The goal of pandemic response is to provide the greatest protection, for the most people, in the least amount of time. Short response times minimize both current and future health impacts for evolving pathogens that pose global threats. To achieve this goal, efficient and effective systems are needed for distributing and administering vaccines, a cornerstone of pandemic response. COVID-19 vaccines were developed in record time in the U.S. and abroad, but U.S. data shows that they were not distributed efficiently and effectively once available. In an effort to "put vaccines on every corner", pharmacies and other small venues were a primary means for vaccinating individuals, but daily throughput rates at these locations were very low. This contributed to extended times from manufacture to administration. An important contributing factor to slow administration rates for COVID-19 was vaccine transport and storage box size. In this paper, we establish a general system objective and provide a computationally tractable approach for allocating vaccines in a rolling horizon manner optimally. We illustrate the consequences of both box size and the number and capacity of dispensing locations on achieving system objectives. Using U.S. CDC data, we demonstrate that if vaccines are allocated and distributed according to our proposed strategy, more people would have been vaccinated sooner in the U.S. Many additional days of protection would have occurred, meaning there would have been fewer infections, less demand for healthcare resources, lower overall mortality, and fewer opportunities for the evolution of vaccine-evading strains of the disease.

Effective disaster response requires time-critical decisions to get personnel and supplies to the right places as quickly as possible. Such operations are complicated by the need for coordination among multiple stakeholders. In this teaching brief, we describe a serious online game for humanitarian logistics courses called the "Disaster Response Game." The game provides students with context, challenges, and command seat experience with simulated life-and-death consequences. Our development effort includes the creation of a flexible digital platform for delivering game-style simulations for undergraduate and graduate education as well as response training within large enterprises. The system has instructional design interfaces for creating a wide range of disaster scenarios including, but not limited to, earthquakes, floods, pandemics, storms, and wildfires. As proof of concept, we built a tropical cyclone scenario and tested the corresponding multiplayer Disaster Response Game with students and subject matter experts. Students found it valuable to be assigned different roles that required collaboration, found the experience to be realistic, and were motivated to minimize casualties as best as they could with available resources. Subject matter experts also found the game to be sound in its instructional design. We seek a community of interested instructors to help develop more scenarios.

We reformulate the problem of determining support vectors directly as an application of Bayes' classifiers rather than as the dual program to a binary geometric separation problem. The primary purpose of the reformulation is to create a simpler exposition of the support vector machines technique. A secondary advantage is that it immediately and naturally applies to multi-class classification problems where the kernel function can be normalized as a density.

When planning responses to pandemics, public policies should be based on appropriate scientific fields of study. Accurate application of non-pharmaceutical interventions rests on knowledge of pathogens’ biology and pathogenesis; similarly, pharmacological responses to pandemics policies should be designed to adhere to supply chain management principles. The COVID-19 pandemic has spurred efforts to strengthen the scientific foundations for public health response options for this and future outbreaks. For example, policies relating to school or work closure, among others, have been subjected to rigorous analysis using dynamic infectious disease models. Our focus in this paper is on the design and optimal operation of a vaccine allocation system based on fundamental principles that have a long pedigree in operations research. We indicate what attributes a vaccine supply chain should have and how capacities for each element of the supply chain impact, over time, inventory requirements and their allocations. We show that maximizing vaccination person-days is optimal, minimizing dispensing locations maximizes vaccination person-days, and optimal allocations can be made using an approach presented herein.

Fresh produce supply chains present variable and diverse conditions that are relevant to food quality and safety because they may favor microbial growth and survival following contamination. This study presents the development of a simulation and visualization framework to model microbial dynamics on fresh produce moving through postharvest supply chain processes. The postharvest supply chain with microbial travelers (PSCMT) tool provides a modular process modeling approach and graphical user interface to visualize microbial populations and evaluate practices specific to any fresh produce supply chain. The resulting modeling tool was validated with empirical data from an observed tomato supply chain from Mexico to the United States, including the packinghouse, distribution center, and supermarket locations, as an illustrative case study. Due to data limitations, a model-fitting exercise was conducted to demonstrate the calibration of model parameter ranges for microbial indicator populations, i.e., mesophilic aerobic microorganisms (quantified by aerobic plate count and here termed APC) and total coliforms (TC). Exploration and analysis of the parameter space refined appropriate parameter ranges and revealed influential parameters for supermarket indicator microorganism levels on tomatoes. Partial rank correlation coefficient analysis determined that APC levels in supermarkets were most influenced by removal due to spray water washing and microbial growth on the tomato surface at postharvest locations, while TC levels were most influenced by growth on the tomato surface at postharvest locations. Overall, this detailed mechanistic dynamic model of microbial behavior is a unique modeling tool that complements empirical data and visualizes how postharvest supply chain practices influence the fate of microbial contamination on fresh produce.IMPORTANCE Preventing the contamination of fresh produce with foodborne pathogens present in the environment during production and postharvest handling is an important food safety goal. Since studying foodborne pathogens in the environment is a complex and costly endeavor, computer simulation models can help to understand and visualize microorganism behavior resulting from supply chain activities. The postharvest supply chain with microbial travelers (PSCMT) model, presented here, provides a unique tool for postharvest supply chain simulations to evaluate microbial contamination. The tool was validated through modeling an observed tomato supply chain. Visualization of dynamic contamination levels from harvest to the supermarket and analysis of the model parameters highlighted critical points where intervention may prevent microbial levels sufficient to cause foodborne illness. The PSCMT model framework and simulation results support ongoing postharvest research and interventions to improve understanding and control of fresh produce contamination.

We develop a computationally efficient optimization procedure to optimize stock and rationing levels for a model consisting of a single product with two priority–demand classes, given by mutually independent, stationary, Poisson demand processes. Each priority class has its own service levels requirements, defined by the class-specific fill rate and expected waiting-time levels. Order lead times are independent and identically distributed random variables. This is the first study in this setting to consider both waiting-time constraints along with fill rate requirements.

We consider a one-warehouse, N-retailer, multiperiod, stock allocation problem in which holding costs are identical at each location and no stock is received from outside suppliers for the duration of the planning horizon. No shipments are allowed between retailers. The only motive for holding inventory at the central warehouse for allocation in future periods is the so-called risk pooling motive. We apply robust optimization to this problem extending the inventory policy to allow for an adaptive, nonanticipatory shipment policy. We consider two alternatives for the uncertainty set, one in which risk pooling is implicit and another for which risk pooling is explicit. The explicit risk pooling uncertainty set grows by no more than the square of the number of retailers. The general problem can be solved using Benders’ decomposition. A special case gives rise to closed-form solutions for both uncertainty set alternatives. The explicit risk pooling uncertainty set leads to a square root law in which the optimal stock to reserve at the central warehouse grows with the square root of the number of retailers. The experimental results confirm the value of the robust optimization approach and provide managerial insights into the operation of such systems. This paper was accepted by Yinyu Ye, optimization.

In this article we develop a procedure for estimating service levels (fill rates) and for optimizing stock and threshold levels in a two-demand-class model managed based on a lot-for-lot replenishment policy and a static threshold allocation policy. We assume that the priority demand classes exhibit mutually independent, stationary, Poisson demand processes and non-zero order lead times that are independent and identically distributed. A key feature of the optimization routine is that it requires computation of the stationary distribution only once. There are two approaches extant in the literature for estimating the stationary distribution of the stock level process: a so-called single-cycle approach and an embedded Markov chain approach. Both approaches rely on constant lead times. We propose a third approach based on a Continuous-Time Markov Chain (CTMC) approach, solving it exactly for the case of exponentially distributed lead times. We prove that if the independence assumption of the embedded Markov chain approach is true, then the CTMC approach is exact for general lead time distributions as well. We evaluate all three approaches for a spectrum of lead time distributions and conclude that, although the independence assumption does not hold, both the CTMC and embedded Markov chain approaches perform well, dominating the single cycle approach. The advantages of the CTMC approach are that it is several orders of magnitude less computationally complex than the embedded Markov chain approach and it can be extended in a straightforward fashion to three demand classes.

We analyze financial hedging tools for inventory management in a risk-averse corporation. We consider the problem of optimizing simultaneously over both the operational policy and the hedging policy of the corporation in a multi-product model. Our main contribution is a separation result such that for a corporation with multiple products and inventory departments, the inventory decisions of each department can be made independently of the other departments’ decisions. That is, no interaction needs to be considered among different products.

Numerical investigation of turbulent flow around a surface-mounted square cylinder of aspect ratio h/d = 4 is presented in this paper. The aims are to get detailed information about the flow structures around such a cylinder and to establish a suitable turbulent model that could yield accurate and reliable results for practical industrial applications. Calculations were performed at Reynolds number of 13041 using a Reynolds stress model. The present simulation has successfully reproduced the primary flow, as well as the three-dimensional large-scale vortex structure in the wake of this finite wall-mounted body. The existence of a spanwise vortex, tip vortex, and horseshoe vortex is consistent with previous experimental studies. The half-loop shed structure observed in the previous experimental study has been clearly captured in the isosurface of the instantaneous second invariant of the velocity gradient. Furthermore, the evolution of the vertical vorticity fields in the wake is presented with details at middle span (z/d = 2.0), and the Strouhal number estimated from flow visualization is 0.1, which not only equals to the value obtained by statistical analysis of the streamwise drag coefficient, but also perfectly matches with the value obtained by the previous experimental study carried out under a similar flow condition with the same geometry.

The Reynolds Stress Model (RSM) and Detached Eddy Simulation (DES) are used to study the turbulent flow around a surface-mounted square cylinder of aspect ratio h/d=4 at a Reynolds number of 13,041. The performances of the RSM and DES are evaluated by comparing their simulation results against experimental measurements. Both models successfully reproduced the primary flow, as well as the three-dimensional large-scale vortex structure in the wake of finite wall-mounted body. However, RSM produces better predictions in both mean velocity and root-mean-square velocity than DES. The Strouhal number obtained by statistical analysis of the streamwise drag coefficient on the top wall of the cylinder is 0.1 by RSM, which matches the value obtained by experimental study carried out under a similar flow condition with the same geometry.

We investigate the (S-1, S) inventory policy under stuttering Poisson demand and generally distributed lead time when the excess demand is lost. We correct results presented in Feeney and Sherbrooke's seminal paper [Feeney, G. J., C. C. Sherbrooke. 1966. The (S-1, S) inventory policy under compound Poisson demand. Management Sci.12(5) 391--411] and note that the stationary distribution of units on order for the general compound Poisson demand case is still an open question.

AbstractThis paper reports on the findings from interview research conducted by a joint team from the Systems Engineering Directorate at Corning Incorporated and the Systems Engineering Program at Cornell University to test for systems engineering (SE) effectiveness in product development in a commercial setting. Between April 2008 and March 2009, the team conducted 19 interviews of systems engineers and project managers within Corning to evaluate the extent to which they used a range of systems engineering techniques, and the effectiveness of those techniques in improving project performance. Both quantitatively and anecdotally, the expectation of a correlation was met, with strongly performing projects having generally higher use of SE, and struggling projects having difficulties that could be traced back to shortcomings in the use of SE. Both the findings and the underlying methodology are discussed, with the aim of interesting others in the field in repeating this type of research within other enterprises.

Both two-dimensional and three-dimensional numerical simulations of the turbulent flow through a staggered tube bundle are presented. The primary aim of the present study is to search for a turbulent model that could serve as an engineering design tool at a relatively low computational cost. In the present study, the performances of the Spalart-Allmaras model, the k-epsilon model, and large eddy simulation are evaluated by comparing their simulation results against experimental measurements. The turbulence models are assessed mainly based on their ability to resolve time-dependent features of the flow related to vortex shedding. Simulations are performed at a Reynolds number of 9300. Overall, the predicted streamwise mean velocity and transverse mean velocity in the present study are in good agreement with measurements, and the results show that the simple one-equation Spalart-Allmaras model could be a very promising tool for numerical simulation of complex turbulent flows, since the Strouhal number obtained by it agrees well with measurements available in the literature for similar tube geometries.

Our goal is to disseminate the systems engineering process to as broad an audience as possible. This audience includes freshmen engineers, students from non‐engineering majors, as well as working managers and staff from a host of different occupations. Reviewing the impediments to the dissemination effort and the success of the Six Sigma movement, we articulate the requirements that a curriculum for non‐engineers should satisfy and we propose a particular blended curriculum that satisfies these requirements and highlight its features. We point to three implementations of the curriculum. Experience with this form of the curriculum is still in its infancy.

Numerical investigation of flow through a U-shaped channel with a square cross section is presented in this paper. The primary aim of this study is to determine if FLUENT, a commercial computational fluid dynamics software package, is capable of providing a reasonable solution for this kind of flow configuration. Calculations were performed at Reynolds numbers of 109, 382, and 873 using large-eddy simulation. Simulation results indicate that the present study has successfully reproduced the primary flow as well as most of the flow features associated with the secondary flow. The existence of Dean vortices was consistent with previous experimental studies.

Time-based item fill rates, or “channel” fill rates, are the building blocks needed to evaluate steady-state compliance with time-based customer service agreements. Exact computation of channel fill rates is both difficult and time-consuming, yet their accurate assessment is essential for system-wide inventory optimization. We describe and validate a practical method for computing channel fill rates in a multi-item, multi-echelon service parts distribution system. A simulation study is presented which shows that, in a three-echelon setting, our estimation errors are very small over a wide range of base stock level vectors. A more accurate, though less efficient, approximation method is also evaluated for comparison.