2016 INFORMS Annual Meeting Program

WD76

INFORMS Nashville – 2016

WD74 Legends B- Omni Ops Mgt/Marketing IV Contributed Session

developed that minimizes the system’s operational costs, including transportation, stocking and purchasing. The numerical experiments show the benefits of purchasing and ordering options in the systems with high penalty costs and long repair times. The benefits are more significant in systems with tight inventories. Our analysis identifies the best inventory levels that minimize both the operational costs and the initial investments at the stock location. 2 - Procurement Under Price Uncertainty – An Analysis Of Operational Hedging Strategies Ashutosh Sarkar, Associate Professor, Indian Institute of Management Kozhikode, IIM Campus, Kunnamangalam, Kozhikode, Kerala, Kozhikode, 673570, India, asarkar@iimk.ac.in, Goutam Sutar, Arun Kumar Misra Consider a manufacturer procuring one of its raw materials from overseas sources. The manufacturer, while facing the risks of price uncertainty due to exchange rate fluctuations, needs to decide the timing, the source and the quantity of purchase. We modeled the manufacturer’s decision problem as a multi-period inventory problem and showed that the (s, S) policy is optimal when the manufacturer buys only once during the planning horizon. We also evaluated various operational hedging strategies like, switching, postponement and switching with financial options. 3 - Additive Manufacturing In A Bio-medical Supply Chain: A Continuous Approximation Approach Adindu Emelogu, PhD Student, Mississippi State University, Dept of Industrial & Systems Engineering, 479-2 Hardy Road, Mississippi State, MS, 39762, United States, aae39@msstate.edu, Sudipta Chowdhury, Mohammad Marufuzzaman, Linkan Bian The fabrication of biomedical devices close to hospitals via Additive Manufacturing (AM) technology has been gaining popularity due to the many potential benefits it provides such as patient-customized parts, fast response, and reduced delivery cost. However, not much attention has been given to AM deployment methods which impact the supply chain and the amount reaped of these benefits. We propose a continuous approximation (CA) model that quantifies the supply chain network costs of AM-produced biomedical implants. We present an algorithm that optimizes the location of the AM centers and raw material inventory to satisfy the customers. We use hospitals in the southeastern USA as our case study. 4 - Considering Dynamic Demand On The Supply Chain Optimization Via Bargaining Models On A Common Replenishment Epochs Environment José Velásquez, Universidad de los Andes, Calle 44D # 45-86 Int.1 Apto-503, Bogotá, 111321, Colombia, jl.velasquez1322@uniandes.edu.co, Jose Fidel Torres In this work we present five different linear programing models to coordinate the supply chain inventories on a single-supplier, single-buyer environment for a variety of products. We considered the common replenishment epochs (CRE) approach on different scenarios where the demand faced by the buyer is dynamic. Depending on the case, one of the actors offers a compensation to the other, in order to accept a strategy for a fixed replenishment period. Finally, we conducted a numerical study to evaluate the benefits of the proposed coordination strategies. 5 - Heuristic Algorithm For Multi-criteria Procurement In Energy Systems Each distributed energy system has several agents, including customer, storage units, and energy source centers that produce energy. Each of these agents are connected to other agents either directly or through other agents and these connections form a network called energy distribution network. To find the optimal route, several heuristic algorithms have been developed. All of the developed algorithms are for a single objective function. However, in reality, decision-makers have to consider several criteria simultaneously. Therefore, the problem is a multi-criteria problem. In this study, we propose multi-criteria heuristic algorithm based on Dijkstra’s algorithm. Mohammad Komaki, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH, 44106, United States, gxk152@case.edu

Chair: Zhenyu Gao, Tsinghua University, Room 430B, Zijing Department #14, Beijing, 100084, China, pjgzy1@163.com 1 - Omni-channel Retail In The Presence Of Operational Frictions Xiaomeng Guo, Assistant Professor, The Hong Kong Polytechnic University, M628, Li Ka Shing Tower, Hung Hom, Kowloon, Hong Kong, xiaomeng.guo@wustl.edu, Panos Kouvelis, Danko Turcic Some firms are implementing omni-channel strategies by offering consistent products and prices across their multiple channels, and some other firms essentially prevent seamless transition between different channels. Our paper provides a game-theoretical model to compare the traditional multi-channel and omni-channel strategies by focusing on product and price consistency. 2 - The Study About Crowdfunding Flight Zihao Zhang, Master, University of Science and Technology of China, 96 Jinzhai Road, Management Research Building 613,Room, Hefei, 230026, China, zhangzih@mail.ustc.edu.cn, Liuyi Ling Airline may have a loss when passenger load factor is very low due to high costs.So we try to solve the problem with crowdfunding.The study investigates the optimal pricing decision and other decisions about crowdfunding for OTA and airline with Stackelberg game.Results indicate that OTA can determines the optimal price and the lowest tickets sales to maximize its profit,in addition,we can get that lead time will be decided by rent and price. Airline will find the optimal rent according to decisions of OTA. The study also establish a contract to coordinate the supply chain consisting of OTA and airline. 3 - Equilibrium Power Structures In The Presence Of Stochastic Learning Guowei Liu, Tianjin University, 92 Weijin Road, Nankai District, Tianjin, 300072, China, gwliu@tju.edu.cn, Yunchuan Liu, Jianxiong Zhang This paper studies equilibrium power structures in a two-period model, where a manufacturer produces a product with stochastic batch learning and sells it to end consumers through a retailer facing a linear demand. The manufacturer and the retailer can implement a dynamic or commitment contract over both periods. We show that when the learning efficiency is sufficiently high, Vertical Nash and Retailer Stackelberg are the equilibrium power structures under the dynamic and commitment contracts, respectively. Meanwhile, the equilibrium power structures are beneficial to consumers. We also extend our main model to the continuous learning and non-linear demand cases. 4 - Storage Assignment In Mobile Fulfillment System Zhenyu Gao, Tsinghua University, 14#430, Zijing Department, Beijing, 100084, China, gaozy14@mails.tsinghua.edu.cn, Chen Wang In the thriving E-commerce market with expanding scale of customers and items, fulfilling large volume of small orders accounts for most of the operational cost. The mobile fulfillment system provided by Amazon dispatches large scale of robots to assemble multiple inventory pods with items needed simultaneously, which makes storage assignment more flexible by storing multiple items in one pod and saves cost significantly compared to traditional warehouses. We extract correlation information among items with factorization machines, which is then integrated into a clustering model for the assignment solution. Algorithms are developed both for the factorization machine and assignment model.

WD76 Legends D- Omni Supply Chain Optimization Contributed Session

Chair: Mohammad Komaki, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH, 44106, United States, gxk152@case.edu 1 - Dynamic Decision Making In A Two Echelon Repairable Inventory System With Purchase And Order Options Rana Afzali-Baghdadabadi, Operation Researcher, General Motors, 2462 John R Rd, # 107, Troy, MI, 48083, United States, rana.afzali@gm.com, Wooseung Jang In this study, we consider a two echelon repairable parts inventory system, where emergency purchasing and ordering from a central warehouse are allowed to deliver high service levels to customers. A dynamic decision making model is

479