Determining Air Traffic Complexity – Challenges and Future Development

Abstract

Air traffic complexity is one of the main drivers of the air traffic controllers’ workload. With the forecasted increase of air traffic, the impact of complexity on the controllers' workload will be even more pronounced in the coming years. The existing models and methods for determining air traffic complexity have drawbacks and issues which are still an unsolved challenge. In this paper, an overview is given of the most relevant literature on air traffic complexity and improvements that can be done in this field. The existing issues have been tackled and new solutions have been given on how to improve the determination of air traffic complexity. A preliminary communication is given on the future development of a novel method for determining air traffic complexity with the aim of designing a new air traffic complexity model based on air traffic controller tasks. The novel method uses new solutions, such as air traffic controller tasks defined on pre-conflict resolution parameters, experiment design, static images of traffic situations and generic airspace to improve the existing air traffic complexity models.

Defining Terminal Airspace Air Traffic Complexity Indicators Based on Air Traffic Controller Tasks

“…The main driver of ATCO workload is air traffic complexity [4]. It is defined as the difficulty of monitoring and managing a specific air traffic situation [5] and it can be explained as the ATCO's subjective experience of how complex a certain traffic situation is [6]. Air traffic complexity influences the workload of every type of air traffic control service within different airspace structures: area control in en-route airspace, approach control in terminal airspace, and aerodrome control in the control zone.…”

A Deep Unsupervised Learning Approach for Airspace Complexity Evaluation

“…Therefore, evaluation of airspace complexity is a non-trivial problem, attracting many investigations from scholars and field practitioners [6]- [16], [22]- [40]. Currently, the state-of-the-art methods for airspace complexity evaluation fall into two main categories: (1) Airspace complexity is described through a single indicator [6]- [14]; (2) Airspace complexity is evaluated based on a multi-indicator system [15], [16], [22]- [31].…”

“…In 1998, NASA Ames Research Center defined the dynamic density as a linear Biyue Li, Wenbo Du, Member, IEEE, Yu Zhang, Jun Chen, Ke Tang, Senior Member, IEEE, Xianbin Cao, Senior Member, IEEE A Deep Unsupervised Learning Approach for Airspace Complexity Evaluation W combination of 9 complexity factors [15]; this has become a standard and been implemented in the real system to evaluate airspace complexity until now [16]. The recent advent of machine learning technologies [17]- [21] encourages scholars to explore the nonlinear mapping between varying complexity factors and airspace complexity [22]- [31]. A pioneering work by Chatterji used the artificial neural network (ANN) to establish the nonlinear correlation for airspace complexity evaluation [22].…”

“…Aiming at reducing the high cost in obtaining labeled data, Zhu et al conducted a series of studies based on small samples [28]- [30] using integrated learning, semi-supervised learning, and transfer learning respectively, and obtained good evaluation results on six airspace sectors in Southwestern China. Similarly, the air traffic controller's tasks are incorporated into training data in order to mitigate the issue of small samples [31]. However, extraction of controller's tasks from traffic situation data is a non-trivial task.…”

Establishing the Correlation Between Complexity and Performance for Arrival Operations

“…However, the civil aviation industry will regain its vigour with several control measures. At this time, the ever-increasing demand for air traffic is doomed to cause airspace congestion, substantial flight delays, excessive fuel consumption and consequential air pollutant emissions, which pose a significant challenge to the ATM system [1]. Therefore, measuring air traffic control (ATC) performance and identifying the affecting factors could help enhance the levels of safety, capacity, efficiency and environmental sustainability [2].…”