International Journal of Reliability, Quality and Safety Engineering

Systems can be classified as dynamic or static depending on the nature of the performance. A static system has a fixed target value for the performance characteristic while a dynamic system has a variable performance characteristic. This variable characteristic can be adjusted by a signal factor based on the requirement of the user. Due to this nature, modeling and optimization methods of dynamic systems may be different from those for static systems. This paper presents a two-step approach to the optimization modeling and design of linear dynamic systems. The approach is based on the concepts of quality engineering, experimental design, loss function, signal-to-noise (SN) ratio and optimization methodology. The motivation, justification and computation of SN ratios for linear dynamic systems will be discussed. As an illustrative example, a circuit design problem is given to demonstrate the method.

Basic approaches of some of the reliability models available for analyzing the effect of operating conditions (or covariates) on the lifetime of a system are shortly discussed, and a general guideline for how to select an appropriate model for a given data set is provided. Some of the models have theoretical and computational difficulties which make them difficult to apply. Models that appear to be suitable for practical applications can broadly be classified as the class of proportional hazards models and the class of accelerated failure time models. In the class of proportional hazards models, e.g. the proportional hazards model and the proportional odds model, the effect of the covariates is assumed to act multiplicatively on the hazard rate or its transformations. In the class of accelerated failure time models, e.g. the parametric regression models, the effect of the covariates is assumed to act multiplicatively on the failure time or its transformations. Models from the proportional hazards family appear to be the better ones for analyzing the effect of the covariates due to the method used for estimating the parameters of these models.

In this paper, a two-stage control chart for monitoring the defective rate of high-yield processes is proposed and studied. The Cumulative Count of Conforming control chart is generalized by using the number of items inspected until two defective items are observed. As this will increase the time to alarm, a two-stage approach combining both schemes is proposed. The occurrence of a defective within n₁ items inspected in the first stage indicates that the process is out of control. If no defective occurs within n₁ items inspected, the occurrence of two defectives within the next n₂ - n₁ in the second stage also indicates that the process is out of control. The probability of making a false alarm at the first and second stages are equal to α₁ and α₂, respectively. This procedure improves the sensitivity of the control chart in detecting shifting of the process defective rate p when p is at the parts-per-million order of magnitude.

Accurate analysis of the reliability of a system requires that all major variations in the system's operation are accounted for. Most reliability analyses assume that the system's configuration, success criteria, and components' behavior remain the same during an entire mission. However, system requirements and behavior vary over the time. Moreover, individual components may be repaired (if failed) such that repairs do not conflict with the operation of the system. Thus repairs are independent of the system state. Thus, multiple phases are natural. For repairable systems, Markov analysis techniques are used but they suffer from state space explosion. This limits the size of system that can be analyzed and it is expensive in computation. We present a new computationally efficient technique for the analysis of phased-mission systems where the operational states of a system can be described by combinations of component states (such as fault trees or assertions). We avoid the state space explosion. The phase algebra is used to account for the effects of variable configurations, repairs, and success criteria from phase to phase. Our technique yields exact (as opposed to approximate) results. We demonstrate our technique by means of several examples and present numerical results to show the effects of phases and repairs on the system reliability/availability.

This paper presents a set of related quasiconvex quality loss functions. Characteristics of quasiconvex functions that are desirable for modeling quality loss are noted. Three frequently used univariate quasiconvex quality loss functions are discussed. Bivariate and multivariate quasiconvex quality loss functions are developed. A set of necessary and sufficient conditions is established for the quasiconvexity of multivariate quality loss functions. An industrial product example is used to illustrate the development of a bivariate quadratic quality loss function.

The rapid growth of air traffic density has long demanded the Federal Aviation Administration (FAA) to design an effective safety inspection system. Well defined thresholds are essential to such an inspection system since they provide standards for both monitoring and regulating purposes. In this paper, we use control chart techniques to derive thresholds and standards for inspection measures, and to provide charts for monitoring them continuously. These thresholds charts play the same role as control charts do in statistical quality control for the monitoring of manufacturing processes. In quality control, the centerline of the chart indicates the target value and the control limits determine whether the process is out of control. In safety inspections, we view the centerline as the safety standard, and justify some properly chosen levels of control limits as meaningful thresholds. For FAA safety inspection surveillance results concerning air carriers, these thresholds are termed alert, advisory, expected, and informational. They provide a concrete measure of the inspection results in terms of the severity of potential flaws, and serve as a guideline for the general rating of the safety performance of each carrier. Furthermore, we can now use the so-called average run length to measure the effectiveness of the inspection system with the proposed thresholds. This approach is implemented on a dataset of an operational performance measure collected from ten air carriers over a period of 6 months. The results are very supportive.