2016 INFORMS Annual Meeting Program

TD60

INFORMS Nashville – 2016

TD59 Cumberland 1- Omni Green Vehicle Routing General Session Chair: Mesut Yavuz, University of Alabama, Box 870226, Tuscaloosa, AL, 35487, United States, myavuz@cba.ua.edu 1 - Electric Vehicle Routing Problem With Time Windows And Multiple Charger Types Bulent Catay, Prof., Sabanci University, FENS, Tuzla, Istanbul, 34956, Turkey, catay@sabanciuniv.edu, Merve Keskin The electric vehicle charging stations may be equipped with chargers having different power supply, power voltage, and maximum current configurations. The type of the charger affects the recharge duration. In this study, we extend the Electric Vehicle Routing Problem with Time Windows by allowing partial recharges using three different charger types. The objective is to minimize total energy costs while operating minimum number of vehicles. We formulate this problem as a mixed integer linear program and propose a matheuristic approach to solve it effectively. The proposed approach uses an Adaptive Large Neighborhood Search algorithm to construct the routes and utilizes a solver to improve them. 2 - Cost Minimization And Fleet Sizing For Multifunction Electric Bus Fleets Amanda Farthing, Clemson University, Clemson, SC, United States, adfarth@g.clemson.edu, Nora Harris, Robert Riggs, Scott J. Mason We address the unique barriers facing university campus fleet managers considering a transition to electric bus fleets. Specifically, the logistical issues pertaining to multifunction vehicle fleets with fixed daytime routes and nighttime dial-a-ride service are addressed. A university vehicle fleet is analyzed in order to integrate real-world constraints, industry perspectives, and previous optimization research to develop a vehicle selection and fleet-sizing model that minimizes total cost. The model considers electric vehicle and infrastructure purchases, operation costs, and environmental benefits in this setting. 3 - The Maximum Profit Mixed-fleet Electric Vehicle Routing Problem This talk presents a maximum profit mixed fleet electric vehicle routing problem. A mixed fleet consists of traditional gasoline or diesel and electric vehicles. Electric vehicles enable the fleet operator to reduce their operating costs as well as carbon emissions. In addition, a set of customers are willing to pay a premium to receive service by electric vehicles to reduce their supply chain carbon footprint. We formulate the emerging problem as a mixed integer linear program, and present a route first cluster second and a greedy algorithm as well as their computational evaluation from our preliminary experiment. 4 - Greening Patrol Routing Via Extended-range Electric Vehicles Mesut Yavuz, University of Alabama, myavuz@cba.ua.edu, Burcu B Keskin, Cameron Harvey, Patrick Mitchell This study investigates patrol routing on state highways with hybrid electric vehicles, which operate in electric mode until battery depletion, and then switch to the more expensive gasoline mode. We present a mixed-integer non-linear programming formulation of the problem as well as analyze some special cases in which the problem reduces to one of minimum cost network flow. The objective is a weighted combination of “hot spot” coverage maximization and cost minimization. The model is tested on real data from Alabama State Troopers. TD60 Cumberland 2- Omni Understanding Shared Mobility and Autonomous Vehicles: Data, Models and Optimization Sponsored: TSL, Urban Transportation Sponsored Session 1 - Studying Trip Planning Behavior For Taxi Drivers Xian-Biao Hu, Metropia, Inc., Tucson, AZ, 85718, United States, xb.hu@metropia.com, Song Gao Taxi cabs account for a significant portion of traffic in megacities. However, research on taxi driver behaviors are limited and mostly formulated to maximize the probability of picking up or minimize search time to find next passenger. Such myopic approach departs from the driver’s actual objective to maximize profit over the entire operation period, and may fail to explain the search behavior around certain hotspots with high customer demand. This research aims to bridge this gap by studying the daily trip planning behavior for taxi drivers with the goal of maximizing profit over the entire operation period. Numeric analysis based on one-month taxi trajectory data will also be presented. Isil Koyuncu, University of Alabama, Tuscaloosa, AL, United States, ikoyuncu@crimson.ua.edu, Mesut Yavuz

aversion assumptions. We found that, theoretically, investment should decrease with firm revenue under specific settings and preference conditions, while experiments suggest the reverse. We also uncover dynamics in decisions where the setting is independent over time, counter to the theory. 2 - Incentivizing Suppliers Using Scorecard Sina Shokoohyar, University of Texas at Dallas, 800 West Campbell Road, Richardson, TX, Jindal School of Management, Richardson, TX, 75080, United States, Sina.shokoohyar@utdallas.edu, Elena Katok, Anyan Qi Suppliers’ scorecard is a tool for manufacturers to track suppliers’ performance. We investigate the effectiveness of two approaches for a manufacturer to incentivize suppliers to improve their performance based on the evaluation of their scorecard performance, the absolute and relative approaches. Under the absolute approach, the manufacturer provides incentive to the supplier if the supplier reaches a targeted score. Under the relative approach, the manufacturer incentivizes suppliers based on the suppliers’ scorecard ranking in the supplier base. Comparing the suppliers’ resultant scores under the two approaches, we characterize conditions on which approach is preferable. Chair: Sara Saberi, Worcester Polytechnic Institute (WPI), Washburn Rm 217, Foisie School of Business, Worcester, MA, 01609, United States, ssaberi@wpi.edu 1 - Server Scheduling Policies For The Queues With Abandonment Sina Ansari, Northwestern University, McCormick School of Engineering, 2145 Sheridan Road, Evanston, IL, 60208, United States, sinaansari2013@u.northwestern.edu, Seyed Iravani, Laurens G Debo We study the optimal server scheduling policy in a two-class service system with abandonment. With the objective of minimizing the total average abandonment cost per unit time, we characterize the optimal control policy at the server using Markov Decision Process. 2 - A Data_driven Approach To Model Fatigue At The Workplace Zahra Sedighi Maman, PhD Student , Research Scientist, Auburn University, Auburn, AL, 36849, United States, zzs0016@auburn.edu, Mohammad Ali Alamdar Yazdi, Fadel Megahed, Lora Cavuoto This paper presents feature selection and predictive modeling approaches for physical workload that can improve the fatigue prediction. The goal of this feature selection is to reduce the number of the used sensors and variables obtained from multiple sensors. The results show that the proposed approaches perform well both in prediction performance and more importantly in feature reduction. 4 - A Network Economic Game Theory Model Of A Service-oriented Internet With Price And Quality Competition In Both Content And Network Provision Sara Saberi, Assistant Professor, Worcester Polytechnic Institute, Foisie School of Business, 100 Institute Road, Worcester, MA, 01609, United States, ssaberi@wpi.edu, Anna B Nagurney, Tilman Wolf This paper develops both a basic and a general network economic game theory model of a quality-based service-oriented Internet to study the competition among the service providers. We derive the governing equilibrium conditions and provide the equivalent variational inequality (VI) formulations. In order to illustrate the modeling framework and the algorithm, we present computed solutions to numerical examples. The results show the generality of the proposed network economic model for a future Internet. TD58 Music Row 6- Omni Service Science Contributed Session

353