Prediction of Actual Takeoff Weight Using Machine Learning Algorithms and Automatic Dependent Surveillance-Broadcast Data

The Actual Takeoff Weight (ATOW) is a significant and critical aircraft performance parameter that has a huge impact on not only fuel consumption and derived gaseous emissions but also on flight trajectory prediction. Aircraft mass data, including Actual Takeoff Weight, is typically considered sensitive information by airlines due to its reliance on factors such as passenger and cargo loads, as well as operational strategies and as such have a competitive meaning. As a result, this data is generally not disclosed to researchers or entities outside the operating airline. Automatic Dependent Surveillance-Broadcast (ADS-B) technology offers precise, up-to-the-minute data on aircraft velocity and acceleration. This information is publicly accessible and not exclusively controlled by any particular airline. This study demonstrates the potential of machine learning algorithms, which using relevant ADS-B data, can predict the Actual Takeoff Weight of the aircraft with high accuracy.