Informs Annual Meeting 2017

WA51

INFORMS Houston – 2017

2 - CEO Scapegoating and Consumer Product Recalls Kevin Mayo, Indiana University, 1309 E 10th Street, Bloomington, IN, 47405, United States, mayok@indiana.edu, George Ball, Alex Mills Product recalls have significant impact to a firm’s financial and legal standing as well as on consumer health and safety. In this study, we explore the relationship between CEO turnovers and product recalls. Do recalls lead to CEOs being fired or quitting? Do new CEOs scapegoat and blame problems on their predecessors by recalling soon after taking over? Using a recurrent hazard model with 40 years of consumer product recalls and CEO turnover data, we examine the hazard of a CEO turnover after a recall and a recall after a CEO turnover. Results support a recall scapegoating effect. 3 - Leadership, Social Preferences and Productivity: Evidence from Truck Drivers We explore how the traits of leaders affects team productivity in the context of 53 customer-delivery trucks where the driver leads a team of two/three helpers. We collect social preferences from lab-in-the-field behavioral games and personality traits from a big-five survey. We show that “generalized trust” and “altruism” improves daily performance across several measures of productivity and quality. In contrast, “trusting helpers” is detrimental, with Big five traits having negligible impact. We also show that going from employee-driver to owner-driver, and thus directly employing its helpers, increases performance, particularly if the driver is “altruistic” and “open to experience”. 4 - Asymmetry and Sharing of Sales Information under Revenue Sharing Contract Ummuhan Akbay, Visiting Instructor, Sabanci University, Istanbul, Turkey, ummuhana@sabanciuniv.edu We study revenue sharing contract on a one manufacturer-one retailer supply chain under sales information asymmetry. The sales quantity to be disclosed to the manufacturer is determined by the retailer who has an opportunity to underreport. We investigate how the truth-telling behavior and ordering decisions of the retailer are affected by the contract parameters and the demand realization; how the manufacturer’s contract parameters are affected by the possible distortion in the shared sales information; and how both firms’ profits and the contract efficiency is affected in this scenario. 5 - Does Advance Warning Help Mitigate the Impact of Supply Chain Disruptions? An Experimental Investigation Sourish Sarkar, Assistant Professor, Penn State-Erie, The Behrend College, 5500 Copper Dr, Apt 301, Erie, PA, 16509, United States, sourishs@gmail.com, Sanjay Kumar We explored the impact of an advance warning of a disruption on supply chain performance. Using controlled laboratory settings, we considered disruptions at both upstream and downstream echelons. For each of these disruptions, scenarios with and without advance warning were compared. We also investigated the impact of communicating such warning to other echelons of the supply chain. We found a behavioral bias that results in worse performance in advance warning scenario for downstream disruptions. 6 - Gamified Newsvendor Experiment: the Impact of Leaderboards on Learning and Behavioral Biases Sina Zare, University of Texas-Arlington, 1020 W. Abram, Apt 289, #103, Arlington, TX, 76013, United States, sina.zare@uta.edu We present research examining how a gamified traditional Newsvendor experiment impacts individual’s decision making performance. We hypothesized that a gamified Newsvendor experiment with leaderboards positively influence newsvendor’s performance through improving learning process and mitigating the influence of behavioral biases. 361C Shared Mobility Systems Sponsored: TSL, Urban Transportation Sponsored Session Chair: Neda Masoud, University of Michigan, Ann, MI, 48109, United States, nmasoud@umich.edu 1 - Stochastic Control of Traffic Flow to Manage Ride-Sharing Platform Qi Luo, University of Michigan, 2465 Lancashire Drive, Apt 2A, Ann Arbor, MI, 48109, United States, luoqi@umich.edu, Romesh Saigal We investigate the Ride-Sharing Platform management using a continuous-time continuous-space approach. A monopolistic principal matches two sides of the market, i.e. supply (vacant vehicles) and demand (customers’ trip requests) dynamically. The supply model prevents the exponential growth of state space by Francisco Brahm, PhD Student, University of Cambridge, Trumpington 10, Cambridge, CB2 1AG, United Kingdom, fb435@cam.ac.uk WA50

adopting multi-population traffic flow theory. The demand side employs a queuing model. We further formulate the platform’s problem as an optimal stochastic control to maximize the revenue by redistributing initial conditions. This simple formulation enables to incorporate myopic agents’ local optimum decisions without treating combinatorial explosion. 2 - Ride-Matching in Large-Scale Transportation Networks Neda Masoud, University of Michigan, 2350 Hayward St., 2124 GG Brow, Ann Arbor, MI, 48109, United States, nmasoud@umich.edu, Yuexi Tu This study presents a divide-and-conquer algorithm for solving the ride-matching problem in a large-scale peer-to-peer ridesharing system. The complexity of the ride-matching problem grows with the number of riders, drivers, and the rider- driver pairs that are eligible to be matched. The divide and conquer algorithm divides the ride-matching problem into multiple smaller sub-problems of the same structure, and combines the solutions to these sub-problems to propose an approximation for the original problem. 3 - A Hub and Shuttle Transit System for Ann Arbor Pascal Van Hentenryck, University of Michigan, 1813 IOE Building, 1205 Beal Avenue, Ann Arbor, MI, 48108-2117, United States, pvanhent@umich.edu, Antoine Legrain, Connor Riley This talk describes a novel Hub and Shuttle Transit System for the Ann Arbor Campus of the University of Michigan (about 7.5 million passengers a year). The system addresses both congestion and the infamous first/last mile problem. The talk describes the design and implementation of the new transit system, simulation results, and early feedback from deployment. 4 - Matching Assignment Game for Mobility as a Service Transportation Assignment Saeid Rasulkhani, New York University, 328 91st Street, Apt 2R, Brooklyn, NY, 11209, United States, sr4262@nyu.edu, Joseph Y.J. Chow The emergence of more transportation options in smart cities requires more generalized transportation assignment models. We develop a transportation assignment model based on a route-based bipartite matching assignment game. For any mobility as a service operating design, we can determine not just the flow of travelers, but also the route service provision decisions of the operator(s) and cost allocations for all parties. Stability conditions under congestion are explored. 5 - Semi-door-to-door On-demand Transit System: Analytics and Practice We will discuss Ford’s various ongoing efforts in shared mobility, in particular, semi-door-to-door on demand ride sharing service. In contrast to traditional on- demand transit system that provides customers’ door-to-door rides, semi-door-to-door ride-sharing service dynamically clusters customers and thus have higher ride-sharing potential. We will discuss various analytics aspects of such ride-sharing system, efficiency evaluation, demand forecasting, dynamic pricing, etc. Hao Zhou, Ford Motor Company, 2101 Village Road, Dearborn, MI, 48124, United States, hzhou35@ford.com

WA51

361D Design and Planning of Ports Sponsored: Transportation Science & Logistics Sponsored Session

1 - A Framework for Building a Smart Port and Smart Port Index Anahita Molavi, University of Houston, 4722 Calhoun Rd, E224, Engineering Bldg 2, Houston, TX, 77204-4007, United States, anahita.molavi@gmail.com, Gino J. Lim, Bruce Race Globally, ports and harbors are facing stiffer competition for market share and delivering more effective and secure flow of goods. High performing ports are implementing smart technologies to better manage operations. We present a comprehensive literature review of the studies regarding Smart Port as a basis for a holistic definition for a Smart Port Index. The Smart Port Index in this paper is based on Key Performance Indicators and organized around four key elements: operations, environment, energy, and safety and security. Case studies and data analysis are used to test the KPIs to understand their characterization of the Smart Port performance of some busiest ports.

459