Knowledge that Transforms

We study a two‐stage decision problem, namely, the allocation and deployment of reserved inventories (RIs) in a supply network with random demand surges. The demand surge follows a time‐dependent stochastic process and our objective is to minimize the expected total unmet demand in the presence of positive transshipment lead times. We first solve the optimal deployment problem given that the demand surges have occurred at some locations. We show that the optimal deployment policy is a “nested” policy with respect to the shadow price at each location, where a shadow price represents the marginal reduction of the expected total unmet demand due to a marginal increase of RIs. Specifically, locations with higher shadow prices have higher priority in inventory allocation. We then consider the optimal allocation problem in the pre‐positioning stage. We show that under certain conditions the optimal allocation is increasing in the total amount of RIs. We introduce a new stochastic order for distributions defined on sets called the first‐order stochastic dominance and use it to show that the expected total unmet demand is higher when one of the following is true: the demand surges tend to occur simultaneously at more locations, the post‐surge delivery takes a longer time, more demand arrives earlier, or the demand has a higher volatility.

We investigate the impact of behavioral ordering on profits under competition. Specifically, we use controlled laboratory experiments to evaluate the differences in profits between a behavioral competitor (where a human places orders), and a management science‐driven competitor (where orders are placed according to one of several plausible policies based on existing literature and managerial practice). Unlike the full‐information game‐theoretic models that assume rational decision‐makers, these policies mimic practical situations by using less information and do not assume that their human competitors make fully rational decisions. Most prior literature focuses on non‐competitive settings, where behaviorally biased deviations from optimal order quantities result in small expected profit losses. In contrast, under competition, we find that human decision‐makers receive a substantially lower profit than the equilibrium expected profit, even as their competitors receive substantially higher profit.

The Keogh Report of 2013 proposed a major reconfiguration of the accident and emergency (A&E) system under National Health Service (NHS) England to improve service. The proposed reconfiguration includes centralized facilities with multiple specialties as well as small local minor‐injury facilities. We use stylized queuing models to analyze cost and service implications of the proposed reconfiguration. We find that increasing numbers of specialty patients that require admission to hospital makes splitting off specialty A&Es from general ones more attractive. The same applies for patients with minor injuries. Our work generally supports the reconfiguration recommended in the Keogh report but with some fine‐tuning: For instance, a merger of A&Es (pooling) does not always make sense even though it increases patient numbers when the patients in the two A&Es are of different types. We provide simple quantitative rules to indicate whether the proposed reconfiguration could lower costs in any particular region of the country. The results here are consistent with some NHS England providers attempting specialty A&Es for geriatric patients and mobile drunkenness treatment centers on weekends. Our rules and approach can be useful for identifying candidate reconfiguration opportunities not only for NHS England but also for any other context where pooling and arrival heterogeneity are important considerations.

Our study contributes to the scholarly debate whether organizational units should have a narrow focus and admit a homogeneous patient cluster or whether they should admit a pool of patient clusters. We investigate whether the benefits of increased volume through pooling patients outweigh the disadvantages of increased heterogeneity and pursue our analysis in the context of neonatal care. Our empirical studies relies on 4020 patient episodes collected in 18 German neonatal intensive care units and we distinguish between two patient clusters that differ with respect to the inherent medical risk and operational heterogeneity. Cluster 1 consists of very‐low birth weight (VLBW) infants with increased risk of complications but similar service trajectories and lower operational heterogeneity. Cluster 2 contains non‐VLBW infants with lower risk of complications but more diversity in disease patterns and higher operational heterogeneity. Our analysis shows that cluster volume, that is, the unit's absolute patient volume in a cluster, is positively related to process outcomes as indicated by decreasing length of stay. This relationship is found for both clusters. Regarding focus, we do not find any evidence of positive effects. In fact, we even find that cluster focus, that is, the unit's relative volume of the cluster, is detrimentally related to process outcomes for non‐VLBW patients with lower risk of complications and more operational heterogeneity. This indicates that organizational units providing services for complex patients should not have a narrow focus, but should rather provide services for related patient clusters in order to achieve higher volume levels within the unit.

This study investigates a novel mechanism—multiple‐winner award rules—that are widely used in e‐procurement auctions and crowdsourcing sites. In many e‐procurement auctions, the auctioneer (i.e., the buyer) specifies three rules before the auction starts: (i) the size of the finalist set (from which the winner[s] will be chosen); (ii) the number of winners; and (iii) the allocation of the contract among the winners. We examine how these three rules affect auction performance using a dataset of online procurement auctions across a variety of product categories. We find that the multiple‐winner award rules significantly impact the suppliers’ participation decisions, which is an important factor in determining the economic performance of the auction (i.e., buyer's savings). Most interestingly, these three rules systematically induce opposite effects on auction participation for two types of suppliers: experienced and inexperienced bidders. For example, increasing the number of winners encourages experienced suppliers, but discourages inexperienced suppliers from participating in the auction. On the other hand, raising the disparity in the contract allocation among winning bidders (e.g., from 50/50 to 90/10 split) deters experienced suppliers, but motivates inexperienced suppliers to participate. These findings provide guidelines for industrial buyers and crowdsourcing hosts on how to effectively make use of multiple‐winner design levers to promote suppliers’ participation when designing procurement auctions and crowdsourcing contests.

This study considers the scheduling of limited resources to a large number of jobs (e.g., medical treatment) with uncertain lifetimes and service times, in the aftermath of a mass casualty incident. Jobs are subject to triage at time zero, and placed into a number of classes. Our goal is to maximize the expected number of job completions. We propose an effective yet simple index policy based on Whittle’s restless bandits approach. The problem concerned features a finite and uncertain time horizon that is dependent upon the service policy, which also determines the decision epochs. Moreover, the number of job classes still competing for service diminishes over time. To the best of our knowledge, this is the first application of Whittle’s index policies to such problems. Two versions of Lagrangian relaxation are proposed in order to decompose the problem. The first is a direct extension of the standard Whittle’s restless bandits approach, while in the second the total number of job classes still competing for service is taken into account; the latter is shown to generalize the former. We prove the indexability of all job classes in the Markovian case, and develop closed‐form indices. Extensive numerical experiments show that the second proposal outperforms the first one (that fails to capture the dynamics in the number of surviving job classes, or bandits) and produces more robust and consistent results as compared to alternative heuristics suggested from the literature, even in non‐Markovian settings.

The purpose of this paper is to investigate the learning from rare events and the knowledge management process involved, which presents a significant challenge to many organizations. This is primarily attributed to the inability to interpret these events in a systematic and “rich” manner, which this paper seeks to address. We start by summarizing the relevant literature on humanitarian operations management (HOM), outlining the evolution of the socio‐technical disaster lifecycle and its relationship with humanitarian operations, using a supply chain resilience theoretical lens. We then outline theories of organizational learning (and unlearning) from disasters and the impact on humanitarian operations. Subsequently, we theorize the role of middle managers in humanitarian operations, which is the main focus of our paper. The main methodology incorporates a hybrid of two techniques for root cause analysis, applied to two related case studies. The cases were specifically selected as, despite occurring twenty years apart, there are many similarities in the chain of causation and supporting factors, potentially suggesting that adequate learning from experience and failures is not occurring. This provides a novel learning experience within the HOM paradigm. Hence, the proposed approach is based on a multilevel structure that facilitates the operationalization of learning from rare events in humanitarian operations. The results show that we are able to provide an environment for multiple interpretations and effective learning, with emphasis on middle managers within a humanitarian operations and crisis/disaster management context.

We consider a company that schedules the tasks of its projects to maximize their expected net present value (ENPV) when tasks may fail. The failure of any task terminates the project immediately. We show that for projects with certain decreasing failure rates, the ENPV optimization problem can be solved using a linear program. The main focus of our work is on how constant task failure rates contribute to decreasing project risk as tasks are completed. Under constant task failure rate, earlier completion of a task improves its probability of success and the risk profile of the project. However, it may also accelerate costs, which worsen discounted cash flow. We show the equivalence of cash flow discount rate and failure rate. Further, if task failures are independent, then the failure rates are additive. We develop a model that (a) recognizes the reduction in project risk when a task is completed, (b) implements this risk reduction into the ENPV calculation, and (c) permits optimization of the ENPV through sequencing and timing decisions for the tasks. We design an algorithm to solve the problem optimally. This enables us to validate the contributions of our work using two computational studies. The first study demonstrates a significant increase in maximum project ENPV from improved project scheduling. The second study demonstrates a significant increase in total project portfolio value as a result of better informed project selection. Our work motivates companies to develop more precise information about the failure risks of their project tasks.

We consider inventory management decisions when manufacturing and warehousing are controlled by independent entities. The latter possess private information that affects their choices and are allowed to communicate via a mediator who attempts to streamline their decisions without restricting their freedom. The mediator designs a mechanism based on quantity discounts to minimize the overall system costs, attempting to reach a win–win situation for both entities. Using the Revelation Principle, we show that it is in the entities’ self‐interest to reveal their information and we prove that coordination is attainable even under bilateral information asymmetry. The acceptable cost allocation is not unique, providing adequate flexibility to the mediator during mechanism design; the flexibility may reflect the relative power of the entities and is quantified in our work by a series of computational experiments. Our approach is motivated by inventory management practices in a manufacturing group and, thus, it is directly applicable to real‐life cases.

We examine inventory decisions in a multiperiod newsvendor model. In particular, we analyze the impact of budget cycles in a behavioral setting. We derive optimal rational decisions and characterize the behavioral decision‐making process using a short‐sightedness factor. We test the aforementioned effect in a laboratory environment. We find that subjects reduce order‐up‐to levels significantly at the end of the current budget cycle, which results in a cyclic pattern during the budget cycle. This indicates that the subjects are short‐sighted with respect to future budget cycles. To control for inventory that is carried over from one period to the next, we introduce a starting‐inventory factor and find that order‐up‐to levels increase in the starting inventory.