2016 INFORMS Annual Meeting Program

TD61

INFORMS Nashville – 2016

TD62 Cumberland 4- Omni Data and Decisions for Airline and Air Traffic Management Sponsored: Aviation Applications Sponsored Session

2 - Discovering Relationships Of Round-trip Carsharing Factors With Association Rules Technique Dahye Lee, Texas A&M University, College Station, TX, United States, dahyelee1991@tamu.edu, Luca Quadrifoglio, Benedetta Sanjust di Teulada, Italo Meloni The objective of this research is a comprehensive analysis for discovering relationships between factors of round-trip carsharing with the association rules approach. Results of analysis show that the strongest dependent variables do not have high correlations with the variables of distance from customers’ residence locations. Although the results gave an idea of connections of round-trip operations characteristics, the degree of impact of each variable still need to be investigated. The goal for future studies is to maximize connectivity to public transportation to help in reducing congestion and pollution. 3 - Household Use Of Autonomous Vehicles: Modeling Framework And Traveler Adaptation Yashar Khayati, University at Buffalo, Amherst, NY, United States, yasharkh@buffalo.edu, Jee Eun Kang, Mark Henry Karwan, Chase Murray We define a framework to model and evaluate potential household-level use of Autonomous Vehicles (AVs), to understand advantages, potential issues and negative external effects. We introduce a new formulation, the Household Activity Pattern Problem for AVs, to simulate the travel patterns of people using AVs. The key modeling challenge is to include modeling capabilities of driverless parking, pickup, drop-off and waiting during travelers’ engagement in activities. We develop solution approaches to this NP-hard problem and conduct a scenario analysis to evaluate changes in travel behavior. TD61 Cumberland 3- Omni Routing and Scheduling Sponsored: TSL, Freight Transportation & Logistics Sponsored Session Chair: Ali Ekici, Ozyegin University, TBD, Istanbul, TBD, Turkey, aliekici@gmail.com 1 - Integration Of Passenger And Freight Rail Scheduling Liang Liu, University of Southern California, Los Angeles, CA, United States, liangliu@usc.edu, Maged M Dessouky We study the integration of passenger and freight rail scheduling to improve the efficiency of freight trains while maintaining the punctuality of passenger trains. An optimization model that jointly considers the travel times of freight trains and the tardiness of the passenger trains is formulated. We proposed a decomposition based solution procedure to solve the problem, in which optimization-based or Santiago Carvajal, University of Southern California, 1150 W 29th Street, Los Angeles, CA, 90007, United States, scarvaja@usc.edu The optimization problem for efficiently routing multi-container trucks to better reposition both loaded and empty containers is studied. Our formulation adds the multi-container truck to the empty container reuse problem. Our aim is that by more efficiently routing trucks that the number of truck trips would be reduced, thus decreasing transportation costs, and reducing the natural environmental impact of transporting goods. 4 - A Tour Generation-based Algorithm For An Inventory Routing Problem We study a variant of inventory routing problem and develop an integrated two- phase solution approach. In the first phase, we cluster the customers such that each clustered is served by a single vehicle. Then, in the second phase, we determine the delivery routes and volumes for each cluster using an integer programming based heuristic approach. In this phase, we first generate several tours and solve mixed integer program to choose among these generated tours and determine the delivery volumes. We compare the performance of the proposed algorithm against the ones in the literature. heuristic algorithms are applied on each of the subproblems. 3 - Congestion Reduction Through Efficient Empty Container Movement Ali Ekici, Ozyegin University, Istanbul, Turkey, ali.ekici@ozyegin.edu.tr, Okan Orsan Ozener

Chair: Alexandre Jacquillat, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA, 02116, United States, alexjacq@mit.edu 1 - Flight Scheduling, Flight Planning And Operations Recovery To Minimize Airline Operating Costs Jane Lee, University of Illinois at Urbana Champaign, jjlee1@illinois.edu The focus of this work is to evaluate impact of stochasticity of disruptions on airline’s recovery decision. In particular, we aim to model the dynamic recoverability of flight schedule in response to disruptions based on Stochastic Queueing Model of airport congestion. We consider the typical mechanisms of departure time holdings, flight cancellations, and aircraft swaps used in aircraft recovery in practice today using Integer Programming. Additionally, we also consider dynamic decision making in recovery based on Dynamic Programming model. Our real-world experiments involve the original schedule of a major carrier in the US and disruptions at a secondary hub. 2 - A Combinatorial Auction For Allocation Of Departure And Arrival Slots Alexander Estes, University of Maryland-College Park, College Park, MD, 20742, United States, aestes@math.umd.edu, Michael O Ball, Mark M Hansen, Yulin Liu We present a combinatorial auction mechanism for the allocation of arrival and departures slots at an airport. This mechanism selects a profile of slots that will be available and provides an allocation of these slots to airlines based on their bids in the auction. Vickrey-Clarke-Grove payments are used so that it dominant strategy for airlines to bid truthfully. This provides a way in which the airlines’ valuation We propose models to incorporate customer choice behaviors into the capacity allocation problem under a network revenue management setting, namely, the airline fleet assignment problem. Unlike network revenue management, the capacity allocation problem with customer choice is usually intractable for real- world instances. We thus devise efficient decomposition approaches with provable performance guarantees. Our approach is data-driven in nature, which learns a choice model from transaction data and builds effective fleeting decisions based on that. 4 - A Model For Airport Schedule Coordination Based On The IATA Guidelines Nuno Ribeiro, University of Coimbra, Coimbra, Portugal, nuno_r_@hotmail.com The International Air Transport Association (IATA) slot allocation process is the dominant demand management mechanism used at busy airports worldwide. In this process, each coordinated airport provides its “declared capacity”, the airlines submit their scheduling requests, and a slot coordinator sets the schedule of flights at the airports. This research develops a new modeling approach to support slot coordinators to accommodate airline preferences better, while complying with the IATA guidelines and other constraints. Results are shown from a case study at Madeira airport (FNC). of congestion costs can be incorporated into slot allocations. 3 - Data-driven Choice-based Airline Fleet Assignment Chiwei Yan, Massachusetts Institute of Technology, chiwei@mit.edu, Cynthia Barnhart, Vikrant Vaze

354