A Study on a Difficulty Index of Air Traffic Control

Abstract

Monitoring performance is an essential part of an air traffic management system, and requires metrics such as complexity and safety corresponding to the monitoring objectives. Research aiming to develop a difficulty index of air traffic control (ATC) to predict airspace “hot spots”, that is, areas of potentially high difficulty in a trajectory-based operations environment was carried out as part of a priority research project at ENRI. This report outlines the research on a creating a metric for ATC difficulty and examining its applicability to ATC support systems.

The basic component of the proposed metrics to be used as a basis to calculate the difficulty index focuses on the proximity situation of an aircraft pair calculated based on a kind of four-dimensional distance that includes “lookahead” time as well as physical separation, and its temporal evolution is linearly predicted. The time component of the metric is scaled using air traffic controller subjective judgements of the “difficulty” of proximity situations. The metric for an airspace volume is then synthesized from the individual metrics of all aircraft pairs in the airspace. Values of the difficulty index can be easily estimated from the time, position and velocity vectors of aircraft observed by a surveillance system and trajectory change point (TCP) information from flight plan data, and can be used for the real-time computation of “hot-spots” in the airspace. The sensitivity of the metric to observation error was also clarified.

This report first describes the concept and characteristics of the metrics, and calculation methods that reflect practical flight paths with trajectory change points. Next, several developments for estimating the difficulty metric in practice, such as methods for scale parameter determination and calculation using TCP data, are described. To evaluate the metric, its parameters were first calibrated in a cognitive experiment in which controllers judged the perceived difficulty of a traffic scenario containing a simple conflict. The values for total airspace difficulty index, and its distribution using the calibrated metric were then verified for other more complex scenarios. Finally, concept of a high-difficulty avoidance ATC support tool based on the metric was constructed, and its feasibility was demonstrated by a computer simulation using a simple conflict resolution model (limited to altitude changes). The feasibility of the application was mostly confirmed.