A Plug and Play Framework for Automated Air Traffic Control

Abstract

A framework for air tra‐c control is developed with modularity in mind. This \plugand-play approach is based on deflning key agents in air tra‐c management (ATM) at the center, sector, and aircraft level. Also important are the interfaces between blocks and the uncertainty and robustness of these interfaces. In addition, the ability of this framework to incorporate new airspace structures makes it a useful construct in analyzing the ATM problem as various upgrades are made to the system in the coming years. The sector control block is chosen as a case study. First, the decision process of human controllers is explored. The advantages of human controllers with decision support tools and external information over automated algorithms are quantifled. Under such conditions, the human controller with decision support is found to signiflcantly outperform the automated algorithm and the human controller without decision support tools. With increasing delays and aging infrastructure, the air tra‐c control system in the United States is in need of improvements. The Next Generation Air Transportation System (NextGen) is a multi-agency efiort designed to improve e‐ciency in air travel as tra‐c levels triple in the next 20 years. This efiort includes a push to upgrade tracking and communication capabilities through technologies such as GPS and ADSB, respectively. Additional automation in the computation and implementation of sector- and center-level control strategies is being developed flrst through controller decision support tools and then through a more rigorous process of verifying safety in automated procedures. While the development of an automated air tra‐c system requires solving various challenging problems, the coming decades pose a separate challenge in the mixed-equipment era, as various systems are upgraded at difierent times. Various ideas have been proposed to deal with the process of bringing state-of-the-art technologies and improved control strategies to air tra‐c control. 1{4 Useful constructs such as 4D trajectories, which speciflc waypoints and timestamps for aircraft, have been developed to better link the roles of the difierent agents of air tra‐c. 5 At the same time, added ∞exibility is needed to utilize airspace efiectively. 6 As time progresses various new techniques and technologies can be applied to air tra‐c management (ATM), which is concerned with the control and routing of aircraft through the National Airspace System (NAS). 7 This paper proposes a hierarchical framework for air tra‐c control. This framework is designed to be very similar to the current NAS, but with additional requirements and interfaces between the difierent levels of the hierarchy. The goal of the extended framework is to allow for the automation of ATC but also to lay out a \plug-and-play architecture by which various components can be upgraded or added in stages. One speciflc problem, sector control, is highlighted as a block in this framework. Human controllers, possibly with decision support tools, are compared to automated controllers using a predictive mode estimation model. Given experiential information and better state awareness, humans are shown to react to tra‐c changes quicker than automated models in some situations. These abilities are then quantifled in terms of aircraft delay for a simple arrival ∞ow process. The outline of the paper is as follows. Section II contains a discussion of the current hierarchies of the NAS. Following this is the deflnition of the extended air tra‐c control framework. This discussion includes