Knowledge that Transforms

Under an alternate assumption of the payoff function, we analyze Lee's dynamic game model of R&D rivalry. Both similarities and differences in the equilibrium results of the model are obtained. It is shown that both the reswitching property and the convergence property of R&D rivalry are robust under the alternate assumption. The reswitching property of R&D rivalry states that after gaining a technological edge against a rival, a decision maker stops doing R&D and he will resume doing R&D when his rival succeeds in narrowing the technological gap between the two rivals. The convergence property of R&D rivalry states that when the technology levels of two rivals differ by a wide margin, the one with a lower technology level will be the only one doing R&D to narrow the technological gap between the two rivals. This formalizes the perception of R&D managers that the R&D decision of one firm should depend on the R&D decision of its rivals for competitive reasons. Moreover, when multiple Nash equilibria exist, a different pair of equilibria is obtained under the alternate assumption. Insights to an antitrust puzzle relevant to managers are provided. Finally, this paper provides an explanation of why the market shares of firms in an industry may differ. At stationary states of technologies, asymmetric technologies (market shares) are expected.

In many practical applications or multi-item inventory systems significant economies of scale can be exploited when coordinating replenishment orders for groups of items. This paper considers a continuous review multi-item inventory system with compound Poisson demand processes; excess demands are backlogged and each replenishment requires a lead time. There is a major setup cost associated with any replenishment of the family of items, and a minor (item dependent) setup cost when including a particular item in this replenishment. Moreover there are holding and penalty costs. We present an algorithm which searches for a simple coordinated control rule minimizing the long-run average cost per unit time subject to a service level constraint per item on the fraction of demand satisfied directly from on-hand inventory. This algorithm is based on a heuristic decomposition procedure and a specialized policy-iteration method to solve the single-item subproblems generated by the decomposition procedure. The model applies to multi-location inventory systems with similar cost structures for coordinated deliveries.

Models of decision-making in multi-level, hierarchical systems in the literature have been restricted to structures where the system is strictly pyramidal in nature. We will show that this restriction has been self-imposed and that existing models can be modified to represent a wide range of nonpyramidal structures.

We examine a queueing system with multiple primary servers and a fewer number of auxiliary servers. There are two classes of customers—those who require service from a primary server working alone and those who require service from a primary server who is assisted by an auxiliary server. Though the apparent Markovian state space is five-dimensional, we show that an aggregation results in an exact two-dimensional representation which is Markovian. Matrix geometric theory is used to obtain approximations for the mean delay and blocking probability of each customer type.

The two types of mathematical manpower planning models that appear in the literature involve either longitudinal or cross-sectional formulations. Despite the high degree of realism achieved, the use of longitudinal models is limited because the implementation requires the knowledge of a large amount of historical personnel data that is often unavailable. The value of cross-sectional models requiring a minimal amount of data is diminished due to (1) the difficulty in transferring cross-sectional results into cohort information, and (2) an assumption implicit in the structure of these models stating that the movement of an individual from one grade in the organization to another is independent of that person's organizational age. In this paper, we present a cohort (longitudinal) personnel planning model solved using distributed parameter optimal control theory that requires only cross-sectional data. We derive the optimal hiring, promotion, separation and retirement policies of an organization as functions of time and a person's organizational age and grade. In response to changing goal levels of manpower, we observe changes in the optimal policies and their subsequent effect on the career paths of cohort groups over time.

Since the introduction of the concept of the product life cycle (PLC) some decades ago a great deal has been written on the subject and several empirical studies have been conducted. However, empirical research to date is often limited in scope. It is oriented towards the analysis of the growth and saturation stages and focused primarily on the study of consumer goods. Furthermore, the application of the PLC concept in strategic planning is largely disappointing, the normative orientation of most of this literature being somewhat naive and misleading. The aim of the present research is to study the efficiency of different strategies for achieving market share and cash flow objectives. This study is performed on different types of businesses producing both consumer and industrial products situated at various stages on the product life cycle: growth, maturity and decline. A sample of approximately 1,100 businesses (217 in the growth stage, 315 in the maturity stage, 569 in the decline stage) is drawn from the PIMS data base. A cluster analysis is run to identify natural groups of homogeneous businesses. For each of the nine identified groups, linear regression models are estimated to study the influence of strategic actions—finance, marketings, production, R&D, personnel—on two criteria of performance: market share and cash flow. The research indicates that strategies not only depend on the life cycle stage but also are influenced by the goal orientation of the firm—short-term (cash-flow) or long-term (market share)—and that success strategies appear to be contingent upon the business and the environmental characteristics.

Improved scheduling rules and spares stocking policies are examined for a repair shop supporting a multi-item repairable inventory system with a hierarchical product structure. The results from an earlier paper are extended to the case where more than one unit fails in each item to be repaired. A variety of scheduling rules and methods for initializing the vector of spare parts are evaluated using a simulation of a representative shop. The results indicate that procedures which perform well when only a single unit fails also perform well when multiple failures occur. More complex rules, which recognize assembly mating requirements, do not appear to provide any significant improvement.

Earlier models for the evaluation of assuming and shedding risks of death are extended to the case of possibly extreme disability over a lifetime. Earlier micromort values are supplemented with microdisability values to enable individuals to make consistent decisions in hazardous situations.

The process of converting a tree into finished lumber involves several interrelated steps. The two of greatest potential for improvement are cutting the felled tree into shorter logs (called bucking the tree) and sawing the bucked logs into lumber. Unlike previous work in the area, this paper considers the two activities together as a single production system. The model allows for variations in tree shape and quality which could be recorded by modern electronic scanners. Among the principal techniques used in the paper are dynamic programming for both bucking and sawing, and a procedure for calculating the distance between two polyhedral sets in R 2 . Computational experience is reported on trees with representative defect patterns, taper, and sweep.

Single objective cost minimization linear programming models are used as computerized decision-aids in sausage manufacturing (hot dogs, bologna, salami, etc.). However, sausage blending is clearly a problem with multiple conflicting criteria (cost, color, fat, protein, moisture, etc.) Presented in this paper is a vector-maximum/filtering MOLP (multiple objective linear programming) methodology for use as an improved decision-making approach with single formula sausage blending problems.