Knowledge that Transforms

We study the problem of allocating stocked fibers to made-to-order cables with the goals of satisfying due dates and reducing the costs of scrap, setup, and fiber circulation. These goals are achieved by generating remnant fibers either long enough to satisfy future orders or short enough to scrap with little waste. They are also achieved by manufacturing concatenations, in which multiple cable orders are satisfied by the production of a single cable that is afterwards cut into the constituent cables ordered. We use a function that values fibers according to length, and which can be viewed as an approximation to the optimal value function of an underlying dynamic programming problem. The daily policy that arises under this approximation is an integer program with a simple linear objective function that uses changes in fiber value to take into account the multi-period consequences of decisions. We describe our successful implementation of this integer program in the factory, summarizing our computational experience as well as realized operational improvements.

In preparing a review I recently discovered an important reference for a key result in Graves (1999). Wecker (1979) had previously derived the variance for demand over a deterministic lead-time for an IMA (0, 1, 1) demand process. I develop effectively the same result, given as Equation (8) in Graves (1999). I state this as the variance of the inventory random variable for an inventory system that is subject to an IMA (0, 1, 1) demand process, a deterministic replenishment lead-time and an adaptive base-stock control policy given by (7). But given these assumptions, the variance of the inventory is the same as the variance of the demand over the lead-time. Although the Wecker manuscript has not been published, Eppen and Martin (1988) reference it and use the key result as part of their research.

Paced or Synchronous assembly lines are a popular class of assembly systems consisting of a series of assembly stations arranged in tandem. Every job (or order) visits all assembly stations in the same sequence and spends the same amount of time (known as the production cycle) at each station. Industries such as aircraft, fire-engine, and automobile assembly have production cycles of a few hours and are labor intensive. In spite of increased automation in such industries, human capital remains the most expensive and important contributor to a flexible production system. In this article we formulate the cross-training problem on a paced assembly line with m stations (mCT). We assume that each worker possesses a number of skills referred to as a skill vector. Our objective is to schedule a set of work orders through the assembly system so as to minimize the size of the required workforce and/or the workforce cross-training costs. We analyze the complexity of mCT and identify polynomially solvable cases. A variety of lower bounds is developed based on optimization techniques. These lower bounds are used to develop a branch and bound algorithm as well as to evaluate our heuristics. A computational experiment reports the performance of all algorithms. Using these algorithms, we examine how the formation of skill vectors affects the workforce size and draw guidelines for cross-training programs in organizations with labor intensive assembly operations.

In this paper, we address the product-family design problem of a firm in a market in which customers choose products based on some measure of product performance. By developing products as a family, the firm can reduce the cost of developing individual product variants due to the reuse of a common product platform. Such a platform, designed in an aggregate-planning phase that precedes the development of individual product variants, is itself expensive to develop. Hence, its costs must be weighted against the benefits of its reuse in a family. We offer a model for capturing costs of product development when the family consists of variants based on a common platform. It is shown that the model can be converted into a network-optimization problem, and the optimal product-family can be identified under fairly general conditions by determining the shortest path of its network formulation. We also analytically examine the effect of alternative product designs on product-family composition, and discuss the implications of investing in new-product technology. Finally, we illustrate our model and managerial insights with an application from the electronics industry.

This paper considers serial production-transportation systems. In recent years, researchers have developed a fairly simple functional equation that characterizes optimal system behavior, under the assumption of constant leadtimes. We show that the equation covers a variety of stochastic-leadtime systems as well. Still, many basic managerial issues remain obscure: When should stock be held at upstream stages? Which system attributes drive overall performance, and how? To address these questions, we develop and analyze several heuristic methods, inspired by observation of common practice and numerical experiments. One of these heuristics yields a bound on the optimal average cost. We also study a set of numerical examples, to gain insight into the nature of the optimal solution and to evaluate the heuristics.

We argue that industries and industry segments are characterized by a clockspeed that gauges the velocity of change in the external business environment and sets the pace of their firms' internal operations. Using data from the electronics industry, we develop and validate an integrated metric for clockspeed that takes into account both demand- and supply-side factors. We show that after controlling for product complexity and other factors, higher industry clockspeed is associated with faster execution in product development and manufacturing (e.g., shorter development time, quicker stabilization of production) and more frequent changes in organizational structure. Our findings on the effects of clockspeed can help researchers studying other industries, and our results provide benchmarks against which practitioners can compare and classify their own organizations.

The Quantity Flexibility (QF) contract is a method for coordinating materials and information flows in supply chains operating under rolling-horizon planning. It stipulates a maximum percentage revision each element of the period-by-period replenishment schedule is allowed per planning iteration. The supplier is obligated to cover any requests that remain within the upside limits. The bounds on reductions are a form of minimum purchase commitment which discourages the customer from overstating its needs. While QF contracts are being implemented in industrial practice, the academic literature has thus far had little guidance to offer a firm interested in structuring its supply relationships in this way. This paper seeks to address this need, by developing rigorous conclusions about the behavioral consequences of QF contracts, and hence about the implications for the performance and design of supply chains with linkages possessing this structure. Issues explored include the impact of system flexibility on inventory characteristics and the patterns by which forecast and order variability propagate along the supply chain. The ultimate goal is to provide insights as to where to position flexibility for the greatest benefit, and how much to pay for it.

In this paper, we consider an adaptive base-stock policy for a single-item inventory system, where the demand process is nonstationary. In particular, the demand process is an integrated moving average process of order (0, 1, 1), for which an exponential-weighted moving average provides the optimal forecast. For the assumed control policy we characterize the inventory random variable and use this to find the safety stock requirements for the system. From this characterization, we see that the required inventory, both in absolute terms and as it depends on the replenishment lead-time, behaves much differently for this case of nonstationary demand compared with stationary demand. We then show how the single-item model extends to a multi-stage, or supply-chain context; in particular we see that the demand process for the upstream stage is not only nonstationary but also more variable than that for the downstream stage. We also show that for this model there is no value from letting the upstream stages see the exogenous demand. The paper concludes with some observations about the practical implications of this work.