A Machine Learning Model for Normal and Extended Taxi-Out Time Prediction
Vienna Airport Case Study
M.A. Probyn (TU Delft - Aerospace Engineering)
Paul Roling – Mentor (TU Delft - Air Transport & Operations)
Floris Herrema – Graduation committee member (EUROCONTROL)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
All major airport operators face a similar challenge, namely ensuring maximum throughput and maintaining high runway utilisation. A key part of this is accurately planning aircraft movements on the ground to avoid queueing and associated delays. A primary indicator of the operator performance in this area is the Taxi-Out Time. The research objective of this article is to review whether the application of machine learning can be used to model the departure process in such a way as to provide accurate prediction of TXOT taking into account a wide range of variables. A regression tree type machine learning model is developed using actual data from Vienna Airport and a selected set of significant predictor variables. The taxi-out times of the test set of flights are closely predicted with an RMSE of 2.03 minutes for normal taxi-out and 3.75 minutes for extended taxi-out.