2015 Informs Annual Meeting

WE49

INFORMS Philadelphia – 2015

WE49 49-Room 105B, CC Facility Location II Contributed Session

3 - A Structural Estimation of the Bullwhip Effect using Supply Network Data Vishal Gaur, Cornell University, 321 Sage Hall, Ithaca, NY, 14850, United States of America, vg77@cornell.edu, Maximiliano Udenio, Jan Fransoo We estimate a two-stage structural model of inventory decisions using financial data for 6,040 unique supplier-customer dyads for the years 1984-2013 to investigate downstream inventory adjustments and their influence on upstream firms. Our results show that inventory cost ratios are dynamic, and a significant cause of the bullwhip effect in supply chains.

Chair: Utku Kunter, Research Assistant, Middle East Technical University, ODTU Kampusu Endustri Muhendisligi Bolum, Oda 325 Cankaya, Ankara, 06801, Turkey, kunter@metu.edu.tr 1 - Identifying Stadium Locations for the Qatar 2022 Fifa World Cup Agha Iqbal Ali, Professor And Chairman, University of Massachusetts-Amherst, Isenberg School of Management, 121 Presidents Dr., Amherst, MA, United States of America, aiali@isenberg.umass.edu, Ahmed Ghoniem The number of venues for the twenty FIFA World Cups since 1930 has ranged from 3 to 20 and the number of participating teams has grown from 13 to 32. Due to the compactness of Qatar, the logistical impacts of different numbers and locations of venues can be significant for the Qatar 2022 FIFA World Cup. These impacts are studied in a scenario analysis using mathematical programming. 2 - Modeling Transportation Cost Uncertainty in City Logistics Systems Utku Kunter, Research Assistant, Middle East Technical University, ODTU Kampusu Endustri Muhendisligi Bolum, Oda 325 Cankaya, Ankara, 06801, Turkey, kunter@metu.edu.tr, Cem Iyigun, Haldun Sural City Logistics (CL) deals with freight transportation in urban areas. We consider the fixed charge facility location problem under transportation cost uncertainty in a CL setting. The optimization problem has been framed as a two-stage stochastic MIP. Locations are chosen in the first stage; capacity allocations are made in the second. We propose an evolutionary solution algorithm with local search. The algorithm performs well when compared with exact solution methods in the literature. 3 - Optimizing Hospital Closures H A Eiselt, University of New Brunswick, Fredericton, NB, Canada, haeiselt@unb.ca, Joyendu Bhadury, Mark L. Burkey, Hunkar Toyoglu We examine the effects of hospital closures on some major indicators concerning service level, including the average access time and the proportion of potential patients who are able to access a health care facility within 30 minutes driving time. Results of extensive computations are provided and discussed. WE50 50-Room 106A, CC Empirical Studies in Supply Chain Risk Management Sponsor: Manufacturing & Service Operations Management Sponsored Session Chair: Jun Li, Assistant Professor, Ross School of Business, University of Michigan, 701 Tappan St, Ann Arbor, 48103, United States of America, junwli@umich.edu 1 - Supply Chain Network Structure and Firm Returns Wu Jing, University of Chicago Booth School of Business, 5050 S Lake Shore Dr 3417S, Chicago, IL, United States of America, wujing@chicagobooth.edu, John Birge Using data on the relationships of public US firms, we investigate the effects of supply chain connections on firm performance, as reflected in stock returns, at two interaction levels, first-order from direct connections and second-order from systemic exposures through the network. 2 - An Empirical Model of Inventory Rationing Game Robert Bray, Northwestern University, Evanston, IL, United States of America, robertlbray@gmail.com, Oliver Yao, Achal Bassamboo We study inventory rationing in a one-DC, 80-retailer supply chain. We model the stores’ inventory policies as a dynamic discreet choice game. We estimate whether the stores increase or decrease their order quantities when the threat of DC stock out increases. Increasing order quantities indicate inventory gaming—- stores squabbling over upstream inventories—-whereas decreasing order quantities indicate supply chain coordination—-stores preserving upstream inventories for those most in need.

WE51 51-Room 106B, CC Facilities Planning and Design Contributed Session

Chair: Begün Efeoglu, Research Assistant, Middle East Technical University, ODTÜ Kampüsü Endüstri Mühendisligi, Oda 324 Çankaya, Ankara, 06800, Turkey, begunefeoglu@gmail.com 1 - A Markov Decision Process Model for the Dynamic Block Stacking Problem Hueon Lee, PhD Student, University of Arkansas, Dept. of IE, 4207 Bell Engineering Center, 1 University of Arkansas, The block stacking problem involves determining the depth of a storage row for unit loads that minimizes the sum of space and travel cost. A conventional block stacking problem assumes static row depths and deterministic demand. We allow row depths to change (by relocating product) and demand to be random. The problem is formulated as a Markov decision process and the optimal lane depth is determined for each inventory level. Results obtained provide useful insights for designers. 2 - A Heuristic Method for Storage Location Assignment Problem for a Distribution Center Zeynep Turgay, Solution Architect, Migros IT, Atasehir, Istanbul, 34758, Turkey, zeynept@migros.com.tr, Necati Aras Storage location assignment of products is an important research topic in warehouse design. We solve storage location assignment problem for a distribution center providing day-to-day service for a retail chain. Ordered items and quantities are random. Order pickers collect the products ordered by the stores from their addresses and deliver to the order shipment area. The objective is to optimize the total effort spent by the order pickers in terms of the total traveling distance. 3 - Impact of Demand Stochasticity on Distributed Layouts Begön Efeoglu, Research Assistant, Middle East Technical University, ODTÜ Kampüsü Endüstri Mühendisligi, Oda 324 Çankaya, Ankara, 06800, Turkey, begunefeoglu@gmail.com, Haldun Sural, Melih Celik The facility layout problem is to assign departments to locations. Recent studies show that functional layout configurations do not meet the needs of multiproduct enterprises. We investigate the effect of the stochastic demand and the relayout cost on the choice of a layout type in a dynamic environment. Using a two-stage stochastic integer programming, we simulate its results in order to study system trade-off and present our experimental results. Fayetteville, AR, 72701, United States of America, hueonlee@uark.edu, John A. White, Shengfan Zhang

WE52 52-Room 107A, CC Productivity and Competitiveness Contributed Session

Chair: Misuk Lee, Assistant Professor, Salisbury University, Perdue School of Business, Salisbury, MD, 21804, United States of America, mxlee@salisbury.edu 1 - Two Sides of the Same Coin: A Meta-analysis of Cohesion in Organizational Contexts Angelo Solarino, City Univeristy of Hong Kong, Chee Avenue, Cohesion is one of the most important determinants of group performance; it is reflected through a socio-emotional (i.e., group members’ liking for one another) and a task-oriented (i.e., group members’ shared task commitment) dimension. Exclusively focusing on organizational contexts we conduct a meta-analyses of the cohesion-performance relationship to clarify the effect of individual dimensions, thereby identifying group type and country as contingencies that moderate the relationship. Kowloon, Hong Kong, Hong Kong - PRC, mgangelo@cityu.edu.hk, Frederik Von Briel

494