Lauren B. Davis

Radio frequency identification system optimisation models for lifecycle of a durable product

We address the implementation of radio frequency identification (RFID) technology to support and manage durable products over their entire lifecycle with a focus on optimising the RFID tagging system. We develop general models that optimise the placement of RFID tags on an end product and its components, the allocation of the data on tags, and the selection of RFID tags and their configuration. The primary criteria for optimisation are to maximise the value of RFID data, and to minimise the total cost of the RFID tag system. The total cost of RFID tags comprises costs associated with memory capacity, tag type, software acquisition and maintenance, and number of tags used; the total value for a piece of data is determined by the data usage frequency and the fixed and variable value of data per use across the product lifecycle. The tag placement and data allocation model involves making a decision about these two different objectives–cost and value, which we develop in the form of a multi-objective optimisation problem. The tag selection model aids in the proper selection of RFID tags and their configuration for individual components for a given end product. Given a durable products characteristics along with its bill-of-materials, a system of RFID tags can efficiently be determined for the product and its components, to help manage and support its lifecycle.

Estimating available supermarket commodities for food bank collection in the absence of information

Food banks collect usable, yet unsellable commodities donated by supermarkets.The amounts of each donated food type is not known prior to collection.Four approximation methods are compared to determine which provides best estimates.The impacts of approximation methods on transportation costs is also evaluated.Neural networks provide good estimations for on-hand food and transportation costs. Food insecurity is a widespread concern in the United States. Addressing this concern is a chief goal of many non-profit organizations including food banks. Understanding the availability of donations is beneficial when addressing the demand in local communities, especially when their collection requires food bank-managed vehicles. High-volume donors such as supermarkets, however, do not provide information in regards to what items are available. This can negatively impact inventory management capabilities and cause unnecessary transportation costs.This research evaluates four approximation methods based on their ability to estimate food availability at supermarkets including the multiple layer perceptron artificial neural network, multiple linear regression, and two naive estimates for the average collection amount. Using a subset of the historic data provided by the Food Bank of Central and Eastern North Carolina (FBCENC), the four approximation methods are evaluated in terms of their ability to estimate collection amounts in the next planning period. Transportation cost estimates are then calculated using projections made using each approximation method and compared to those calculated using the actual transportation costs. Results suggest that the MLP-NN models provide the best approximations for each food type and provide closer estimations for transportation cost than other approximation methods.

Pre?positioning commodities to repair maritime navigational aids

Purpose – As the intensity of natural disasters increases, there is a need to develop policies and procedures to assist various agencies with moving aid to affected areas. One of the biggest limitations to this process is damage to transportation networks, in particular waterways. To keep waterways safe, aids to navigation (ATONs) are placed in various areas to guide mariners and ships to their final destination. If the ATONs are damaged, then the waterways are left unsafe, making it difficult to move supplies and recover from a disaster. The aim of this paper is to explore the effectiveness of pre‐positioning strategies for port recovery in response to a natural disaster.Design/methodology/approach – A stochastic facility location model is presented to determine where teams and commodities should be pre‐positioned in order to maximize the number of ATONs repaired, given a constraint on response time. The first stage decisions focus on determining the location of resources. The second stage decisions cons...

Achieving Equity, Effectiveness, and Efficiency in Food Bank Operations: Strategies for Feeding America with Implications for Global Hunger Relief

One in six Americans (14.3 % of households) reported being food insecure at some time during the year 2013 (i.e., they lacked access to enough food for an active, healthy life for all household members). This translates to 17.5 million food insecure households and 49.1 million Americans, 33.3 million adults and 15.8 million children living in food insecure households (Coleman-Jensen et al. 2011, 2014). This slight decrease from 14.5 % in 2012 was not statistically significant and marks the third consecutive year that the USDA’s annual hunger survey has found food insecurity at some of the highest levels since the government started the report in 1995.

Optimization Model For Distributed Routing For Disaster Area Logistics

The problem of transportation in a disaster area can be seen broadly as having two aspects: (a) moving people and materials out of an area, and (b) moving people and materiel into the same area. The common thread here is the use of a limited set of surface and air transportation gateways into and out of the area. The distributed routing problem here is that of assigning one from among the limited set of gateways to various transportation requests in real-time, while maximizing some measure of success for the entire relief mission. We define the general problem, provide an example of the real-time transportation routing problem, and propose an optimization model. The similarity of this problem to a job shop scheduling problem is presented pointing to the fact that a distributed approach based on bio-inspired methods can be developed to counter the large problem size and centralized nature of the integer multi-commodity network model. A small illustrative model of the integer multi-commodity network model is presented and solved

Empirical analysis of volunteer convergence following the 2011 tornado disaster in Tuscaloosa, Alabama

Volunteer convergence refers to the mass movement of volunteers toward affected areas following disaster events. Emergency management professionals sometimes refer to volunteer convergence as “the disaster within the disaster,” which is an indicator of the tremendous challenge that managing the post-disaster influx of spontaneous volunteers presents. In order to develop effective strategies for managing volunteer convergence, it is imperative that emergency managers and coordinators understand the nature of convergence from a quantitative perspective. This paper presents a case study of volunteer convergence following the April 2011 tornado disaster in Tuscaloosa, Alabama, and represents the first academic study to rigorously analyze volunteer convergence data. Specifically, we characterize selected stochastic variables that are relevant to volunteer task assignment within the context of a disaster relief warehouse environment using data collected during tornado relief efforts in May 2011. Time series analysis and a hierarchical clustering method based on the Kruskal–Wallis test revealed both non-stationarity and non-homogeneity in the data with respect to time of day, day of the week, and number of weeks past the disaster event. We also discuss the implications of our findings with respect to modeling relief center convergence as a queuing system.

Impact of batch appointments on no-show rates: a public vs private clinic perspective

Much of the appointment-scheduling literature has characterized the impact of no-show rates with respect to individual appointment requests. However, little is known about the impact of appointments that are grouped by household. This study is concerned with understanding the prevalence of these family group appointments and identifying the appointment characteristics that significantly influence no-show rates. Using historical data from one public and one private pediatric clinic, multiway frequency analysis is used to characterize the association between appointment characteristics and appointment size. A logistic regression model is also developed to identify the factors that contribute to no-show rates for both clinics. The results of the study indicate that more than one-third of the appointments scheduled by both clinics were associated with batch appointment requests. In addition, no-show rates for batch appointments in the public clinic were higher than those of the private clinic. We identify appointment size as a significant predictor in determining no-show rates for both public and private clinics. Since no-show rates adversely impact clinic efficiency, scheduling coordinators should consider the impact of grouping-related appointments when determining how to best allocate resources.

A Markov decision process model for equitable distribution of supplies under uncertainty

Many individuals suffering from food insecurity obtain assistance from governmental programs and nonprofit agencies such as food banks. Much of the food distributed by food banks come from donations which are received from various sources in uncertain quantities at random points in time. This paper presents a model that can assist food banks in distributing these uncertain supplies equitably and measure the performance of their distribution efforts. We formulate this decision problem as a discrete-time, discrete state Markov decision process that considers stochastic supply, deterministic demand and an equity-based objective. We investigate three different allocation rules and describe the optimal policy as a function of available inventory. We also provide county level estimates of unmet need and determine the probability distribution associated with the number of underserved counties. A numerical study is performed to show how the allocation policy and unmet need are impacted by uncertain supply and deterministic, time-varying demand. We also compare different allocation rules in terms of equity and effectiveness.

A Markov chain model for quantifying consumer risk in food supply chains

According to the Centers for Disease Control and Prevention, approximately 48 million people experience foodborne illnesses per year. The majority of the illnesses are attributed to the presence of bacteria in food products. However, there is some concern about the likelihood of food contamination resulting from intentional acts of sabotage. This research presents a stochastic model to quantify food supply chain vulnerability in terms of the number of people who become ill from consuming a contaminated food product. We specifically use a discrete time, discrete state Markov chain model with rewards and estimate consumer illness by product and distribution channel. The results of our computational study show the relationship between purchasing behavior, product shelf life, and consumer risk. We propose a classification scheme that can be used to categorize the level of vulnerability among different food distribution channels. We also show the impact of purchasing behavior on the speed with which the products are sold at each distribution channel. The proposed model has the potential to provide insight into timely interventions and influence how intervention policies would need to be tailored to each distribution channel in the event that a chemical contamination occurs.

Tactical allocation and acceptance of multiple patient classes in the presence of no-shows

Clinics that provide pediatric care are frequently confronted with family group appointment requests, where parents desire their children to be scheduled simultaneously or consecutively. This is potentially beneficial to the family by minimizing the number of trips to the provider’s office. However, offering prescheduled group appointments have the risk of reducing provider utilization, particularly if the entire group fails to meet their scheduled appointment. Similarly, reserving appointment slots for same day group appointment requests may also decrease utilization and impact profitability. This paper explores the impact of family group appointments on clinic performance in terms of provider utilization and profit. A finite-horizon, stochastic dynamic programming problem is presented to determine the optimal scheduling strategy given both individual and group appointment requests can be tactically accommodated via overbooking. On the basis of a computational study, we quantify the risk to clinic profitability and productivity resulting from the no-show behavior of prescheduled appointments. We also characterize the behavior of the optimal scheduling strategy as a function of prescheduled appointment allocations among the patient classes.