A Markovian model for process setup and improvement

We examine the setup and improvement policies for a production process with multiple performance states. Assume that the production process deteriorates randomly over time, following a Markovian process with known transition probabilities. In order to reduce the production cost incurred because of process deterioration, the process is inspected at the end of each period. Then one of three actions may be taken: do nothing, perform routine process setup, or perform routine setup and process improvement. The routine setup operation returns the process to its best performance state, whereas the process improvement action may reduce future production and setup costs and improve the process-state transition probabilities. A discounted Markovian model is formulated to find the strategy that minimizes the total cost of operating the production process. © 1997 John Wiley & Sons, Inc. Naval Research Logistics 44: 383–400, 1997     

A heuristic algorithm for determining replacement policies in k-out-of-n systems

The authors study a discrete-time, infinite-horizon, dynamic programming model for the replacement of components in a binary k-out-of-n failure system. (The system fails when k or more of its n components fail.) Costs are incurred when the system fails and when failed components are replaced. The objective is to minimize the long-run expected average undiscounted cost per period. A companion article develops a branch-and-bound algorithm for computing optimal policies. Extensive computational experiments find it effective for k to be small or near n; however, difficulties are encountered when n ≥ 30 and 10 ≤ k ≤ n − 4. This article presents a simple, intuitive heuristic rule for determining a replacement policy whose memory storage and computation time requirements are O(n − k) and O(n(n − k) + k), respectively. This heuristic is based on a plausible formula for ranking components in order of their usefulness. The authors provide sufficient conditions for it to be optimal and undertake computational experiments that suggest that it handles parallel systems (k = n) effectively and, further, that its effectiveness increases as k moves away from n. In our test problems, the mean relative errors are under 5% when n ≤ 100 and under 2% when k ≤ n − 3 and n ≤ 50. © 1997 John Wiley & Sons, Inc. Naval Research Logistics 44, 273–286, 1997.     

Volume discount decisions: Wealth maximization or EOQ approach?

A recent paper finds that when volume discounts are available, in some cases, reliance on the Economic Order Quantity (EOQ) model can induce purchasers to make wealth reducing decisions, and the Present Value (PV) approach should be preferred. While this finding is theoretically correct, the magnitudes of wealth reductions suggested by the paper's numerical examples seem to be questionable. Furthermore, the paper also finds that, in some other cases, a purchaser using the EOQ approach realizes a net increase in current wealth compared to a purchaser using the PV approach. Logic suggests that such a finding cannot be correct, since by its very definition, it is the PV approach that seeks to maximize the current wealth. We offer an alternative frame of comparison and a modified model to show that, under the paper's assumptions, the EOQ approach can never realize a net increase in current wealth compared to the current wealth generated by the PV approach. On the other hand, we also show that when typical values of the relevant parameters prevail, the additional costs imposed by the EOQ approach are not significant. Finally, we suggest that insofar as the PV approach requires greater administrative costs to implement, it may even be counterproductive to the goal of wealth maximization. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 377–389, 1998     

A general model for accelerated life testing with time‐dependent covariates

This paper introduces a general or “distribution‐free” model to analyze the lifetime of components under accelerated life testing. Unlike the accelerated failure time (AFT) models, the proposed model shares the advantage of being “distribution‐free” with the proportional hazard (PH) model and overcomes the deficiency of the PH model not allowing survival curves corresponding to different values of a covariate to cross. In this research, we extend and modify the extended hazard regression (EHR) model using the partial likelihood function to analyze failure data with time‐dependent covariates. The new model can be easily adopted to create an accelerated life testing model with different types of stress loading. For example, stress loading in accelerated life testing can be a step function, cyclic, or linear function with time. These types of stress loadings reduce the testing time and increase the number of failures of components under test. The proposed EHR model with time‐dependent covariates which incorporates multiple stress loadings requires further verification. Therefore, we conduct an accelerated life test in the laboratory by subjecting components to time‐dependent stresses, and we compare the reliability estimation based on the developed model with that obtained from experimental results. The combination of the theoretical development of the accelerated life testing model verified by laboratory experiments offers a unique perspective to reliability model building and verification. © 1999 John Wiley & Sons, Inc. Naval Research Logistics 46: 303–321, 1999     

Setting military reenlistment bonuses

The United States military frequently has difficulty retaining enlisted personnel beyond their initial enlistment. A bonus program within each service, called a Selective Reenlistment Bonus (SRB) program, seeks to enhance reenlistments and thus reduce personnel shortages in critical military occupational specialties (MOSs). The amount of bonus is set by assigning “SRB multipliers” to each MOS. We develop a nonlinear integer program to select multipliers which minimize a function of deviations from desired reenlistment targets. A Lagrangian relaxation of a linearized version of the integer program is used to obtain lower bounds and feasible solutions. The best feasible solution, discovered in a coordinate search of the Lagrangian function, is heuristically improved by apportioning unexpended funds. For large problems a heuristic variable reduction is employed to speed model solution. U.S. Army data and requirements for FY87 yield a 0-1 integer program with 12,992 binary variables and 273 constraints, which is solved within 0.00002% of optimality on an IBM 3033AP in less than 1.7 seconds. More general models with up to 463,000 binary variables are solved, on average, to within 0.009% of optimality in less than 1.8 minutes. The U.S. Marine Corps has used a simpler version of this model since 1986. © 1993 John Wiley & Sons, Inc.     

Parallel machine replacement

We consider the parallel replacement problem in which there are both fixed and variable costs associated with replacing machines. Increasing maintenance costs motivate replacements, and the fixed replacement cost provides incentive for replacing machines of different ages together in “clusters.” We prove two intuitive results for this problem. First, it is never optimal to split a cluster of like-aged machines, and second, it is never optimal to replace newer clusters before older clusters. By incorporating these two results into an algorithmic approach, we vastly reduce the amount of computation required to identify an optimal replacement policy.     

Learning and costs in airframe production: A multiple-output production function approach

This article examines the influence of both production rate and learning on airframe program costs. A dynamic multiple-output production model is developed and is used to observe the cost impact of changes in production rate and learning. Several simulations are performed to demonstrate the sensitivity of the optimal time path of cost to changes in the model parameters. The model is applied by estimating parameters from the F102 airframe program.     

Book reviews

British Naval Policy and Norwegian Security: Maritime Power in Transition, 1951-60.. By Mats Berdal. Forsvarsstudier No. 2/92, Institutt For Forsvarsstudier (IFS) Oslo, Norway, 1992, ISSN 0333-3981.

The European Security Order Recast: Scenarios for the Post-Cold War Era.. By Barry Buzan, Morten Kelstrup, Pierre Lemaitre, Elzbieta Tromer and OLE Waever. Pinter, London (1990), ISBN 086187143X, £30.00.

Mates and Muchachos: Unit Cohesion in the Falklands War.. By Nora Kinzer Stewart. Brassey's (US), Washington DC (1991), ISBN 008-037-4395, $20.

The Declining Hegemon: The United States and European Defense, 1960-1990.. By Joseph Lepgold. Praeger, New York (1990), ISBN 0-275-93657-0, $15.95 pb.

Inventing Accuracy: A Historical Sociology of Nuclear Missile Guidance.. By Donald Mackenzie. MIT Press, Cambridge MA, and London, England (1990), ISBN 262-132-583, £19.95.     

A Lanchester-type aggregated-force model of conventional ground combat

This article develops a Lanchester-type model of large-scale conventional ground combat between two opposing forces in a “sector”. It is shown that nonlinear Helmbold-type equations of warfare with operational losses may be used to represent the loss-rate curves that have been used in many aggregated-force models. These nonlinear differential equations are used to model the attrition of combat capability (as quantified by a so-called firepower index) in conjunction with a rate-of-advance equation that relates motion of the contact zone (or FEBA) between the opposing forces to the force ratio and tactical decisions of the combatants. This simplified auxiliary model is then used to develop some important insights into the dynamics of FEBA movement used in large-scale aggregated-force models. Different types of behavior for FEBA movement over time are shown to correspond to different ranges of values for the initial force ratio, for example, an attack will “stall out” for a range of initial force ratios above a specific threshold value, but it will “break out” for force ratios above a second specific threshold value. Such FEBA-movement predictions are essentially based on being able to forecast changes over time in the force ratio.