Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TB75

5 - Simulation-based Optimization in Outpatient Appointment Scheduling

n TB76 West Bldg 212C Joint Session PSOR/Practice Curated: OR Applications in Humanitarian Logistics Sponsored: Public Sector OR Sponsored Session Chair: Hafizul Islam, North Carolina State University, Raleigh, NC, 27695, United States 1 - Modeling the Logistics Capabilities of Home Health Agencies to Provide In-home Dispensing to Vulnerable Populations During Public Health Emergencies Ashlea Bennett Milburn, University of Arkansas, 4207 Bell Engineering Center, Fayetteville, AR, 72701, United States, Emre Kirac, Charleen McNeill Public health emergencies may call for medical countermeasures dispensing to affected populations. Many people can visit established dispensing locations en masse, but this is impractical for vulnerable homebound populations. Home health agencies (HHAs) have potential to act as private dispensing entities in this situation, treating homebound persons and their families in their homes. This research establishes the comparative effectiveness of centralized and decentralized operational models of in-home dispensing via HHAs. Results are presented for a mixed urban/rural case study area. 2 - An Integer L-shaped Algorithm for the Integrated Location and Network Restoration Problem in Humanitarian Relief Ece Sanci, University of Michigan, Ann Arbor, MI, United States, Mark Stephen Daskin Prepositioning emergency relief items before an anticipated disaster is a common strategy to increase the effectiveness of relief distribution. In this study, we assume that relief distribution is hampered due to damaged roads. We present a two-stage stochastic programming model integrating facility location and network restoration decisions. We propose an integer L-shaped algorithm which deploys lower and upper bounds to solve our model efficiently. Our computational results show significant improvement in unmet demand and cost measures by integrating location and network restoration models. 3 - Dynamic Network Flow Problems Having ARC Setup Costs Rob Curry, Clemson University, Cole Smith Some network applications require dynamic flows to be transmitted according to a non-simultaneous schedule. In this talk, we consider a dynamic network flow problem that considers the presence of arc setup costs that may exist whenever an arc begins transmitting flow. This problem can be modeled by the minimum cost flow problem having arc-activation costs (MCF-A), in which an arc (i,j) is said to be activated on a path p when (i,j) is included on p but not on p-1. As an alternative to this mixed-integer programming approach, we employ a relaxation- based algorithm for obtaining upper and lower bounds that increases the number paths in a schedule, as needed. 4 - Robust Optimization Approaches for the Equitable and Effective Distribution of Donated Food Irem Sengul Orgut, Lenovo, 8209 Pritchett Farm Lane, Raleigh, NC, 27606, United States, Julie Simmons Ivy, Reha Uzsoy, Charlie Hale Motivated by our partnership with a local food bank, we present a robust optimization model to support the equitable and effective distribution of donated food over the service area. Our model addresses uncertainty in the amount of donated food counties can receive. Letting the capacity of each county vary within a range, the model seeks to maximize total food distribution while enforcing a user-specified level of robustness. We derive structural properties and develop an efficient exact solution algorithm. We illustrate our model using historical data obtained from our food bank partner, summarize the policy implications of our results and examine the impact of uncertainty on outcomes. 5 - A Dynamic Programming Approach for Equitable and Effective Distribution under Uncertain Supply Md Hafizul Islam, North Carolina State University, 111 Lampe Drive, Daniel Hall 373, Raleigh, NC, 27695, United States, Julie Simmons Ivy Food banks in the United States serve people in hunger need within their respective service regions with donated food. Food banks receive donations from various sources and distribute them in an equitable and effective manner. Uncertainty in food donations is a big challenge for food banks in planning the distributions of food donations for future periods. In this work, we present a Markov decision process model and a dynamic programming solution approach of the model to deal with the uncertain supply for a local food bank over a finite number of periods with an aim to make an equitable and effective distribution of donated food.

Payman Jula, Associate Professor, Simon Fraser University, Faculty of Business, Vancouver, BC, V5A 1S6, Canada This talk addresses the challenges of outpatient scheduling in multi-stage healthcare facilities. We consider stochastic service times, the availability and compatibility of resources, and the presence of a variety of patient types. The proposed methods are based on integrating simulation with mathematical programming, and meta-heuristics algorithms to achieve multi objectives of minimizing the waiting time of patients, the completion time of the facility, and the procedures cancellation. The performance of proposed approaches are analyzed an reported. n TB75 West Bldg 212B Operations Research at USAFA General Session Chair: Megan Muniz, USAFA, Monument, CO, 80132, United States 1 - Air Force Business Process Automation Mark J. Williams, USAFA The Air Force has numerous processes operated by humans where modern service management systems could streamline their execution. This research analyzes processes used throughout the Air Force to develop improved workflows, identify areas to automate, and implement them using a modern service management system. The overarching goals of this research effort are to streamline Air Force business processes into a single, consistent look and feel, eliminate duplicate licenses and associated licensing costs, save manpower through improved process automation, and improve business process tracking and transparency. 2 - Cadet Summer Research Program Drew X. Richardson, Colorado Springs, CO, 80920, United States At the United States Air Force Academy, elite students are chosen to participate in the competitive Cadet Summer Research Program. What makes the program selective is that a cadet can only participate between his/her junior and senior year. Those selected are granted the opportunity to do research at military installations such as the Pentagon or Hickam Air Force Base, Hi, as well as work with non-DoD agencies such as the NSA or the DHS. In addition, cadets can also work for defense contractors such as the Institute for Defense Analysis or with companies like Facebook. This talk provides an overview of the selective CSRP process, and a snapshot at some of the high quality research performed by USAFA cadets. 3 - Analysis From the United States Air Force Academy’s Operations Research Capstone Course Gregory Steeger, USAFA, 16932 Park Trail Drive, Monument, CO, 80132, United States At USAFA, OR Majors take a year-long Capstone. In this culminating course, students analyze and solve problems faced by real clients, using the analytical skills they have acquired over the previous three years. Last year, 36 students worked in teams of four for nine clients. Clients included a tissue bank; Air Force Reserve Command Recruiting Service; Air Force Special Operations Command; Air Mobility Command; the Children’s Hospital of Colorado; Lockheed Martin; the National Geospatial Intelligence Agency; Sandia National Labs; and USAF Personnel. This talk provides an overview of the nine problems these 36 students analyzed and discusses the results they delivered to their clients. 4 - Optimal Configurations for Mountain Switchbacks Kristopher Pruitt, United States Air Force Academy, 2 Grass Song Place, Monument, CO, 80132, United States The purpose of this research is to determine the configuration (number, length, and gradient) of paths (i.e., switchbacks) that minimizes the time or energy required for a hiker to ascend a mountain face. We present a mathematical model of the possible path configurations and nonlinear optimization problems which select the optimal paths. The model is parameterized according to the dimensions of the mountain face and the physiological capabilities of the hiker. Subject to these parameters, the optimization problems return the path-gradient and length required for a hiker to minimize either the time or energy to ascend the face. 5 - Faculty Summer Research Megan Muniz, USAFA, CO Springs, CO, United States Currently, the Air Force (AF) is experiencing a pilot shortage, where the operational tempo’s demand for combat pilots exceeds supply. Retention efforts to date have been insufficient to solve this problem. Pilot Training Next (PTN) is the AF’s experimental investigation into two possible responses to the current pilot shortage: increasing Undergraduate Pilot Training throughput and increasing the pool of qualified pilot candidates. A group of pilots will go through a modified training syllabus; IDA and USAFA faculty member is analyzing the data to determine what factors facilitated more efficient learning and skill acquisition with the aim of creating a model of pilot training effectiveness.

302