Informs Annual Meeting 2017

SD06

INFORMS Houston – 2017

SD04

SD05

320A Emerging Topics in Responsible Operations Sponsored: Manufacturing & Service Oper Mgmt, Sustainable Operations Sponsored Session Chair: Jayashankar M. Swaminathan, University of North Carolina, University of North Carolina, Chapel Hill, NC, 27599-3490, United States, msj@unc.edu 1 - Improving Supplier Compliance through Joint and Shared Audits with Collective Penalty Prashant Chintapalli, University of California, Los Angeles, Anderson School of Management, 11000 Weyburn Drive, Apt# 419, Los Angeles, CA, 90024, United States, prashant.chintapalli.1@anderson.ucla.edu, Felipe Caro, Kumar Rajaram, Chris Tang When suppliers fail to comply with safety regulations, buyers improve supplier compliance by auditing and imposing penalties. We analyze three audit mechanisms namely, independent, joint, and shared mechanisms. While in independent audit the buyers audit the supplier individually, in joint audit they audit jointly and share the audit cost. In shared audit they audit individually but share the information. We compare the three mechanisms in terms of audit levels, compliance levels, and the profits of the buyers, the supplier and the supply chain. We identify the cases when joint audit can improve supply chain profit so that both the buyers and the supplier can benefit from joint audits. 2 - Cover-up of Safety Hazards in Product Recalls Woonam Hwang, HEC Paris, Jouy-en-Josas, 78350, France, hwang@hec.fr, Soo-Haeng Cho, Victor DeMiguel Product safety regulators often have to rely on manufacturers’ voluntary disclosure of information when investigating product defects. However, manufacturers may not always truthfully report all safety hazard information. In this paper, we investigate how regulators can induce manufacturers to truthfully report any safety issues in a timely manner. Specifically, we examine the different regulatory policies employed in the automotive industry in the U.S. and the U.K., and compare their effects in inducing truthful reports from manufactures and the resulting social welfare. 3 - Optimal Seeding Policy under Rainfall Uncertainty Ying Zhang, University of North Carolina, Chapel Hill, NC, 27599, United States, ying_zhang@kenan-flagler.unc.edu, Jayashankar M. Swaminathan In this paper, we develop a model to determine the optimal seeding policy under rainfall uncertainty using a finite-horizon stochastic dynamic program. In our model, a farmer needs to decide the amount of seeds to plant in each period given the soil moisture. We show that the optimal planting policy is a time dependent threshold-type policy. Utilizing field weather data from Southern Africa, we investigate the impact of climate conditions on the relative yield advantage of the optimal planting schedule over commonly used heuristics in practice. For a real size planting problem, our computational study demonstrates significant relative yield advantage of the optimal planting schedule. 4 - Sustainable Sourcing Practices: Global Coverage, Characteristics and Drivers Joann De Zegher, jfdezegher@stanford.edu, Tannis Thorlakson, Eric Lambin Companies use a variety of Sustainable Sourcing Practices (SSPs) to address social and environmental challenges in their supply chains. Our current understanding of such SSPs is largely based on theoretical models, literature reviews and case studies. We use a random sample of 450 public companies to study the range of SSPs that companies pursue, addressing the following questions. What SSPs do companies adopt? How many tiers in the supply chain do SSPs cover and to what extent are these SSPs audited? What factors explain the type of SSP a company adopts and are these factors in line with theoretical predictions? This research helps direct the research agenda in sustainable supply chain management.

320B Advances in Healthcare Logistics Sponsored: Health Applications Sponsored Session Chair: Eva Lee, Georgia Tech, Atlanta, GA, 30332-0205, United States, evakylee@isye.gatech.edu 1 - Medical Imaging Log Files: Big Data Analysis & Findings Larissa P.G. Petroianu, University of Washington, 3801 Brooklyn MRIs are important for the detection and monitoring of specific conditions; however, the exam is expensive. Wasted scanner time caused by poor scheduling, failed/low value sequences, or idleness is to be avoided. This Philips Healthcare project analyzes data from scanner logs containing over 230,000 observations. We use machine learning to work with big data and focus on reducing “wasted” time to ascertain the root causes of unintended utilization patterns. Prior work utilized human observation which is subject to error, while this work develops analyses based on machine information. We will discuss the challenges of working with log files, present results and discuss the impact of these findings. 2 - Inventory Management and Scheduling for Blood Products: An Effect of Transfer among Regional Block Blood Centers Mari Ito, Tokyo University of Science, 2641 Yamazaki, Noda-shi, Chiba, 278-8510, Japan, mariito@rs.tus.ac.jp, Ryuta Takashima, Kazuhiro Kobayashi We investigate a management for regional block blood centers, which are established by Japanese Red Cross, by means of an inventory control and scheduling model for blood products. We propose a mathematical programming model which minimizes the number of discarded blood products and maintains a minimum blood supply. This model determines the number of blood products transferred from one regional block blood center to another one at one day and collected in one regional block blood center at one day. We show an effect of regional block blood centers on the number of collected, transferred and discarded blood products. 3 - Vaccine Prioritization for Optimal Population Health Protection Eva Lee, Georgia Tech, Industrial & Systems Engineering, Ctr for Operations Research in Medicine, Atlanta, GA, 30332-0205, United States, evakylee@isye.gatech.edu, Yifan Liu This study is joint with CDC. When the vaccine supply is limited during infectious disease outbreaks, prioritized vaccination is broadly considered as an effective strategy. In this study, we present a general-purpose framework to model the effect of mass vaccination during pandemics. This model integrates compartmental model for disease propagation and stochastic queueing for vaccination operations. Results for a smallpox outbreak will be discussed. 320C Emergency Services Analytics Sponsored: Health Applications Sponsored Session Chair: David L Kaufman, University of Michigan - Dearborn, Dearborn, MI, 48126, United States, davidlk@umich.edu 1 - Emergency Department Capacity Planning in Presence of Competition Elham Torabi, Assistant Professor, James Madison University, 421 Bluestone Dr., MSC 0202, Harrisonburg, VA, 22807, United States, torabiex@jmu.edu, Baback Vaziri Emergency departments (EDs) use historical patient arrival data to forecast future demand. The foretasted demand is then used for capacity planning purposes. However, this task becomes challenging if there is new competition in the region. We analyze the case of two EDs over a five-year period - one is a new ED entering the market and the other is an established ED facing new competition. We derive useful operational and managerial insights for EDs trying to cope with challenges in capacity planning in presence of competition. Avenue NE, Stevens Court - K302, Seattle, WA, 98105, United States, lpetroia@uw.edu, Rebecca J.Mieloszyk, Christina Mastrangelo SD06

101