complex data make accurate regression difficult, let alone monitoring it. As an extension to Dr Laura M. Sangalli’s work, our discussion will focus on three areas which we believe are highly-related and complementary to her work. First, profile monitoring will be reviewed, which is regarded as the next step after estimation; second, other types of formulation for spatial and spatiotemporal dependencies will be summarized; last but not least, a possible future direction extending spatial data to a higher dimension, which forms a tensor, will also be explained.
Defect inspection is important in many industries, such as in the manufacturing and pharmaceutica... more Defect inspection is important in many industries, such as in the manufacturing and pharmaceutical industries. Existing methods usually use either low-resolution data, which are obtained from less precise measurements, or high-resolution data, which are obtained from more precise measurements, to estimate the number of defectives in a given amount of goods produced. In this study, a novel approach is proposed that combines the two types of data to construct tolerance intervals with a desired average coverage probability. A simulation study shows that the derived tolerance intervals can lead to better performance than a tolerance interval that is constructed based on only the low-resolution data. In addition, a real-data example shows that the tolerance interval based on only the low-resolution data is more conservative than the tolerance intervals based on both high-resolution and low-resolution data.
The regression control chart is an effective statistical process control (SPC) tool in situations... more The regression control chart is an effective statistical process control (SPC) tool in situations where the output characteristic of interest is affected by an external covariate. It has many applications, including monitoring and diagnosing multistage processes. Its basic purpose is to remove the influence of the covariate using regression adjustment and then apply a regular control scheme to the regression residual. In practice, the regression model relating the output and the covariate is rarely known and needs to be estimated. Little is known about the performance of the regression control chart when the true parameters are replaced with their estimates. In this paper, the run length performance of regression control charts with estimated parameters is studied. Two types of regression control charts are considered: a Shewhart control for regression residuals, and an exponentially weighted moving average (EWMA) control for regression residuals.
To save some space in the paper, some technical details and numerical results are presented in th... more To save some space in the paper, some technical details and numerical results are presented in this supplemental file. This file has four sections. In Section S1, the proofs of the theorems and propositions are provided. In Section S2, some extra simulation results of the four methods discussed in the paper are presented. In Section S3, some supportive simulation results of the real data analysis are presented. In Section S4, some extensions of the proposed diagnostic methods are discussed.
If it is the author's pre-published version, changes introduced as a result of publishing process... more If it is the author's pre-published version, changes introduced as a result of publishing processes such as copy-editing and formatting may not be reflected in this document. For a definitive version of this work, please refer to the published version.
Journal of the Chinese Institute of Industrial Engineers, 2003
ABSTRACT Most products and services today are the results of several process stages and steps. Wi... more ABSTRACT Most products and services today are the results of several process stages and steps. With the current emphasis in industry on improved quality, control charts are widely used in process monitoring. However, conventional statistical process control (SPC) techniques focus mostly on individual stages in a process and do not consider disseminating information throughout the multiple stages of the process. Such techniques are ineffective in analyzing multistage processes. The cause-selecting chart (CSC), based on output adjusted for the effect of the incoming quality is an effective SPC tool for analyzing multistage processes. This paper discusses several important problems associated with conventional CSCs. First, the model relating input and output measures often needs to be estimated in practice before the CSC is implemented. Little is known about the performance of the CSC when the model parameters are estimated. Second, the simple linear regression model widely discussed in the CSC is insufficient to capture the stochastic behavior of the output. In practice, the process disturbance in a multistage process can be autocorrelated. To deal with autocorrelated disturbance, the cumulative score (Cuscore) chart and the triggered Cuscore chart are proposed. Finally, extensions of the CSC to multivariate CSCs are discussed.
2009 6th International Conference on Service Systems and Service Management, 2009
... [14] Mason, RL, Tracy, ND, Young, JC, Decomposition of T2 for multivariate control chart int... more ... [14] Mason, RL, Tracy, ND, Young, JC, Decomposition of T2 for multivariate control chart interpretation, Journal of Quality Technology, V. 27 No.2, 1995, 99-108. ... [23] Acosta-Mejia, CA, Improved p charts to monitor process quality, IIE Transactions, 31(6), 1999, 509-516. ...
2014 IEEE International Conference on Automation Science and Engineering (CASE), 2014
3D Printing or Additive Manufacturing (AM), refers to a new class of technologies of making produ... more 3D Printing or Additive Manufacturing (AM), refers to a new class of technologies of making products directly from any three-dimensional digital models. Broader applications of AM require to lower the cost of AM machines. However, products fabricated by low-end machines suffer from the issue of low dimensional accuracy. In this paper, we intend to address this issue for Fused Deposition Modeling (FDM) process - one of mostly adopted AM technologies. Based on the FDM mechanism, we attribute the dimensional inaccuracy to two significant error sources that affect the shape of the product consecutively: (i) positioning error of the extruder and (ii) shape deformation induced by processing error including phase change and other variations occurred. We first adopt Kriging method to model the extruder positioning error, which is treated as design input error under a novel framework of modeling the product shape deformation. Experimental results and case studies demonstrate the effectiveness of the predictive modeling framework, which can be applied to compensate dimensional error of 3D printed products.
2006 IEEE International Conference on Systems, Man and Cybernetics, 2006
In recent years, there has been an increasing demand for quality control of multivariate dynamic ... more In recent years, there has been an increasing demand for quality control of multivariate dynamic systems. Conventional directionally invariant charts fail to make use of versatile shift patterns in a multivariate process and are sensitive to general failures only. Directionally variant charts, however, are designed for specific and constant shifts and are not suitable for processes with dynamic and unknown failures. This paper proposes an adaptive T2 scheme, which can successfully capture the unknown shift patterns of a multivariate system via an exponentially weighted moving average (EWMA) forecasting procedure. The adaptive scheme preserves the optimality of a directionally variant chart, while provides a scalable extension to Hotelling's T2 procedure. The smoothing parameter of the proposed scheme can be tuned for desired shift sizes. Significant improvement of sensitivity over an intended range is demonstrated by Monte Carlo simulation.
Supplier selection is an important part of supply chain management. Among the numerous methods th... more Supplier selection is an important part of supply chain management. Among the numerous methods that have been proposed, process capability index is considered to be the most effective technique for identifying quality parts. However, supplier selection should be carried out on the basis of quality and cost together. There is no easy tool available to evaluate the price and quality in an integrated manner. In this article, a new approach to supplier selection using capability index and price comparison (CPC) chart is presented. The CPC chart integrates the process capability and price information of multiple suppliers and presents them in a single chart. It provides a simple but effective method to consider quality and price simultaneously in the supplier selection process.
Monitoring the vibration behaviour of large rotating machinery is an effective way to reduce prod... more Monitoring the vibration behaviour of large rotating machinery is an effective way to reduce production losses and enhance safety, efficiency, reliability, availability and longevity in manufacturing processes. As large rotating machinery is increasingly employed in continuous operations at high speeds and with heavy loads, the vibration behaviour of the rotor systems is more and more complex. Focusing on the defects of different joint time-frequency representations, we present an adaptive time-frequency decomposition (ATFD) technique to describe transient vibration. This approach provides a precise interpretation of complex signals that consist of various time-frequency structures through decomposing it into paramedic, redundant and well-localised components in the time-frequency plane. Both computer simulation and an actual case show that this technique has high time-frequency resolution and no interference terms. The analysis results proved this approach could specify critical speed as well as acceleration rate accurately and is very effective in ensuring machinery passing through critical speed field safely.
Journal of the American Statistical Association, 2006
To detect and estimate nonconstant, time-varying mean shifts, statistical process control (SPC) t... more To detect and estimate nonconstant, time-varying mean shifts, statistical process control (SPC) tools, such as the cumulative score (Cuscore) and generalized likelihood ratio test (GLRT) charts, have recently been proposed. However, their efficiency is based on previous and exact knowledge of a reference pattern. In this article a reference-free Cuscore (RFCuscore) chart is proposed that can trace and detect dynamic mean changes quickly without knowing the reference pattern. In addition, a unified framework that contains most of the control charts is presented and applied for a theoretical comparison of the RFCuscore, Cuscore, GLRT, and CUSUM charts in detecting dynamic mean changes. Moreover, numerical simulations and a real example are used to illustrate and verify the results. Both theoretical analysis and numerical results show that the RFCuscore chart performs not only robustly, but also quickly in detecting both small and large dynamic mean changes.
This paper investigates how to adaptively predict the time-varying metrology delay that can reali... more This paper investigates how to adaptively predict the time-varying metrology delay that can realistically occur in the semiconductor manufacturing practice. In the presence of metrology delays, the expected asymptotic double exponentially weighted moving average (dEWMA) control output, by using the EWMA and recursive least squares prediction methods, is derived. It has been found that the relationships between the expected control output and target in both estimation methods are equivalent, and six cases are addressed. Within the context of time-varying metrology delay, a new time update scheme to the recursive least squares-linear trend (RLS-LT) controller, combined with zone tests and the moving average (MA) control chart, is proposed. Simulated single input-single output (SISO) run-to-run processes subject to two time-varying metrology delay scenarios are used to assess the effectiveness of the proposed controller.
Quick detection of unanticipated changes in a financial sequence is a critical problem for practi... more Quick detection of unanticipated changes in a financial sequence is a critical problem for practitioners in the finance industry. Based on refined logarithmic moment estimators for the four parameters of a stable distribution, this article presents a stable-distribution-based ...
International Journal of Six Sigma and Competitive Advantage, 2005
In Hong Kong, more than 500,000 container units per year are handled by small midstream cargo ope... more In Hong Kong, more than 500,000 container units per year are handled by small midstream cargo operators located at the Public Cargo Working Areas (PCWAs). Unfortunately, over 600 fall accidents were reported at PCWAs over the past ten years resulting in more than 45 deaths. This paper reports the results of a Six Sigma study to reduce fall accidents at PCWAs. Work procedures leading to falls of cargo handlers from cargo containers are described and the sigma levels for the rates of such accidents are determined. With help from the Logistics Cargo Supervisors Association and the Hong Kong Occupational Safety and Health Council (OSHC), we conducted focus group interviews and questionnaire surveys at the PCWAs. Data indicated that jumping from one container to another and resting and standing on a lifted container are the most hazardous procedures. The three most critical factors are identified. They are: 1 not concentrating at work 2 disregarding safety regulations 3 using worn-out hooks and slings. Detailed improvement actions to reduce the fall accident rates were identified. Control procedures were documented and passed to OSHC.
International Journal of Production Research, 2013
ABSTRACT To monitor covariance matrices, most of the existing control charts are based on some om... more ABSTRACT To monitor covariance matrices, most of the existing control charts are based on some omnibus test and thus usually are not powerful when one is interested in detecting shifts that occur in a small number of elements of the covariance matrix. A new multivariate exponentially weighted moving average control chart is developed for the monitor of the covariance matrices by integrating the classical [Inline formula]-norm-based test with a maximum-norm-based test. Numerical studies show that the new control chart affords more balanced performance under various shift directions than the existing ones and is thus an effective tool for multivariate SPC applications. The implementation of the proposed control chart is demonstrated with an example from the health care industry.
Uploads
Papers by Fugee Tsung