Robert G. Batson

Control Charts for Monitoring Field Failure Data

One responsibility of the reliability engineer is to monitor failure trends for fielded units to confirm that pre-production life testing results remain valid. This research suggests an approach that is computationally simple and can be used with a small number of failures per observation period. The approach is based on converting failure time data from fielded units to normal distribution data, using simple logarithmic or power transformations. Appropriate normalizing transformations for the classic life distributions (exponential, lognormal, and Weibull) are identified from the literature. Samples of size 500 field failure times are generated for seven different lifetime distributions (normal, lognormal, exponential, and four Weibulls of various shapes). Various control charts are then tested under three sampling schemes (individual, fixed, and random) and three system reliability degradations (large step, small step, and linear decrease in mean time between failures (MTBF)). The results of these tests are converted to performance measures of time to first out-of-control signal and persistence of signal after out-of-control status begins. Three of the well-known Western Electric sensitizing rules are used to recognize the assignable cause signals. Based on this testing, the ―X-chart with fixed sample size is the best overall for field failure monitoring, although the individual chart was better for the transformed exponential and another highly-skewed Weibull. As expected, the linear decrease in MTBF is the most difficult change for any of the charts to detect. Copyright

Constrained Multivariate Loss Function Minimization

Modifying Taguchis one-dimensional quadratic loss function to a multidimensional form has attracted many researchers because a more accurate estimate of total loss due to variation from multiple quality targets would result. This article includes a..

Characteristics of R&D management which influence information needs

The information needs of mid- and upper-level R&D managers are identified by reviewing their environment, functions, and communication patterns from an information systems perspective. The information problems in R&D, such as inability to respond rapidly and failure of analytical support methods, are seen as symptoms of the mismatch of information needs and the information systems employed. The new technology of decision support systems is discussed as the ideal solution approach to R&D information management and decision-making. A comprehensive list of candidate data bases is provided.

A survey of best practices in automotive supplier development

This article is intended to survey and synthesise best practices in supplier development in the US automotive industry. The supplier development literature up through the early 1990s consisted mostly of case studies. Since then, enough has been determined through industry surveys and academic research to enable the identification here of success factors (the prerequisites) and best practices – process and methods – used in todays automotive supplier development efforts. The conclusions of our research are the following: supplier development is a widespread and well-justified business practice among automotive OEMs today, but not among lower tiers; there is a consensus that supplier development improves supplier performance, but most companies limit development efforts to suppliers who are underperforming but with which some extended relationship exists; as the supplier relationship progresses towards partnership, OEMs are more willing to engage in supplier development and indeed in direct involvement requiring commitment of their time and resources.

A Program for Retrospective Change-Point Analysis of Individual Observations

In the retrospective (preliminary) analysis of individual observations collected over time, the in-control parameters of the process are unknown, and there is the possibility of a shift in the mean or standard deviation at any point in the observations. In many cases a special cause of variation will produce a single, sustained shift in the mean, standard deviation, or both. Presented here is a FORTRAN program for detecting such shifts, developed from a likelihood ratio test under the assumption of normality. The advantages over use of the standard X- chart and moving range chart are much higher probabilities of detection of sustained shifts and much improved diagnostic information. When a shift is detected, the most likely location is given, as well as whether it is primarily due to a shift in the mean, standard deviation, or both. Multiple shifts can often be detected by recursive application of the algorithm under the users direction. The program calculates an approximate upper control limit, using an expression that is determined from simulation by Sullivan and Woodall (1996). Alternatively, the user can provide the desired upper control limit.

Quality Planning for the Manufacturing Supply Chain

Quality as a supply chain performance measure has not been adequately explored in either the supply chain or quality management literature. Planning for quality in the manufacturing supply chain requires a context, and begins with documented customer needs. An original schematic model depicting the interrelationship between strategic production planning and strategic supply chain planning provides this context. Another result is a listing of the departmental customers for the manufacturing supply chain and their respective needs. The quality requirements for each supplied part or module, derived from these needs, should become critical criteria in supplier selection and other sourcing decisions when starting up a new plant. Although supply chains are being created today using planning processes that focus primarily on logistics and production planning, driven by cost and schedule considerations, the authors argue that there is a need to formally “plan quality in” as well.

Project Risk Identification Methods for Construction Planning and Execution

Risk identification is the often neglected first step in the risk management process: potential problems specific to the project must be identified before quantification of the magnitude of the risk, or consideration of prevention or mitigation actions. This paper presents two research conclusions: systematic project risk identification during the transition from the design to construction (planning) phase benefits from a taxonomy of risk areas/potential problems for the specific type of construction involved; during construction execution, earned value analysis is an effective method to identify emerging risks from readily available project data bases. A taxonomy was developed from published references for infrastructure construction projects, validated by construction managers, and is the main contribution of this paper. Fifteen “risk areas” were used to organize 96 potential problems.

Critical path acceleration and simulation in aircraft technology planning

Critical path methods have been in common use in the aircraft industry for a number of years, but most often after a program enters preliminary design. In aircraft technology development studies, critical path acceleration is shown to be a natural and, in fact, necessary method for planning. An implementation of a time-cost trade-off algorithm of Tufekci is discussed. Because of inherent time uncertainty, network simulation should also be performed to develop a confidence band on the traditional time-cost trade-off curve. The network simulation package VERT is available for this purpose. Two instances of applying these two methods to plans for advanced materials technology development are reported as case studies.

Supplier Cost Evaluation: Estimating the Excess Costs of Purchased Components

The performance of an original equipment manufacturer (OEM) is driven in part by the quality, delivery, and cost performance of its first-tier suppliers. Many different methods for evaluating supplier-related costs have appeared in the purchasing and quality literature. After identifying four supplier cost performance systems in current use at OEMs, the author focuses on total cost of ownership (TCO) systems. He develops an innovative, activity-based method of estimating model-year “excess costs” for a component. Such excess costs are shown to be the incremental supply chain costs of the component, as compared to ideal. The method is implemented in spreadsheet software, and an example application of this tool is provided. Potential uses for such component-specific excess cost estimates include sourcing, price negotiations, supplier rating, analysis of improvement proposals, and sharing in the costs and benefits of supplier improvement.

Application of total quality management techniques to safety analysis in software product development

The safety level of any system may be defined as the probability that no catastrophic accident is expected to occur during system operation, for a specified period of time. The purpose of software safety analysis is to reduce system risk, due to software malfunctions, to an acceptable level. Conventional software safety analysis techniques are reactive in orientation. They are intended to identify software faults as part of a post-programming evaluation procedure. A more cost-effective approach would be to avoid introducing the faults into the computer code. This paper discusses the ongoing establishment of a development methodology that would incorporate this proactive approach to software safety analysis, by applying the concepts of total quality management. This research provides a linkage of software engineering to such classical industrial engineering activities as safety analysis and quality control.