J. D. Esary’s research while affiliated with Naval Postgraduate School and other places

Several conditions are considered that extend to a multivariate setting the univariate concept of an increasing hazard rate average. The relationships between the various conditions are established. In particular it is shown that if for some independent random variables $X_1, \cdots, X_k$ with increasing hazard rate average and some coherent life functions $\tau_1, \cdots, \tau_n$ of order $k, T_i = \tau_i(X_1, \cdots, X_k)$, then the joint survival function $\bar{F}(\mathbf{t}) = P(T_1 > t_1, \cdots, T_n > t_n)$ has the property that $\alpha^{-1} \log \bar{F}(\alpha\mathbf{t})$ is decreasing in $\alpha > 0$ whenever each $t_i \geqslant 0$. Various other properties of the multivariate conditions are given. The conditions can all be stated in terms of inequalities in which equality implies that the one dimensional marginal distributions are exponential. For most of the conditions, the form of the multivariate exponential distributions that satisfy the equality is exhibited.

The calculation of the exact reliability of complex systems is a difficult and tedious task. Consequently simple approximating techniques have great practical value. The hazard transform of a system is an invertible transformation of its reliability function which is convenient and useful in both applied and theoretical reliability work. A simple calculus for finding an approximate hazard transform for systems formed by series and parallel combinations of components is extended so that it can be used for any coherent system. The extended calculus is shown to lead to conservative approximations. A first order version of the extended calculus is also discussed. This method of approximation is even more simple to use, but is not always conservative. Examples of its application indicate that it is capable of giving quite accurate results.

The life distribution H(t) of a device subject to shocks governed by a Poisson process is considered as a function of the probabilities $P_k$ of not surviving the first k shocks. Various properties of the discrete failure distribution $P_k$ are shown to be reflected in corresponding properties of the continuous life distribution H(t). As an example, if $P_k$ has discrete increasing hazard rate, then H(t) has continuous increasing hazard rate. Properties of $P_k$ are obtained from various physically motivated models, including that in which damage resulting from shocks accumulates until exceedance of a threshold results in failure. We extend our results to continuous wear processes. Applications of interest in renewal theory are obtained. Total positivity theory is used in deriving many of the results.

A fairly common failure model in a wide variety of contexts is a cumulative damage process, in which shocks occur randomly in time and associated with each shock there is a random amount of damage which adds to previously incurred damage until a breaking threshold is reached. The multivariate life distributions that are induced when several components, each with its own breaking threshold, are exposed to the same cumulative damage process are of interest in their own right, and are important examples in the general study of multivariate life distributions. The paper is a summary of some results about the very special, but central, case in which the cumulative damage process is a compound Poisson process. It is focused on the multivariate life distributions that arise when the component breaking thresholds are random and have a Marshall-Olkin multivariate exponential distribution. The results have application to the life distribution of a coherent system whose components are exposed to the damage process. (Modified author abstract)

We consider some unresolved relationships among various notions of bivariate dependence. In particular we show that $P\lbrack T > t \mid S > s\rbrack \uparrow$ in s (or alternately, $P\lbrack T \leqq t \mid S \leqq s\rbrack \downarrow$ in s) implies $S, T$ are associated, i.e. $\operatorname{Cov} \lbrack f(S, T), g(S, T)\rbrack \geqq 0$ for all non-decreasing f and g.

: A systematic study is made of the role of the hazard transform in developing new results in reliability theory, and simplifying the proofs of known results. The hazard transform of a coherent system gives the hazard function (or cumulative hazard rate) of the system in terms of component hazard functions. (Author)

In this article the minimal cut lower bound on the reliability of a coherent system, derived in Esary-Proschan [6] for the case of independent components not subject to maintenance, is shown to hold under a variety of component maintenance policies and in several typical cases of component dependence. As an example, the lower bound is obtained for the reliability of a “two out of three” system in which each component has an exponential life length and an exponential repair time. The lower bound is compared numerically with the exact system reliability; for realistic combinations of failure rate, repair rate, and mission time, the discrepancy is quite small.

It is customary to consider that two random variables S and T are associated if $\operatorname{Cov}\lbrack S, T\rbrack = EST - ES\cdot ET$ is nonnegative. If $\operatorname{Cov}\lbrack f(S), g(T)\rbrack \geqq 0$ for all pairs of nondecreasing functions $f, g$, then S and T may be considered more strongly associated. Finally, if $\operatorname{Cov}\lbrack f(S, T), g(S, T)\rbrack \geqq 0$ for all pairs of functions $f, g$ which are nondecreasing in each argument, then S and T may be considered still more strongly associated. The strongest of these three criteria has a natural multivariate generalization which serves as a useful definition of association: DEFINITION 1.1. We say random variables $T_1,\cdots, T_n$ are associated if \begin{equation*}\tag{1.1}\operatorname{Cov}\lbrack f(\mathbf{T}), g(\mathbf{T})\rbrack \geqq 0\end{equation*} for all nondecreasing functions f and g for which $Ef(\mathbf{T}), Eg(\mathbf{T}), Ef(\mathbf{T})g(\mathbf{T})$ exist. (Throughout, we use $\mathbf{T}$ for $(T_1,\cdots, T_n)$; also, without further explicit mention we consider only test functions $f, g$ for which $\operatorname{Cov}\lbrack f(\mathbf{T}), g(\mathbf{T})\rbrack$ exists.) In Section 2 we develop the fundamental properties of association: Association of random variables is preserved under (a) taking subsets, (b) forming unions of independent sets, (c) forming sets of nondecreasing functions, (d) taking limits in distribution. In Section 3 we develop some simpler criteria for association. We show that to establish association it suffices to take in (1.1) nondecreasing test functions f and g which are either (a) binary or (b) bounded and continuous. In Section 4 we develop the special properties of association that hold in the case of binary random variables, i.e., random variables that take only the values 0 or 1. These properties turn out to be quite useful in applications. We also discuss association in the bivariate case. We relate our concept of association in this case to several discussed by Lehmann (1966). Finally, in Section 5 applications in probability and statistics are presented yielding results by Robbins (1954), Marshall-Olkin (1966), and Kimball (1951). An application in reliability which motivated our original interest in association will be presented in a forthcoming paper.

It is well known that the future life distribution of a device remains the same regardless of the time it was previously in use, if and only if the life distribution of that device is exponential. For this reason exponential life distributions are accepted as characterizing the phenomenon of no wear. The problem of finding a class of life distributions which would similarly reflect the phenomenon of wear-out has been under investigation for some time. In answer to this problem we introduce in this paper the class of IHRA (Increasing Hazard Rate Average) distributions and show that it has, among others, the following optimal properties: (i) it contains the limiting case of no wear, i.e., all exponential distributions, (ii) whenever components which have IHRA life distributions are put together into a coherent system, this system again has an IHRA life distribution, i.e., a system wears out when its components wear out, and (iii) the IHRA class is the smallest class with properties (i) and (ii).