Chau-Chen Torng

The Performance of Double Sampling Control Charts Under Non Normality

In this study, the performances of double sampling (DS) chart under non normality are presented and compared with Shewhart chart and VP chart. The comparison result shows the DS chart has performance as good as the VP chart, but it is more effective than the Shewhart chart in detecting small process mean shifts.

A Design of s Control Charts with a Combined Double Sampling and Variable Sampling Interval Scheme

This study extended the idea of Carot et al. (2002) to combine double sampling and variable sampling interval s charts (DSVSI s chart) for improving efficiency in detection of small standard deviation shifts. The performance of DSVSI s charts in detecting small shifts was measured and compared with double sampling, variable sampling intervals, EWMA, and Cusum control charts. This comparative study presented found the DSVSI s chart to be efficient in detecting small shifts.

ARL Performance of the Tukey's Control Chart

A Tukeys control chart was designed to monitor single observation data. Its easy control-limits setting and simple statistical concept are the most important advantages. In this study, we setup the Tukeys control chart based on known probability distribution and construct its average run length calculator. ARL performance of the Tukeys control chart is evaluated under a normal distribution and non normal distributions, respectively. The comparison of ARL performance reveals that Tukeys control chart is not sensitive to shift detection when the process heavily violates the normal assumption. The number of observations needed for Tukeys control chart setup is less than Shewharts control chart. Tukeys control chart is a better choice for the process mean monitoring.

A Note on “Modified Tukey's Control Chart”

Tercero-Gomez et al. [Communications in Statistics - Simulation and Computation, 2012, 41(9), pp. 1566–1579] modified the Tukeys chart to improve the insensitivity of signaling mean shifts when the population is skewly distributed. They examine the efficiency for several skewed populations with the modified Tukeys chart (MTCC). The parameters of the MTCC were intuitively calibrated to fit several gamma distributions without following a specific mathematical procedure. Moreover, they performed no optimization for specific distributions. This study optimizes the MTCC with a statistical model for each skewed population and reexamines its efficiency. A numerical comparison shows that the optimal MTCC is sensitive to detect the shift of the means.

AN ECONOMIC DESIGN OF TUKEY'S CONTROL CHART

ABSTRACT The significant features of Tukeys control chart include characteristics such as: easy setup, use of single observation to monitor process, and suitability to monitor destructive testing processes. The control limit width and sampling interval should be determined before Tukeys control chart can be used. In this study, Duncans cost function is modified to construct an economic design model of Tukeys chart and to attain optimal design. Furthermore, methods of calculating error probabilities of Tukeys chart under normal conditions are constructed and applied in the economic design model. A real-world example of IC packaging illustrates the application of Tukeys chart to economics. The results of sensitivity analyses demonstrate that shorter in-control time or larger process variation cost incurs higher control cost.

Prioritization of product design tasks using QFD, TRIZ and DSM

Quality Function Deployment (QFD) is widely used application utilized in many fields such as product design, manufacturing process design, and project development. It can convert customer requirements into a firms engineering characteristics which results to achievements showing great success and House of Quality (HOQ) is the manipulation tool for QFD. However, the consideration of correlation among engineering characteristics is often ignored in most of QFD application related researches. This usually affects the implementation sequence of the project tasks and results to delay or queuing in product design/project development. This paper presents a methodology based on Theory of Inventive Problems Solving (TRIZ), Design Structure Matrix (DSM) and the absolute importance of engineering characteristics in HOQ which deals with the weakness mentioned. An example is illustrated to evidence the proposed methodology.

An economic-statistical design of the joint and S control charts with double sampling and variable sampling intervals

A combination of double sampling and variable sampling interval (DSVSI) charts can increase the efficiency of signaling small shifts . This study proposes an economic-statistical model for designing joint DSVSI and s charts. The proposed approach uses the economic model developed by Duncan in 1956. Research results indicate that the economic-statistical design of the joint DSVSI and s charts outperformed traditional economic design approach in detection abilities of process shifts; variable cost (a2), expected costs resulting from defective products when the process was in control (C0), and the expected number of assignable causes (λ) have the major influence on cost E(C′); deviations (δ, γ) significantly influenced cost E(C′); and statistical limitations () do not substantially affect cost E(C′).

Lean Production Implement Model for Aerospace Manufacturing Suppliers

Aerospace industry is a value-added and technology integrated industry. The aerospace manufacturing suppliers have to coordinate the demand of quality, delivery, cost and flexibility of aircraft to achieve customer satisfaction. This study integrates the lean concepts and summaries the Lean Cycle for lean implement and practice, and also develops the Lean Production Implement Model to strengthen the competitiveness of the manufacturing suppliers in the aerospace market.

Using Fuzzy Analytic Hierarchy Process to Construct Green Suppliers Assessment Criteria and Inspection Exemption Guidelines.

Since the announcement of Restriction of Hazardous Substances Directive (RoHS) in 2003, manufacturers of electrical and electronic equipments are restricted of containing any harmful substances such as: Pb, Cd, Hg, Cr6+, PBB and PBDE, in their products importing to European market, starting from July 1, 2006. How to avoid the use of materials containing hazardous substances has become a major issue in the electrical and electronic industry. The purpose of this research is to establish green supplier’s assessment criteria through a modified Delphi method, and to determine the weights of criteria by the Fuzzy Analytic Hierarchy Process. A sequential sampling plan is proposed as inspection exemption guideline based on the MIL-STD-1916 procedure. This could save 12.13% in material inspecting time for the case company.

Bottleneck Monitoring and Resource Allocation of Aerospace Sheet Metal Shop

Aerospace industry is a value-added and technology integrated industry. Sheet metal parts are important portion of aerostructures. The sheet metal production line combines five different routing types together. This study introduces an indicator to identify and monitor bottleneck among work centers, then develops an algorithm to reallocate the resources of sheet metal shop. It helps to construct a balance production line and satisfy the dynamic demand of aerospace market.