Informs Annual Meeting 2017

MC48

INFORMS Houston – 2017

MC46

2 - Optimal Pricing and Workforce Composition for Service Delivery using a Hybrid Workforce Monica Johar, University of North Carolina, 9201 University City Blvd, Friday 352 C, Charlotte, NC, 28223, United States, msjohar@uncc.edu Organizations are increasingly able to use an on-demand workforce to complement in-house workers to perform sophisticated workflows in a Hybrid Service environment. Such sophisticated workflows include simple as well as more complex integration tasks. We present an optimization model that integrates pricing, workforce composition and task allocation decisions for such a hybrid service environment. 3 - How Much to Open, How Fast to Fix and Develop? Impacts of Openness on Software Development and Maintenance Rakesh Mallipeddi, TX, United States, mallipeddi@tamu.edu, Emre Muzaffer Demirezen, Subodha Kumar, Ram Gopal Proprietary software vendors have recently begun to emulate the open source software community in opening up part of their software. We analyze the impact of software openness in the context of resource allocation between bug fixing and new version development, an important operations issue that many software vendors face in light of prevalence of software defects. We first empirically analyze the impact of openness on time it takes to fix a defect. Using the findings from the empirical study, we formulate optimal control models to examine the effects of making the software code open (fully or partially) on the overall quality of the software and the development efforts of its next version. 4 - Dynamic Optimization of Software Maintenance Requests with Duplications Yonghua Ji, University of Alberta, Business School, Edmonton, AB, T6G 2R6, Canada, yji@ualberta.ca, Said Samiedaluie We study the problem of software request maintenance by using a dynamic optimization approach. We consider both duplication of and prioritization of requests. Duplicated requests can arrive independently and are grouped before being worked on. Also requests are classified into different priority classes. Our work explores various characteristics of an optimal policy. 361A Opportunities and New Directions for Behavioral OM Sponsored: Behavioral Operations Management Sponsored Session Chair: Robert Batt, Wisconsin School of Business, UW-Madison, Madison, WI, 53706, United States, rbatt@bus.wisc.edu Co-Chair: Jordan Tong, University of Wisconsin-Madison, Madison, WI, United States, jordan.tong@wisc.edu 1 - Humans are not Machines: Queueing with Human Servers Amy R.Ward, University of Southern California, Marshall School of Business, Bridge Hall BR.I.401H, Los Angeles, CA, 90089-0809, United States, amyward@marshall.usc.edu There is a long history of using queueing models for system design purposes. This is because there is a well-developed framework that provides insightful analytic expressions for many performance measures of interest, such as expected customer wait time. The issue is that traditional queueing models assume processing rates are fixed. However, in many systems that arise both in manufacturing and service applications, people rather than machines are doing the processing, and people may change their processing rates. In this talk, we discuss how to incorporate behavioral elements in queueing models. We highlight some surprising results, and propose some open questions. 2 - Beyond the Lab: Next Frontiers in Behavioural Research Serguei Netessine, University of Pennsylvania, Philadelphia, PA, United States, netessin@wharton.upenn.edu While vast majority of behavioural research in our field uses lab experiments, there are important complementary approaches that we need to utilise going forward to make stronger practical impact. I will discuss these approaches using recent papers as examples. MC48

360E Scheduling and Logistics Invited: Project Management and Scheduling Invited Session Chair: Xiangtong Qi, Hong Kong University of Science & Technology (HKUST), Kowloon, Hong Kong, ieemqi@ust.hk 1 - Approximation Algorithms for Capacitated Vehicle Routing Problems with a Fixed Number of Depots Liang Xu, Associate Professor, Southwestern University of Finance & Economics, 555 Liutai Avenue, Tongbo Tower A.202, Chengdu, China, arecxuliang1112@gmail.com, Zhou Xu We study problems of routing capacitated vehicles to serve customers with an objective to minimize the total travel distance. In this talk, we present several novel constant ratio approximation algorithms, which run in polynomial time Yifu Li, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, 999077, Hong Kong, ylick@connect.ust.hk, Tinglong Dai, Xiangtong Qi Sequence decision of an experiential service is intensively studied by researchers from various disciplines. While empirical research shows that the ideal schedule is usually with the peak activity early in the interior, behavior study suggests that a sequence with the peak activity by the side is optimal. To address this inconsistency, we introduce a modification of the acclimation and memory decay model. We examine the condition of the interior peak optimal schedule, and we find the optimal position of the peak activity may change non-monotonically as its memory decay rate increases. Furthermore, we explore the shape property and propose an optimal algorithm for the service scheduling problem. 3 - Due Date Assignment and Scheduling on a Single Machine with Limited Information of Job Processing Times Guohua Wan, Shanghai Jiao Tong University, 535 Fahua Zhen Road, Antai College of Economics and Management, Shanghai, 200052, China, ghwan@sjtu.edu.cn, Qing Yue, Shenghai Zhao We study a single machine scheduling problem with due date assignment and stochastic job processing times, where only the mean and support of job processing times are known but not the distribution. The objective is to minimize the total costs of earliness, tardiness and due date assignment. We develop an approximation for the problem and propose both branch and bound and heuristic algorithms to find the optimal solution. We also conduct computational experiments to evaluate the efficiency of approximation and the performance of proposed heuristic algorithms. 4 - Stabilizing Unbalanced Cooperative Scheduling Games Xiangtong Qi, Hong Kong Univ. of Science & Technology (HKUST), Dept of IELM, HKUST, Kowloon, Hong Kong, ieemqi@ust.hk We consider cooperative parallel machine scheduling games. In many cases such games are unbalanced, i.e., having an empty core. Hence the grand coalition of the game is not stable. We study how an outsider can take some action to stabilize the game so that the grand coalition can be sustained. We will discuss different schemes and show how each of them works. 360F Managing Service Delivery and Software Processes in an ebusiness Environment Sponsored: EBusiness Sponsored Session Chair: Yonghua Ji, University of Alberta, Edmonton, AB, T6G 2R6, Canada, yji@ualberta.ca 1 - Does Online Real Estate Information Affect Property Price? Evidence from Zillow Zhengrui Jiang, Iowa State University, Iowa CIty, IA, United States, zjiang@iastate.edu, Yuheng Hu, Arun Rai, Hua Sun Using data collected from Zillow.com, this study examines how online property information affects offline property price. Our results reveal that property information has a larger influence on the sale price of uncommon properties, i.e., those valued significantly lower or higher than neighborhood averages, than on the price of common properties, i.e., those valued close to neighborhood averages. Furthermore, we find that Zestimate, property value assessed by Zillow, influences the price of all properties. Finally, sellers tend to provide more property information when a property is more expensive, less comparable to surrounding ones, or priced higher than its estimated property value. MC47 when vehicles are located at a constant number of depots. 2 - Scheduling Service Package with Acclimation and Non-homogeneous Memory Decay

219