Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MC31

4 - A Last Mile Delivery Paradigm using Microhubs with Crowdshipping Jane Lin, University of Illinois-Chicago, 842 W. Taylor Street (M/C 246), Chicago, IL, 60607, United States, Sudheer Ballare This study investigates the feasibility of a delivery paradigm in microhubs with crowdshipping. Performance was evaluated by comparing with the traditional Hub-and-Spoke paradigm in terms of vehicle miles traveled, numbers of trucks and crowdshippers dispatched, fuel consumption, and total daily operating cost. The study also investigates the effect of key operational parameters such as network size, customer demand density, crowdshipper payment and penalty rate on the performance of M+C. It is found that M+C generally outperforms H+C with the economy of scale. It is also found that higher penalty rate increases the attractiveness of the proposed M+C delivery paradigm. 5 - Uncertainty at Scale: The Technician Routing Problem with Hard Time-Windows, Time-Dependent Travel, and Stochastic Service Times Ishai Menache, Microsoft, Cambridge, MA, United States We revisit the VRP for field services. Our goal is to account for inherent uncertainty in work-duration or travel-time. To that end, we introduce risk as an additional measure. Risk quantifies the probability of “bad outcomes in the schedule; for example, missing a work-order time-window. We design efficient algorithms for estimating the risk. Based on that, we design and evaluate a scalable optimization framework, which allows operators to choose the sweet point of risk vs. expected performance, akin to portfolio management. n MC30 North Bldg 221C Recent Advances in Facility Location and Supply Chain Design Sponsored: TSL/Facility Logistics Sponsored Session Chair: Lian Qi, Rutgers University, Piscataway, NJ, 08854, United States 1 - Two-stage Stochastic Programming of Facility Location Problem with Endogenous Uncertainty Mengnan Chen, University of Central Florida, Orlando, FL, United States, Qipeng Zheng Our study aims to improve hospital efficiency by allocating the hospital resource (physician and clinic), as well as matching the patient preference to maximize the patients’ satisfaction. We use the two-stage stochastic programming to model the physician/clinic facility location and patient assignment problem, where the patient preference is considered as the endogenous uncertainty. To solve this model, we design the hybrid heuristic algorithm, that the Tabu Search (TS) and Large Neighborhood Search (LNS) are used to solve the facility location problem, and Sample Average Approximation (SAA) is used to handle the exponential Dincer Konur, Missouri University of Science and Technology, 206 Emse, 600 W. 14th Street, Rolla, MO, 65401, United States This paper studies store location decisions of a retail chain. The located stores will be competing for the demand in the end consumer market. The retail chain can prefer centralized or decentralized approach for determining the store locations. In the case of decentralized approach, the stores will engage in a competitive facility location game. We analyze the location decisions under each approach. Numerical studies are conducted to demonstrate the effects of different approaches and parameters on the outcome of the store location decisions. 3 - Product Geographical Distribution under Recall Risk Ying Rong, Shanghai Jiao Tong University, No. 1 Lane 9, Yunwushan Road, Shanghai, 200051, China, Long He, Zuo-Jun Max Shen When product recalls happen, companies not only have to incur additional logistics costs but also su?er from damaged reputation. In this paper, we discuss how to alleviate the consequences of product recalls in the perspective of (outbound) product geographical distribution strategy in joint with (inbound) sourcing decision. increasing scenario size of patient assignment problem. 2 - On Retail Chain Store Location Decisions

4 - Locating Distribution Centers in a Collaborative Logistics Network Chase Rainwater, University of Arkansas, 4207 Bell Engineering Center, Fayetteville, AR, 72701, United States Researchers worldwide have concluded that meeting a shift towards smaller, fast- moving, unit loads calls for a new paradigm for future logistics systems. A potential cornerstone of a proposed new paradigm is collaboration and, in particular, horizontal collaboration. The vision embraces advancing technologies and is consistent with the issues challenges addressed by Industry 4.0. This talk focuses specifically on the design of a collaborative supply chain network as it relates to the location of shared distribution centers and warehouse. The impact of horizonal collaboration on the design choices made to construct such network are specifically discussed. n MC31 North Bldg 222A New Mobility Services Sponsored: TSL/Intelligent Transportation Systems (ITS) Sponsored Session Chair: Roksana Asadi, Stony Brook University, 100 Nicolls Rd, Stony Brook, NY, 11794, United States 1 - Ridesharing and Self-parking with Autonomous Vehicles: A Novel Multimodal Ridesharing and Parking User Equilibrium Bo Zou, Assistant Professor, University of Illinois at Chicago, 2073 Engineering Research Facility, 842 W. Taylor Street, Chicago, IL, 60607, United States, Mohamadhossein Noruzoliaee This study explores the impacts of Autonomous Vehicles (AVs) on transportation system equilibrium focusing on ridesharing with shared AVs and self-parking with private AVs. A Stackelberg game is developed where a profit-maximizing Transportation Network Company (TNC) chooses shared AV fleet size, sets fare rates, and decides on fleet allocation/relocation. The associated system equilibrium is formulated as a novel Multimodal Ridesharing and Parking User Equilibrium (MRPUE) problem, which determines market share of AV in a mixed autonomous/human driving environment. 2 - User and System Optimal Matching for Dynamic Ride Sharing Systems Pramesh Kumar, University of Minnesota, Minneapolis, MN, 55414, United States, Alireza Khani This research develops an optimization model for matching riders with drivers in a ride sharing problem with dynamic travel times. The objective is to reduce travel delay and schedule deviation, and multiple rider matching is allowed. In addition to the system optimal model, a user optimal model is presented in which users cannot improve their match unilaterally. 3 - A Modeling Framework for the Integrated Vehicle-drone Routing Problem Aline Karak, PhD Student, Southern Methodist University, Amesbury, dallas, TX, 75206, United States, Khaled Abdelghany We present a mathematical formulation for the integrated vehicle-drone routing problem in the context of pick-up and delivery services. An efficient heuristic- based solution methodology that captures the cost trade-off and operational characteristics of the two modes is presented. The methodology introduces a multi modal version of the classical Clarke and Wright algorithm. The results of a set of experiments are presented, which compare the solution performance against drone-driven and vehicle-driven solutions. 4 - Joint Optimization of Electric Power Distribution and Electric Vehicle Charging Infrastructure Design under Network Traffic Equilibrium Roksana Asadi, Graduate Student, Stony brook university, 100 Nicolls Rd, Stony Brook, NY, 11794, United States, Leila Hajibabai This study investigates an integrated plan for the power distribution network (PDN) design and electric vehicle (EV) charging station deployment in urban transportation networks with the underlying traffic flows. The problem is characterized by a bi-level model structure with PDN design and facility location decisions in the upper level and traffic equilibrium in the lower level. The objective is to minimize the total cost due to PDN operation, charging facility deployment, and transportation. The problem is converted into a single-level model based on the Karush-Kuhn-Tucker conditions and solved using a column and constraint generation algorithm.

195