In This Issue

The rapidly aging population has resulted in an urgent need for long-term care programs and facilities to develop capacity expansion plans to meet this need. Current practice has been to use a fixed ratio of beds per person over age 75 for planning. This approach lacks rigor or empirical basis and is insensitive to regional demographic differences. It has resulted in long wait times for admission to care or to excess capacity. In “A Simulation Optimization Approach to Long-Term Care Capacity Planning,” Y. Zhang, M. L. Puterman, M. Nelson, and D. Atkins present a methodology for setting long-term care capacity levels over a multiyear planning horizon to achieve target wait time service levels. Their approach integrates demographic and survival analysis, discrete event simulation, and optimization. They illustrate this approach and derive policy recommendations through two case studies: one for a regional health authority in British Columbia, Canada, and the other for an individual long-term care facility. As consumers, we all struggle with technology adoption decisions. Should I buy a new cell phone (or a new computer, new television, new car, and so on) now or wait for the next generation of the product, which might be better or cheaper? Firms face similar difficulties: Should I install a new power plant (or new piece of equipment, new manufacturing process, and so on) now or wait for a future generation of the technology, which might be better or cheaper? In “Technology Adoption with Uncertain Future Costs and Quality,” J. E. Smith and C. Ulu study these kinds of technology adoption decisions, with uncertainty about future costs and quality. They compare three models of the adoption decision: (i) a simple NPV model where the consumer adopts any technology whose NPV of benefits exceeds the cost of adoption; (ii) a stochastic dynamic programming model where the consumer can wait to adopt; and (iii) a more sophisticated (and more realistic) dynamic programming model where the consumer can wait to adopt and may also repeatedly adopt, by upgrading over time. Although the first two models have many intuitive structural properties (e.g., improving the technology makes consumers better off and encourages adoption), the more realistic third model has much more complicated policies, and some of these intuitive properties need not hold. For example, improving the technology may make the consumer better off and discourage adoption. In short, the technology adoption decisions that we consumers find difficult to make can also be quite difficult to study. Economic dispatch is the problem of determining the most efficient, low-cost, and reliable operation of a power system by dispatching the available electricity generation resources to the load on the system. The primary objective of economic dispatch is to minimize the total cost of generation while satisfying the physical constraints and operational limits. This problem is formulated as a nonconvex nonlinear programming problem, which is, in general, difficult to solve. In “Lagrangian Duality and Branch-and-Bound Algorithms for Optimal Power Flow,” D. T. Phan proposes an efficient method to solve the problem to optimality. Stochastic dynamic inventory problems with fixed costs are technically challenging because the objective function is neither concave nor convex. The problems have traditionally been analyzed by showing that the objective function is K-concave or non-K-increasing. In “On the Quasiconcavity of Lost-Sales Inventory Models with Fixed Costs,” Q. Li and P. Yu identify a set of conditions under which both the objective function and the maximal value function are quasiconcave. Not only is the quasiconcavity useful in computation, it leads to sharper and more intuitively appealing characterization of the optimal policies. This work presents a stochastic optimization approach to placement of passive acoustic sensors for underwater threat detection in a surveillance zone with complex environmental noise. In “Stochastic Optimization of Sensor Placement for Diver Detection,” A. Molyboha and M. Zabarankin suggest an approach that is efficient in numerical experiments with real data and can be used in underwater security systems. They also propose decision-aid software for detecting a hostile swimmer or diver in an urban harbor or estuary. Two important problems in retail demand forecasting are estimating turned away demand when items are sold out and properly accounting for substitution effects among related items. For simplicity, most retail demand forecasts rely on time-series models of observed sales data, which treat each stock keeping unit (SKU) as receiving an independent stream of requests. However, if the demand lost when a customer's first choice is unavailable (referred to as “spilled” demand) is ignored, the resulting demand forecasts may be negatively biased; this underestimation can be severe if products are unavailable for long periods of time. Concurrently, stockout-based substitution will increase sales in substitute products that are available (referred to as “recaptured” demand); ignoring recapture in demand forecasting leads to an overestimation bias among the set of available SKUs. Correcting for both spill and recapture effects is important to establishing a good estimate of the true underlying demand for products. In “Estimating Primary Demand for Substitutable Products from Sales Transaction Data,” G. Vulcano, G. van Ryzin, and R. Ratliff propose a method for estimating substitute and lost demand when only sales and product availability data are observable, not all products are displayed in all periods (e.g., due to stock-outs or availability controls), and the seller knows its aggregate market share. The model combines a multinomial logit (MNL) choice model with a nonhomogeneous Poisson model of arrivals over multiple periods and uses the expectation-maximization (EM) method to compensate for incomplete observations of primary demand; that is, the demand that would have been observed if all products had been available in all periods. The procedure is shown to be effective on both simulated and real-world data sets. As government intervention in the marketplace suffers increasing criticism in recent years, the authors analyze the impact of antitrust laws on the welfare of society. Although it is well known that in a perfectly competitive market, free competition is optimal for society, in practice many markets are oligopolies: they are controlled by only a few price-setting firms. In such an environment, free competition does not necessarily lead to an optimal social outcome. In “Generalized Quantity Competition for Multiple Products and Loss of Efficiency,” J. Kluberg and G. Perakis evaluate the impact of competition on both social welfare and firms' profit. For a market in which firms compete for market share, they compute the social benefit and the cost for the firms of enforcing competition as a function of market characteristics such as the number of competing firms and the intensity of competition. Cargo carriers face the same fundamental trade-off as passenger carriers. Given a request for a certain amount of capacity at a certain price, should the carrier accept this request to generate revenue or keep the capacity for a more profitable request that may arrive in the future? Although the fundamental trade-off is the same, management of cargo capacity brings unique challenges when compared with management of passenger capacity. To begin with, cargo capacity is measured in multiple dimensions such as weight, volume, and location of the cargo in the body of the aircraft. Furthermore, the amount of capacity that will eventually be used by an accepted cargo request is usually not known until the departure time of the aircraft, resulting in an additional source of uncertainty. Finally, carriers do business through periodically signed contracts that can generate a steady stream of cargo from large customers and through the spot market that can be somewhat unpredictable. In “Cargo Capacity Management with Allotments and Spot Market Demand,” Y. Levin, M. Nediak, and H. Topaloglu develop a model that allows a carrier to decide which portion of the capacity to sell through contracts and which portion of the capacity to reserve for the spot market. The model has both tactical and operational implications, as it can be used not only to make relatively infrequent contract decisions, but also to manage the cargo capacity reserved for the spot market on a day-to-day basis. Integrated supply chain design problems are normally modeled as complex integer or mixed-integer nonlinear optimization problems. In the past decade, column generation and Lagrangian relaxation methods have been the solution methods of choice for different versions of these problems. However, these algorithms are problem-specific and time consuming to design and implement. In “A Conic Integer Approach to Stochastic Joint Location-Inventory Problems,” A. Atamtürk, G. Berenguer, and Z.-J. Shen suggest conic programming as an alternative approach for solving these problems. Conic optimization has been employed in a wide arrange of areas, including basic uncapacitated facility location problems, but this paper is the first to apply conic integer programming formulations to various supply chain design problems. The advantage of the new approach is that it not only provides a more general modeling framework but also leads to fast solution t

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.