Gittins-index heuristics for research planning

The classical work of Gittins, which resulted in the celebrated index result, had applications to research planning as an important part of its motivation. However, research planning problems often have features that are not accommodated within Gittins's original framework. These include precedence constraints on the task set, influence between tasks, stopping or investment options and routes to success in which some tasks do not feature. We consider three classes of Markovian decision models for research planning, each of which has all of these features. Gittins-index heuristics are proposed and are assessed both analytically and computationally. They perform impressively. © 1995 John Wiley & Sons, Inc.     

A bayesian approach to environmental stress screening

The article presents a Bayesian analysis for the environmental stress screening problem. The decision problem of deriving optimal stress screen durations is solved. Given a screen duration, the optimal stress level can also be determined. Indicators of the quality of a screen of any duration are derived. A statistical model is presented which allows a posterior density for the rate of early failures of the production process to be calculated. This enables the user to update his opinion about the quality of the process. © 1994 John Wiley & Sons, Inc.     

The simple multiserver loss system with time-varying intensities

For the loss system with time-varying intensities and a finite number of identical servers we derive formulas for the expected number of refused customers in a given time interval and for the expected number of servers busy at a given time instant. © 1994 John Wiley & Sons, Inc.     

A two-machine repair model with variable repair rate

A two-unit cold standby production system with one repairman is considered. After inspection of a failed unit the repairman chooses either a slow or a fast repair rate to carry out the corresponding amount of work. At system breakdown the repairman has an additional opportunity to switch to the fast rate. If there are no fixed costs associated with system breakdowns, then the policy which minimizes longrun average costs is shown to be a two-dimensional control limit rule. If fixed costs are incurred every time the system breaks down, then the optimal policy is not necessarily of control limit type. This is illustrated by a counterexample. Furthermore, we present several performance measures for this maintenance system controlled by a two-dimensional control limit rule. © 1993 John Wiley & Sons, Inc.     

Bounds and properties of the expected value of sample information for a project-selection problem

In this article we extend the work of Mehrez and Stulman [5] on the expected value of perfect information (EVPI) to the expected value of sample information (EVSI) for a class of economic problems dealing with the decision to reject or accept an investment project. It is shown that shifting the mean of the underlying a priori distribution of X, the project's monetary value from zero in either direction will decrease the associated EVSI of Y, the random sampled information. A theorem is then presented which gives an upper bound on the EVSI over all distributions of Y, as well as the structure of the posterior mean E[X|Y] for which this upper bound is achieved. Finally, the case where E[X|Y] is linear in Y is discussed and its performance compared with that of the optimal case.     

Safety stock reduction by order splitting

In this article we consider an item for which a continuous review, reorder point, order quantity inventory control system is used. The amount of safety stock required depends upon, among other factors, the average value and variability of the length of the replenishment lead time. One way to reduce these quantities is to split orders among two or more vendors. In this article the random lead times are assumed to have Weibull distributions. This permits the development of analytic expressions for the reduction in the expected value and variability of total demand until the critical first (earliest) delivery received from a vendor. An expression is also obtained for the reorder point that provides a given probability of no stockout prior to the first delivery. Lower bounds are given on the order quantity so as to ensure that the probability of a stockout before any one of the later (second, third, etc.) deliveries is sufficiently small to be considered negligible. The analytic and tabular results can be used to estimate the benefits (reduced carrying costs and/or increased service level) of order splitting.     

Is human trafficking the dark side of economic freedom?

Economic freedom has increased living standards worldwide. Concurrent with such gains are rising concerns about potential human costs associated with free markets. This paper uses data on human trafficking and anti-trafficking policies, in conjunction with a measure of economic freedom, to examine whether free markets exacerbate or attenuate the incidence of human trafficking and policies designed to combat it. We do not find evidence suggesting that economic freedom is associated with human trafficking. In addition, our results suggest that economically free countries are more likely to enact and enforce policies to fight human trafficking.     

Modeling the performance of a scanning radio communications sensor

A model which allows the prediction of long-term average report rates against an array of radio emitters by a scanning sensor is presented. The model is based on Markov renewal theory.     

Markovian multiechelon repairable inventory system

In this article we model a two-echelon (two levels of repair, one level of supply) repairable-item inventory system using continuous-time Markov processes. We analyze two models. In the first model we assume a system with a single base. In the second model we expand this model to include n bases. The Markov approach gives rise to multidimensional state spaces that are large even for relatively small problems. Because of this, we utilize aggregate/disaggregate techniques to develop a solution algorithm for finding the steady-state distribution. This algorithm is exact for the single-base model and is an approximation for the n-base model, in which case it is found to be very accurate and computationally very efficient.