Informs Annual Meeting 2017

SB32

INFORMS Houston – 2017

SB32

2 - Probabilistic Modeling of Airline Network Evolution Reed Harder, Dartmouth, 14 Engineering Drive, Hanover, NH, 03755, United States, reed.haseltine.harder.TH@dartmouth.edu Airline networks evolve as airlines enter and exit segments, based on the profitability of the routes related to this segment, competitive considerations, and the function of a given segment in the overall network. However, forecasting such segment entry and exit remains challenging due to the complexity of the air transportation system and the relative sparsity of entry and exit data. We present and compare probabilistic methods for predicting segment entry and exit by an airline, using data on past airline operations, network structure, and regional demographics. 351D Operations Research and Cyber Sponsored: Military Applications Sponsored Session Chair: Andrew Oscar Hall, United States Military Academy, West Point, NY, 10996, United States, andrew.hall@usma.edu 1 - A Sequential Weighted Laplacian Regularized Optimal Design of Experiments for Response Surface Modeling of Expensive Tests Stanford Martinez, Graduate Research Assistant, University of Texas at San Antonio, San Antonio, TX, United States, ddfe530tx@gmail.com, Adel Alaeddini In the design of experiments, space-filling methods can be used to reduce the number of tests needed to learn the response surface intrinsic to an experiment. Classical methods include preset designs that only take into account currently measured samples. Laplacian regularization adds a penalty term that will allow the learning of the interconnectivity of both measured and unmeasured data. In addition, further weighting can be applied using weighted regression techniques to mitigate effects of outliers present in an experiment. The effects of these two additions to the classical form of optimal experimental design will be compared to other methods in terms of performance and practicality-of-use. 2 - Critical Nodes in Interconnected Networks Venkat Venkateswaran, Georgia Institute Of Technology, Room 4143, 800 West Peachtree, NW, Atlanta, GA, 30308, United States, venkat.venkateswaran@scheller.gatech.edu We consider a system of interconnected networks. Initial damage to nodes can have far-reaching effects through cascading failures. We develop the methodology to identify nodes that may produce various levels of network degradation through such cascading. Computational results from sample test problems are presented. 3 - Queue Inference with Uncertain Observations Although the effect of various queue types on system performance has been well studied, the internal structure and parameters of a queuing system of interest may be completely unobservable in military and competitive commercial applications. Furthermore, external input and output information may be observable but subject to substantial uncertainty due to measurement error in the adversarial environment. This analysis quantifies the uncertainty in arrival and departure times, which are then used to estimate the number of servers and service rates of an internally unobservable GI/G/c queue. 351E Supply Chain Management Contributed Session Chair: Juhi Bhardwaj Sapra, Rensselaer Polytechnic Institute, Troy, NY, United States, bhardj@rpi.edu 1 - The Interactive Effects of User Generated Contents on the Determinants of Tourist Destinations Strategies Jinjin Xu, Tianjin University, Tianjin, China, x_jinjin@outlook.com, Liming Zhao Vigorous development of tourism industry and sharing economy has led to fierce competition among China’s tourist destinations, which accelerates tourism organizations to re-evaluate their competitive strategies. This study conducted from a strategy perspective and contributes both theoretically and practically to the tourism literature by validating artificial neural network model to investigate the relationships among user generated content and tourist destinations strategies. The empirical findings reveal four hidden nodes that have a significant impact on tourist destination performance. SB34 Andrew Keith, PhD Student, USAF, Dayton, OH, 45433, United States, andrew.keith@afit.edu, Darryl K. Ahner SB35

351B Drone Delivery Systems - II Invited: InvitedDrone Delivery Systems (tentative title) Invited Session Chair: Chase Murray, University at Buffalo, Amherst, NY, 14260-2050, United States, cmurray3@buffalo.edu 1 - Establishing Energy Efficient Paths for Delivery Drones in Direction Dependent Uniform Wind-Fields Abhishake Kundu, Texas Tech University, Dept of Ind Eng, Box 43061, Lubbock, TX, 79409-3061, United States, abhishake.kundu@ttu.edu, Tim Matis The purpose of this research is to propose an efficient minimum-energy path planning algorithm that combines global and local waypoints for parcel delivery using drones. Constraints on turn rate and turn acceleration is considered for the vehicle with limited battery power, flying in a uniform wind-field on the horizontal plane. The minimum energy path of the delivery drone pursues Dijkstra’s arcs subject to curvature constraints for transitional headings, in contrast to minimum time traversal route that follows a Dubins path. 2 - The Customer-centric Vehicle Routing Problems with Drones Mohammad Moshref-Javadi, PhD Candidate, Purdue University, This research proposes a new drone delivery system that focuses on minimizing the total waiting time of recipients. In this delivery system, a single truck takes drones and supplies and follows a route to visit some demand locations or dispatch the drones to customers for package delivery. The truck and drones can work in parallel, that is, the truck does not have to wait for the drones to return to the truck at the same location. We also compare this system with a static system in which the truck has to wait for the drones to return to the truck at the same location. This drone delivery system is specifically useful for disaster relief logistics and customer-oriented commercial systems. 3 - Augmenting a Truck with a Drone for Parcel Delivery John Gunnar Carlsson, University of Southern California, 3750 McClintock Avenue, Los Angeles, 90089, United States, jcarlsso@usc.edu Abstract We determine the efficiency of a delivery system in which an unmanned aerial vehicle (UAV) provides service to customers while making return trips to a truck that is itself moving. In other words, a UAV picks up a package from the truck (which continues on its route), and after delivering the package, the UAV returns to the truck to pick up the next package. 4 - Drone Delivery with Time Dependent Travel Chase Murray, University at Buffalo, Dept of Industrial & Systems Engineering, 309 Bell Hall, Amherst, NY, 14260-2050, United States, cmurray3@buffalo.edu A new variant of drone delivery problems is presented. In this problem, multiple drones are available to support last-mile delivery of small parcels. The drones coordinate with a single delivery truck. Travel times for the truck are time-of-day dependent. A mathematical programming formulation is described, as well as a heuristic approach for solving this problem. 315 N. Grant St., Room 324E, West Lafayette, IN, 47907, United States, moshref@purdue.edu, Seokcheon Lee 351C AAS Best Student Presentation Competition I Sponsored: Aviation Applications Sponsored Session Chair: Virginie Lurkin, Ecole Polytechnique Fédérale de Lausanne, Route Cantonale, Lausanne, 1015, Switzerland, vlurkin@ulg.ac.be Co-Chair: Peng Wei, Iowa State University, Ames, IA, 50011, United States, pwei@iastate.edu 1 - Predicting Performance of Traffic Management Initiatives Alexander Estes, University of Maryland, 3802 Calverton Blvd, Apartment 25, Beltsville, MD, 20705, United States, aestes@math.umd.edu The Federal Aviation Administration uses traffic management initiatives to manage excess demand for resources in the national airspace system. We present methods for estimating the performance that a given traffic management initiative will produce under a given set of weather and traffic conditions. These methods combine a weighting scheme from Geographic Weighted Regression with forest-based regression methods. Estimates of the conditional density can be produced along with standard regression estimates. SB33

46