Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MD30

2 - Inferring Multimodal Day-to-day Travel Dynamics using Smartphone App Data Shanjiang Zhu, George Mason University, Nguyen Engineering Building, Suite 1300, 4400 University Drive Ms 6c1, Fairfax, VA, 22030, United States, Zhuo Yang, Guanqi Liu, Lei Zhang This study investigates the day-to-day travel dynamics on a multi-modal network. Longitudinal travel trajectories data were collected using a smartphone app among commuters during Washington Metro Safetrack project. Algorithms were developed to infer the travel choices before, during, and after the network disruptions, which are crucial for assessing the impact of the unprecedented maintenance work at the Washington Metro. Findings from this study could help to inform agencies who are struggling with the aging infrastructure across the country and help them to develop better strategies. 3 - Early Warning Signal for Congested Large-scale Traffic Networks Xiaozheng He, Rensselaer Polytechnic Institute, 110 8th St., JEC 4034, Troy, NY, 12180, United States, Chunhen Jiang, Jihui Nie, Jianxi Gao This study proposes an early warning signal to indicate whether the traffic condition of a congested network is getting close to its critical degradation threshold, beyond which congestion is difficult to mitigate. The early warning signal is developed based on the critical slowing down theory for perturbed dynamical systems. Using the collected field data, we validate the proposed signal and show that the system recovery process becomes increasingly slow when the traffic condition approaches the critical point. 4 - Estimating Probability Density of Origin-destination Matrices on Congested Networks Yudi Yang, 1420 Lake Boulevard, Apartment 17, Davis, CA, 95616, United States, Yueyue Fan, Johannes Royset To understand the stochastic nature of travel demand is gaining more attention in transportation studies as reliability and resilience become important performance measures for transportation project evaluation. In this study, we aim to infer the probability density function (pdf) of Origin-Destination (O-D) demand variables by integrating (potentially) multiple data sources. Unlike most traditional statistical approaches that are only applicable to non-congested networks, the proposed method is designed to be capable of incorporating traffic network flow rules/models in a congested network and determining route choice proportion and O-D matrix simultaneously. 5 - Transportation Big Data: Promises, Issues, and Implications Rong Fan, University of Washington, Seattle, WA, United States, Xuegang (Jeff) Ban Big data and related data analytics methods have received much attention recently in transportation for various planning and operational applications. This talk summarizes the promises of big data and illustrates the issues of some commonly used big data sources in transportation. We then briefly discuss the implications of such issues and suggest a pathway that may help address those issues. n MD30 North Bldg 221C Facility Layout Sponsored: TSL/Facility Logistics Sponsored Session Chair: Churlzu Lim, UNC-Charlotte, Systems Engineering & Engineering Management, 9201 University City Boulevard, Charlotte, NC, 28223, United States 1 - An Evaluation Method for Layout Design of a Spacecraft Using Canonical Correlation, Data Envelopment Analysis, and Choquet Integral Claudia Ramirez, Doctoral Student, University of North Carolina- Charlotte, 9201 University City Blvd, Charlotte, NC, 28223, United States, Simon Hsiang, Churlzu Lim The ideal arrangement of task spaces in a spacecraft reflects a minimal overall volume while maximizing the crewmembers’ health, performance, and safety for the duration of a mission to Mars. Their “happiness is evaluated by considering design factors that rank and score different layout designs based on the factors’ priority order designated by the user. The evaluation is achieved through the relationship between the spacecraft’s physical characteristics and their psychophysical attributes by using Canonical Correlation Analysis, Data Envelopment Analysis, and Choquet Integral. 2 - Dominance Results for the M Finite-size Facility Placement Problem Rakesh Nagi, University of Illinois at Urbana-Champaign, Department of Industrial & Enterprise Systems, 117 Transportation Building, MC-238, Urbana, IL, 61801, United States, Ketan Date We study the M Finite-size Facility Placement Problem: Optimally placing M new finite size rectangular facilities (NFs) in a layout with N existing rectangular

facilities (EFs). Pairs of facilities interact through Input/output points located on the facility boundary. We propose lower bounds and dominance in an implicit enumeration scheme that significantly reduces the number of candidate solutions, making this a viable solution approach in practice. 3 - Effect of Dynamic Demand Uncertainty on Facility Layout Design: Revisiting Integrated Facility Layouts Melih Celik, Middle East Technical University, ODTU Endustri Muh. Bolumu, Universiteler Mh. Dumlupinar Bl. No: 1, Ankara, 06800, Turkey, Beg n Efeoglu, Haldun S ral We consider the integrated facility layout problem (IFLP), where the aim is to determine the material or product flows among (duplicates or copies of) work centers and machines, in conjunction with their locations within the facility. In many manufacturing environments, the amount of demand may not be known in advance. For this end, we consider the stochastic IFLP in a dynamic environment where the jobs have different demand distributions in each period. We propose a dynamic programming approach based on the enumeration of possible machine assignments with given machine types, machine copies, and locations in each period. We show the effectiveness of our approach using instances from the literature. 4 - Overlap Cuboid Packing Optimization for Spacecraft Layout Design and Computational Enhancements Churlzu Lim, University of North Carolina-Charlotte, Systems Engineering & Engineering Management, 9201 University City Boulevard, Charlotte, NC, 28223, United States, Richard Alaimo, Claudia Ramirez, Simon M. Hsiang Consider a variant of the packing problem, called overlap cuboid packing problem, which allows for cuboid items to share space in order to further minimize the dimensions of the container. The objective of this study is to investigate a mixed-integer linear program that effectively provides a packing solution in the context of designing the layout for a spacecraft module. This talk will briefly review the bi-objective function that enforces specific adjacency requirements between gradient cuboids, and discuss about computational enhancements that display promising results in a numerical study. n MD31 North Bldg 222A Logistics, Supply Chain, and Multi-modal Transportation Sponsored: TSL/Intelligent Transportation Systems (ITS) Sponsored Session Chair: Nasibeh Zanjirani Farahani, University of Missouri, E3437 Thomas and Nell Lafferre Hall, Columbia, MO, 65211, United States 1 - A Sim-heuristic Approach for Near Real-time Optimization of Single-track Railway Scheduling Qi Tian, Dalian University of Technology, 2 Linggong Road, Ganjingzi District, Dalian, 116024, China, Rutgers, The State University of New Jersey, 96 Frelinghuysen Road, Piscataway, NJ, 08854, United States, Weihong Guo, Wenyuan Wang, Zijian Guo The wide applications of real-time sensing devices and big data methods are providing unprecedented opportunities for developing new solutions for single- track railway scheduling. Considering near real-time information such as unexpected delays caused by inclement weather or train breakdowns, this research develops a sim-heuristic solution algorithm to minimize the total weighted delay time. The proposed method provides near real-time optimization for single-track railway scheduling. The effectiveness of the proposed algorithm is demonstrated in both simulation and a real-world case study. 2 - A Scalable Non-myopic Atomic Game for Smart Parking Mechanism Hamid R. Sayarshad, Cornell University, Ithaca, NY, 14850, United States, Shahram Sattar, H. Oliver Gao We propose a new smart parking mechanism which seeks to reduce cruising for parking at multiple parking facilities in a crowded area. A non-myopic atomic game is formulated to search the equilibrium solution through allocating travelers to candidate parking facilities which considers travel time difference for the vehicles, walking time, dynamic pricing, cruising time and parking facilities’ occupancy. This study integrates a social efficiency price which accounts customer waiting time for parking searching. We involve the network travel time variation into our dynamic policy and make the game model reflect reality better by the competition of vehicles in both road resources and parking spaces.

224