Informs Annual Meeting 2017

MB16

INFORMS Houston – 2017

4 - A Quasi Experimental Study of Environmental Benefits of Craigslist Suvrat Dhanorkar, Pennsylvania State University, 466 Business Building, University Park, PA, 16802, United States, ssd14@psu.edu Matching platforms such as Craigslist can facilitate the creation of C2C closed loop supply chains (CLSCs) for used goods, which can enhance product reuse and limit reliance on recycling and disposal alternatives.We use a quasi-experimental setup to examine how Craigslist’s entry into various U.S. geographic markets impacts environmental outcomes. 332F Inventory Management Contributed Session Chair: Zhongjie Ma, Purdue University, West Lafayette, IN, United States, zhongjiema@purdue.edu 1 - A Reinforcement Learning Approach for Perishable Product Supply Chain with Markov-modulated Demand Ahmet Kara, Erciyes University, Erciyes University, Department of Industrial Engineering, kayseri, 38039, Turkey, ahmet.kara@erciyes.edu.tr, brahim Do an Inventory management of perishable products with limited shelf life is different than the classical inventory management. The inventory control of perishable products requires different ordering policies that consider both the quantity and the age of the stocks. This study analyzes a two-echelon supply chain model for perishable inventory management with Markov-Modulated Demand. We examine different ordering policies to emphasize the importance of the age information. The problem has been formulated using a reinforcement learning model. 2 - Newsvendor Problems with Demand Forecast Updating and Supply Constraints Meimei Zheng, Shanghai Jiao Tong University, Shanghai, 200240, China, meimeizheng2009@gmail.com, Kan Wu, Yan Shu This study investigates an extension of the newsvendor model with demand forecast updating under supply constraints. A retailer can postpone order placement to obtain a better demand forecast. However, the manufacturer would charge the retailer a higher cost for a shorter lead time and set restrictions on the ordering times and quantities. Two supply modes are investigated: supply mode A, which has a limited ordering time scale, and supply mode B, which has a decreasing maximum ordering quantity. For supply mode A, it can be proven under justifiable assumption that a retailer should order either as early or as late as possible. For supply mode B, an algorithm is proposed to simplify the ordering policy. 3 - Optimal Order and Shipment Policies on Classic Periodic-review Two-echelon Distribution Models Youyi Feng, Professor, Business School, Sichuan University, Wangjiang Campus, Sichuan University, Wuhou District, Chengdu, Sichuan, China, Chengdu, China, yyfeng@live.com, Lei Xie We characterize optimal dynamic inventory policies for the classic single-item periodic-review two-echelon distribution inventory models over a finite planning horizon. Consisting of a warehouse and a number of retailers, the system periodically procures inventory from an outside supplier to replenish the warehouse and delivers on-hand inventory from the warehouse to the downstream retailers. We innovate the computing procedures of linear complexity to obtain optimal order/shipment quantities respectively. Infinite- horizon models with discounted cost and long-term average cost criteria are also explored. 4 - Multi Season Production Planning under Sales Quota Tianxiao Chen, The Chinese University of Hong Kong, ERB615, CUHK, Shatin, Shatin, Hong Kong, chentx@se.cuhk.edu.hk, Xiting Gong, Qing Li, He Xu Motivated by the export quota in international trading, we study a periodic- review multi-season production planning problem where the total sales quantity in each season is restricted by a sales quota. By formulating the problem as a dynamic program, we characterize the structure of the optimal production policy and exam the impact of sales quota. MB16

5 - Dynamic Coordination in a Supply Chain with Production Capacity Uncertainty Zhongjie Ma, Purdue University, 403 W. State Street, West Lafayette, IN, 47906, United States, zhongjiema@purdue.edu, Qi Feng, J. George Shanthikumar We study the effect of upstream supply uncertainty on the inventory decisions in a two-stage supply chain under both centralized and decentralized controls. Applying the concept of directional concave ordering and extending the notion of stochastic linearity, we develop a concave transformation of the problem for a general class of stochastic supply functions, which allows us to analyze the dynamic problem and characterize the optimal policy. Furthermore, we discuss the mechanism to dynamically coordinate the decentralized systems when the upstream and downstream make the ordering decisions based on self-interest. 340A Queueing Models in Applied Probability Sponsored: Applied Probability Sponsored Session Chair: Baris Ata, University of Chicago, Chicago, IL, 60637, United States, Baris.Ata@chicagobooth.edu Co-Chair: Xiaoshan Peng, University of Chicago, Booth School of Business, Chicago, IL, 60637, United States, x-peng@chicagobooth.edu 1 - Managing Service Platforms with Self-scheduling Agents and Impatient Customers Rouba Ibrahim, University College London, MS& I.department, UCL, Gower Street, London, WC1E 6BT, United Kingdom, rouba.ibrahim@ucl.ac.uk We study the operational management of service platforms with self-scheduling agents and impatient customers. Since the customer impatience distribution plays an important role, we propose controlling it using delay announcements, and characterise the interaction between three controls at the manager’s disposal: the staffing level, the compensation offered to agents, and the announcements made to customers. 2 - A Constrained Optimization Problem for a Two-Class Queueing Model Mark E. Lewis, Cornell University, School of Ops Research & Information Engin., 221 Rhodes Hall, Ithaca, NY, 14853, United States, mark.lewis@cornell.edu, Cory Girard, Linda Vogel Green, Jingui Xie We discuss dynamic server control in a two-class service system under a constraint on the number of high-priority customers. A class of randomized threshold policies is defined, and is proven to contain an optimal policy in the case without abandonments. The proof of optimality is then used to construct heuristic policies for the case of low-priority abandonments, which we test numerically. The experiments we run suggest that, even when abandonments are introduced, these classes of policies outperform priority policies, and, in some cases, are near-optimal. 3 - Delay Predictability Tradeoffs in Reaching a Secret Goal Kuang Xu, Stanford Graduate School of Business, Stanford, CA, United States, kuangxu@stanford.edu, John N.Tsitsiklis We formulate a model of dynamic decision-making to study an agent’s predictability as she attempts to reach a final goal through a sequence of intermediate actions, while watched by an adversary who tries to predict the goal before it is reached. We are motivated by the increasing ubiquity of large-scale data collection infrastructures capable of predicting an agent’s intentions and future actions, in juxtaposition with an agent’s desire for privacy. We show the predictability of the agent’s goal can be made inversely proportional to the time she spends reaching it, and that this is the best possible. This characterization does not depend on the structure of the agent’s state space beyond the diameter. 4 - Managing the Callback Option under Arrival Rate Uncertainty Xiaoshan Peng, Indiana University, Kelley School of Business, Bloomington, IN, 47405, United States, xp1@iu.edu, Baris Ata We study how to manage the callback option effectively to mitigate congestion due to temporary surges in the arrivals to a call center. When a customer arrives, the call center manager decides whether to keep him in the online queue or to offer the callback option. Some customers may reject the callback offer. We show that a modified lookahead policy that looks into the future arrivals and service completion times for the next p/h time units and uses the current number of customers in the system who previously rejected a callback offer is pathwise optimal among the complete foresight policies. MB17

170