A k-out-of-n system with n identical and independent components is considered in which the components takes two types of function: 0 (open-circuit) or 1 (closed) on command (e.g. electromagnetic relays and solid state switches). Components are subject to two types of failure on command: failure to close or failure to open. In our k-out-of-n system, failure of (n-k)＋1 or more components to close causes to the close failure of the system, or failure of k or more components to open causes the open failure of the system. The long-run average cost rate is obtained. We find the optimal k minimizing the long run average cost rate for given n. A numerical example is presented.

A non-parametric test based on U-statistic for testing exponentiality against the new better than used renewal failure rate (NBURFR) alternatives is introduced and the percentiles of this test statistic are tabulated for sample size 5(1)50. Its properties are also discussed including the Pitman asymptotic efficiency relative to the tests of the new better than used and new better than used failure rate (Ahmed (1994) and Hendi (2000)). The powers of this test are also calculated for some used life distributions. An example from blood cancer patients demonstrates a practical application of our test in the medical sciences is presented. Finally the problem when right-censored data is available is handled.

We propose, in this paper, the discrete version of NBU$_{Mg}$and show that the NBU$_{Mg}$ class is preserved under both the non-homo geneous poisson shock model and the general shock model.

Manufacturers collect and analyze field reliability data to enhance the quality and reliability of their products and to improve customer satisfaction. To reduce the data collecting and maintenance costs, the amount of data maintained for evaluating product quality and reliability should be minimized. With this in mind, some industrial companies assemble warranty databases by gathering data from different sources for a particular time period. This “marginal count failure data” does not provide (i) the number of failures by when the product entered service, (ii) the number of failures by product age, or (iii) information about the effects of the operating season or environment. This article describes a method for estimating age-based claim rates from marginal count failure data. It uses covariates to identify variations in claims relative to variables such as manufacturing characteristics, time of manufacture, operating season or environment. A Poisson model is presented, and the method is illustrated using warranty claims data for two electrical products.