Dissertation/Thesis Abstract

Development and Validation of a New Air Carrier Block Time Prediction Model and Methodology
by Litvay, Robyn Olson, Ph.D., The Ohio State University, 2012, 159; 10631204
Abstract (Summary)

Commercial airline operations rely on predicted block times as the foundation for critical, successive decisions that include fuel purchasing, crew scheduling, and airport facility usage planning. Small inaccuracies in the predicted block times have the potential to result in huge financial losses, and, with profit margins for airline operations currently almost nonexistent, potentially negate any possible profit. Although optimization techniques have resulted in many models targeting airline operations, the challenge of accurately predicting and quantifying variables months in advance remains elusive. The objective of this work is the development of an airline block time prediction model and methodology that is practical, easily implemented, and easily updated. Research was accomplished, and actual U.S., domestic, flight data from a major airline was utilized, to develop a model to predict airline block times with increased accuracy and smaller variance in the actual times from the predicted times. This reduction in variance represents tens of millions of dollars (U.S.) per year in operational cost savings for an individual airline.

A new methodology for block time prediction is constructed using a regression model as the base, as it has both deterministic and probabilistic components, and historic block time distributions. The estimation of the block times for commercial, domestic, airline operations requires a probabilistic, general model that can be easily customized for a specific airline's network. As individual block times vary by season, by day, and by time of day, the challenge is to make general, long-term estimations representing the average, actual block times while minimizing the variation.

Predictions of block times for the third quarter months of July and August of 2011 were calculated using this new model. The resulting, actual block times were obtained from the Research and Innovative Technology Administration, Bureau of Transportation Statistics (Airline On-time Performance Data, 2008-2011) for comparison and analysis. Future block times are shown to be predicted with greater accuracy, without exception and network-wide, for a major, U.S., domestic airline.

Indexing (document details)
Advisor: Benzakein, Meyer
Commitee: Gregorek, Gerald, Shen, Mo-How Herman, Taneja, Nawal
School: The Ohio State University
Department: Aero/Astro Engineering
School Location: United States -- Ohio
Source: DAI-B 78/11(E), Dissertation Abstracts International
Subjects: Statistics, Engineering, Industrial engineering, Transportation planning, Systems science, Operations research
Keywords: Air carrier operations, Airline block time, Block time estimation, Block time prediction
Publication Number: 10631204
ISBN: 978-0-355-01484-6
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy