Search

While semiconductors are being used in an increasing number of products, semiconductor manufacturers continually look for ways to make their processes more efficient. This paper will focus on an issue in the manufacturing process called the class-constrained lot-to-order matching problem (CLOMP), where individual lots of microprocessors are matched to customer orders, while seeking to optimize multiple objectives. Due to its complexity, the problem is decomposed into two stages - the first identifies which customer orders to fill while the second assigns specific lots to the chosen orders. We design an experiment with four first-stage sorting rules, four second-stage heuristics and two production cases. Based on our simulation results, this paper will recommend the first-stage sorting rule and second-stage heuristic which attain the best results with regards to our measures of effectiveness. [ABSTRACT FROM AUTHOR]

Many companies have set Parts per Million (PPM) or Parts per Million Opportunities (PPMO) goals in their quest for continuous improvement. The time-between-events (TBE) CUSUM has been suggested for monitoring the number of good units or the number of opportunities that occurred between discoveries of consecutive bad units. We focus on the robustness of the TBE CUSUM. Robustness, in this case, refers to sensitivity of the procedure to make the proper decisions regarding a shift in the mean defect rate when, in fact, the time between events is not exponential. We examine and report average run lengths (ARLs) under both a Weibull and a lognormal time between events distribution. Our results indicate that the TBE CUSUM is extremely robust for a wide variety of parameter values for both the Weibull and lognormal distributions. The implications of these results in practice imply that users of the TBE CUSUM procedure need not be concerned about departures from the exponential TBE distribution. Practical implementation of the TBE CUSUM procedure is also discussed. [ABSTRACT FROM AUTHOR]

Variation exists in all processes. Significant work has been done to identify and remove sources of variation in manufacturing processes resulting in large returns for companies. However, business process optimization is an area that has a large potential return for a company. Business processes can be difficult to optimize due to the nature of the output variables associated with them. Business processes tend to have output variables that are binary, nominal or ordinal. Examples of these types of output include whether a particular event occurred, a customer's color preference for a new product and survey questions that assess the extent of the survey respondent's agreement with a particular statement. Output variables that are binary, nominal or ordinal cannot be modeled using ordinary least-squares regression. Logistic regression is a method used to model data where the output is binary, nominal or ordinal. This article provides a review of logistic regression and demonstrates its use in modeling data from a business process involving customer feedback. Copyright © 2006 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

An important question within industrial statistics is how to find operating conditions that achieve some goal for the mean of a characteristic of interest while simultaneously minimizing the characteristic's process variance. Often, people refer to this kind of situation as the robust parameter design problem. The robust parameter design literature is rich with ways to create separate models for the mean and variance from this type of experiment. Many times time and/or cost constraints force certain factors of interest to be much more difficult to change than others. An appropriate approach to such an experiment restricts the randomization, which leads to a split-plot structure. The paper modifies the central composite design to allow the estimation of separate models for the characteristic's mean and variances under a split-plot structure. The paper goes on to discuss an appropriate analysis of the experimental results. It illustrates the methodology with an industrial experiment involving a chemical vapour deposition process for the manufacture of silicon wafers. The methodology was used to achieve a silicon layer thickness value of 485 Å while minimizing the process variation. [ABSTRACT FROM AUTHOR]

Response surface methodology (RSM) is a collection of statistical design and numerical optimization techniques traditionally used to optimize industrial processes. In this paper, we demonstrate that the methodology can be successfully applied to the domain of networking. Specifically, we obtain increased throughput with a significant decrease in delay in a ns-2 simulation model of a mobile ad hoc network (MANET) by using RSM to optimize protocol interaction found by factor screening. Whether the experimentation is with a stochastic simulation model or a physical system, such as a MANET or a wireless sensor network test-bed, RSM provides a general and practical methodology to screen factors and robustly and jointly optimize responses. [ABSTRACT FROM AUTHOR]

Response surface methodology (RSM) is widely used for optimizing manufacturing processes and product designs. Most applications of RSM involve several response variables. In a typical RSM study, the experimenter will build an empirical model such as the second-order model to each response and use these models to determine settings on the design variables that produce optimal or at least acceptable values for the responses. In most multiple-response RSM problems, the experimenter fits a model to each response using ordinary least squares (OLS). This article illustrates another estimation technique useful in multiple-response RSM problems, seemingly unrelated regressions (SUR). This technique can be very useful when response variables in a multiple-response RSM problem are correlated. Essentially, SUR produces more precise estimates of the model parameters than OLS when responses are correlated. This improved precision of estimation can lead to a more precise estimate of the optimum operating conditions on the process. [ABSTRACT FROM AUTHOR]

In most statistical process control (SPC) applications, it is assumed that the quality of a process or product can be adequately represented by the distribution of a univariate quality characteristic or by the general multivariate distribution of a vector consisting of several correlated quality characteristics. In many practical situations, however, the quality of a process or product is better characterized and summarized by a relationship between a response variable and one or more explanatory variables. Thus, at each sampling stage, one observes a collection of data points that can be represented by a curve (or profile). In some calibration applications, the profile can be represented adequately by a simple straight-line model, while in other applications, more complicated models are needed. In this expository paper, we discuss some of the general issues involved in using control charts to monitor such process- and product-quality profiles and review the SPC literature on the topic. We relate this application to functional data analysis and review applications involving linear profiles, nonlinear profiles, and the use of splines and wavelets. We strongly encourage research in profile monitoring and provide some research ideas. [ABSTRACT FROM AUTHOR]

Non-normal responses are common in many industrial experiments. When there are factors whose levels are difficult and/or costly to control, an experiment is typically run within a split-plot context. Industrial split-plot experiments have received a great deal of attention in the literature in the normal response, linear model setting. Generalized linear models have been proposed for the analysis of completely randomized designs when the response is non-normal. When one uses a completely randomized design, there is an implication that responses are independent. For split-plot experimentation there is an assumption that responses within a given whole plot are correlated. To account for this correlation, generalized linear mixed models (GLMM) can be used for analysis. Generalized linear mixed models fall into two categories: population-averaged models and batch-specific models. We give an expository treatment of GLMM and discuss the differences in interpretations of the population-averaged and batch-specific models using an example from film manufacturing. [ABSTRACT FROM AUTHOR]

Variance dispersion graphs are useful tools for evaluating various types of designs, including mixture and mixture-process designs. They allow an experimenter to see patterns of scaled prediction variances (SPV) throughout a design space. We introduce a complementary fraction of design space (FDS) plot that provides additional information on the distribution of the SPV throughout a design space. These plots display the fraction of design space where the SPV is less than or equal to specific values. The FDS plots for combined mixture-process experiments also show which of the two types of variables has more influence on the SPV. The FDS plots for mixture and mixture-process experiments are developed and then demonstrated with several examples. [ABSTRACT FROM AUTHOR]

In this paper we demonstrate the use of control charts based on orthogonal contrasts for a semiconductor manufacturing process. The statistical performance of individuals control charts based on these contrasts and their exponentially weighted moving average counterparts is compared to that of the traditional approach. Contrast-based charts have superior detection capabilities than the traditional approach when targeted at specific process disturbances. Copyright © 2001 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Feedforward control is a particular form of engineering process control. If a known input variable zt can be measured and appropriate relationships built between the input variable, a compensatory variable xt, and the desired output yt, then a feedforward control system can be developed. Using feedforward control, we show under both sudden jump shifts and trends in the mean of the input and compensatory variables that the use of statistical process monitoring tools, such as the exponentially-weighted moving average and the cumulative sum (CUSUM), significantly reduces variability in both the output variable and the controller relative to the use of feedforward control alone. Copyright © 2000 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Multivariate statistical process control (SPC) procedures are useful in cases where several process variables are monitored simultaneously. A significant disadvantage of these techniques is that the time required to detect a process shift increases with the number of variables being monitored. We show how the shift detection capability of one popular multivariate SPC scheme, the multivariate analogue of the exponentially weighted moving average control chart, can be significantly improved by transforming the original process variables to a lower-dimensional subspace through the use of a U-transformation. Copyright © 1999 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

This paper presents an application of the sample autocorrelation function to statistical process control where the process data are serially correlated. Two innovative control charts are illustrated: the sample autocorrelation control chart and the group autocorrelation control chart. The important feature is that these control charts will detect shifts in the autocorrelative structure as well as shifts in the mean of the process. The sample autocorrelation function is typically used to identify an appropriate ARIMA model for a time series. The sample autocorrelation function may also be used as the basis of control charts to detect process upsets. Two unique features distingush this application of the sample autocorrelation function to statistical process control. First, the sample autocorrelations are exponentially smoothed estimates. This allows the user to control the sensitivity of the sample autocorrelation control chart. Secondly, the sample autocorrelation control chart is applied to a continuous stream of data-rather than to a static set of data that has been used to fit an ARIMA model. [ABSTRACT FROM AUTHOR]

Statistical process control methods are usually applied in an environment when periodic sampling and rational subgrouping of process output is appropriate. The resulting summary statistics can be graphically displayed and analysed using either traditional Shewhart control charts or other charts such as those based on the cumulative sum. This article presents an alternative. approach, based on time series analysis of all the real-time process data. The time series approach is employed because the sequence of process observations may not be statistically independent. The autocorrelative structure in the data may be captured using an ARIMA model, and the residuals from this model are shown to be an effective input signal for a variety of statistical process control procedures. [ABSTRACT FROM AUTHOR]

Most industrial processes are characterized by a system of several variables, all of which are subject to drifts, disturbances, and assignable causes of variation. In the chemical and process industries, there are often inertial forces arising from raw material streams, reactors and tanks that introduce serial correlation over time into these variables. This autocorrelation can have a profound impact on the effectiveness of the statistical monitoring methods used for such processes. This paper reviews some of the available methodology for multivariate process monitoring and shows the effectiveness of principal components in this context. An application of the principal components approach with con-elated observation vectors is presented. The effectiveness of this procedure to indicate process upsets is discussed. [ABSTRACT FROM AUTHOR]

This paper demonstrates the use of principal components in conjunction with the multivariate exponentially-weighted moving average (MEWMA) control procedure for process monitoring. It is demonstrated that the number of variables to be monitored is reduced through this approach, and that the average run length to detect process shifts or upsets is substantially reduced as well. The performance of the MEWMA applied 10 all the variables may he related to the MEWMA control chart that uses principal components through the non-centrality parameter. An average run length table demonstrates the advantages of the principal components MEWMA over the procedure that uses all of the variables, An illustrative example is provided. [ABSTRACT FROM AUTHOR]

An expected cost model for the fraction defective control chart for the case where there are several out-of-control states is developed. A procedure is presented for determining the sample size, control limit or critical region, and interval between samples which minimizes the expected cost of control per unit of product. The expected cost of control consists of the cost of sampling, the cost of investigating and possibly correcting the process when an out-of-control condition is indicated, and the cost of producing defective product. Direct search techniques are used to optimize the expected cost function. Numerical examples and sensitivity analysis of the model are presented. [ABSTRACT FROM AUTHOR]

A new model is presented for the optimal design of X charts utilized for the statistical monitoring of processes where production runs have a finite duration. The proposed model considers the effect of the setup operation on the chart design. The model contains both Duncan's model and a model due to Ladany as particular cases, yet it allows the user to consider more realistic production environments. Two types of finite-length production process are considered: a repetitive manufacturing process and a job-shop process. New relationships between the length of the production run, the power of the chart and the nature of the process setup are found by numerically analyzing the behavior of the model. [ABSTRACT FROM AUTHOR]

Adaptive sampling is an enhancement to a classical statistical process control scheme in which the time between samples from the process is varied based on the available information. Unlike the classical Shewhart control schemes, the performance of an adaptive sampling scheme for detecting a process which is off-target before the first observation is selected can differ from the steady-state performance, which is the performance of the scheme for detecting a process shift after the control scheme has been operating for some time. Often steady-state performance is the more important criterion for evaluating statistical process control schemes. Previous work has considered initial performance and steady-state performance, but the charts examined were optimized for initial performance. The development of charts optimized for steady-state performance and a comparison to previously developed schemes are the topics of this paper. Tables providing performance information on adaptive schemes are included. [ABSTRACT FROM AUTHOR]

Quality and productivity improvement has become an essential element of the overall strategic plan for most organizations. This has sparked renewed interest in statistical methods for quality improvement. This paper reviews some recent developments in statistical methodology that have application in product and process improvement, concentrating on statistical process control and design of experiments. Some directions for future research are identified. [ABSTRACT FROM AUTHOR]

Suggested modifications to Q charts are true enhancements from a run length distribution point of view. We would like to clarify and respond to certain statements made by Dr. Quesenberry. [ABSTRACT FROM AUTHOR]

The author presents insights on the importance of lean six sigma for technological innovations. He mentions the two kinds of innovation, breakthrough innovation and incremental innovation. He cites ways of promoting lean six sigma projects, including posting a descriptive summary of the project and the key results on an internal company website and participating in internal technical meetings and conferences that the organization conducts.

The first two tables in Reference 1 are incorrect. Corrected tables are provided below. The errors do not affect the conclusions or the discussion in that paper. [ABSTRACT FROM AUTHOR]