Informs 2015

90 / 552

INFORMS Philadelphia – 2015
88
SB66
66-Room 113C, CC
Aviation Applications Section: Best Student
Presentation Competition 1
Sponsor: Aviation Applications
Sponsored Session
Chair: Bo Zou, University of Illinois at Chicago, 2095 Engineering
Research Facility, 842 W. Taylor Street (M/C 246), Chicago, IL, 60607-
7023, United States of America,
1 - Predicting Airport Arrivals using Data Mining Techniques
Sreeta Gorripaty, Student, University of California Berkeley,
116 McLaughlin Hall, UC Berkeley, Berkeley, CA, 94720,
United States of America, , Yi Liu,
Alexey Pozdnukhov, Mark Hansen
One way to define similarity between two days in the NAS is measuring the
difference in their airport arrival counts. Similar historical days can be extracted
by predicting the airport arrivals for a given day and can be subsequently used in
decision support tools. In our work, we use machine learning algorithms of
support vector regression and random forest to predict the number of arrivals.
These models also quantify the contribution of factors such as weather and
demand, on airport arrivals.
2 - Airline Alliance and Product Quality: The Case of the U.S.
Domestic Airline Industry
Jules Yimga, Kansas State University, KS,
United States of America, , Philip Gayle
Collusion on price and service levels tends to be the main concern of
policymakers when appraising an airline alliance formation. We posit that product
quality is an important dimension to be considered in alliance appraisals. This
paper investigates the product quality implications of the
Delta/Continental/Northwest codeshare alliance with a particular focus on the
codeshare effects in markets where the alliance partners competed prior to the
alliance.
3 - Statistical Analysis of Dispatcher Fuel Loading Behavior
Lei Kang, University of California, Berkeley, 107D McLaughlin
Hall, UC Berkeley, Berkeley, Ca, 94720, United States of America,
, Mark Hansen, Megan Ryerson, Lu Hao
Airlines are moving aggressively to reduce fuel consumption. Thus, gaining a
better understanding of dispatcher fuel loading behavior regarding discretionary
fuel (i.e. contingency fuel and alternate fuel) is of great interest to airlines. By
combining a large flight level fuel loading dataset with a dispatcher survey
provided by a major US airline, dispatcher heterogeneity and its impacts in
loading discretionary fuel regarding domestic and international flights will be
quantified and compared.
SB67
67-Room 201A, CC
Working Towards the Physical Internet
Sponsor: TSL/Freight Transportation & Logistics
Sponsored Session
Chair: Tom Van Woensel, Full Professor, Eindhoven University of
Technology, Den Dolech 2, Eindhoven, NB, Netherlands,
1 - Exploring Operational Problems for Future Delivery Service
Operations using Unmanned Aerial Vehicles
Heng Chen, University of Massachusetts Amherst,
Isenberg School of Management, Amherst, MA, 01003,
United States of America, Senay Solak
It is well accepted that commercial use of UAVs in the near future will involve
delivery service operations by retailers and courier companies. We outline some
strategic and tactical decisions that these companies will face in UAV based
delivery operations, and use currently available data to develop preliminary
models for guiding such decisions by a firm. Specifically, optimal policies on
certain capacity and revenue management decisions under stochastic demand are
studied and derived.
2 - Crowdshipping for Same-day Delivery
Niels Agatz, Rotterdam School of Management, Erasmus
University, Burgemeester Oudlaan 50, Rotterdam, Netherlands,
Alp Arslan
Crowdshipping entails obtaining transportation services from approved drivers
and carrier companies with spare capacity rather than from traditional logistics
partners. The key idea is to exploit unused capacities and existing transportation
flows of the crowd to save delivery cost and provide faster delivery. This study
investigates the use of crowdshippers for last-mile delivery in a same-day delivery
setting.
3 - The Pickup and Delivery Problem with Time Windows,
Scheduled Lines and Stochastic Demands
Tom Van Woensel, Full professor, Eindhoven University of
Technology, Den Dolech 2, Eindhoven, NB, Netherlands,
This paper concerns scheduling a set of vehicles to serve a set of requests, whose
expected demands are known in distribution when planning, but are only
revealed with certainty upon vehicles’ arrival. In addition, a part of the transport
plan can be carried out on limited-capacity scheduled public transportation lines
using bus, train, tram, metro, etc. We present the model, solution approach and
numerical results.
SB68
68-Room 201B, CC
Joint Session TSL/Public Sector: Transportation
Issues in Emergency Response
Sponsor: Transportation, Science and Logistics & Public Sector
Sponsored Session
Chair: Kash Barker, Associate Professor, University of Oklahoma, 202
W Boyd St, Rm. 124, Norman, OK, 73019, United States of America,
1 - Combining Worst and Average Case Considerations in an
Integrated Emergency Response Network Design
Jyotirmoy Dalal, Assistant Professor, Indian Institute of
Management Udaipur, IIM Udaipur, MLSU Campus, Udaipur,
313001, India, , Halit Uster
We design an emergency response network, integrating relief (supply) and
evacuation (demand) sides under uncertainties in demand amount and location.
We formulate a MIP, combining stochastic and robust optimization concepts as
the weighted sum of the corresponding objectives. For varying relative weights,
we devise alternative approaches to solve large scale problem instances. We
present computational result and insights gained by applying the model to a GIS-
based case study on coastal Texas.
2 - A Multi-period Dynamic Location Planning Model for
Emergency Response
Burcu Keskin, Associate Professor, The University of Alabama,
345 Alston Hall, 361 Stadium Drive, Tuscaloosa, AL, 35487,
United States of America, Jianing Zhi,
Sharif Melouk
We propose a deferred service model to analyze the daily operations of
ambulances that involve dispatch and redeploy. Given incident priority levels and
patterns, we decide which incidents to serve immediately or to defer to following
periods while incurring deferral penalty costs. Considering network size, fleet size,
incident patterns, and time-dependent parameters, we compare the results in
terms of service, response time, and cost through experimentation and offer new
dispatch policies.
3 - Replenishment Location Planning for Service Trucks under
Network Congestion and Routing Constraints
Leila Hajibabai, Clinical Assistant Professor, Washington State
University, 405 Spokane St., Sloan 119, P.O. Box 64291, Pullman,
WA, 99164, United States of America,
It is often challenging to plan service trucks operations under network design
constraints, especially on congested roadways. An integrated model is developed
to simultaneously determine the optimal location design of replenishment
facilities, minimize routing cost in the proposed network, assign traffic, and select
candidate links for capacity expansion. A genetic algorithm is proposed that
integrates a continuous approximation model for routing cost estimation and
traffic assignment algorithm.
SB66