Agent-Based Modelling and Analysis of Non-Autonomous Airport Ground Surface Operations

Abstract

The highly fuel-inefficient aircraft taxi-phase causes the industry to consider alternative ground propulsion systems. External systems like the Taxibot are preferred over onboard systems as they do not require aircraft structural additions and recertifications. However, the operational implementation of external taxi-systems is expected to increase both traffic complexity and human workload for Air Traffic Control (ATC). Systematic assessment of operational safety and efficiency consequences of novel taxi-concepts employing automated external systems within a hub airport’s ground surface operation has not yet been considered in existing research. This study has taken the first steps into the design and evaluation of a novel taxi-concept for outbound aircraft enabled by autonomous Taxibots. An Agent-Based model is created to explore the novel concept of operations within the ground surface operations of Amsterdam Airport Schiphol in the Netherlands. Four days of real-world data from ADS-B ground tracks have been used to simulate the novel taxi-concept for different operational scenarios. The aircraft taxi-time was found to significantly increase for outbound aircraft in the novel taxi-concept compared to real-world taxi operations, especially for large taxi-distances. The number of necessary Taxibots in the novel taxi-concept and the aircraft engine-off times significantly depended on the considered operational scenario. Distributed coordination and planning has shown to allow for safe and efficient guidance of heterogeneous vehicle types within an airport ground surface operation for increasing traffic complexity.