Understanding Controller-Pilot Interaction Dynamics in The Context of Air Traffic Control

Abstract

Literature New capabilities to modernize the U.S. National Airspace System (NAS) include support of real-time information streams derived from many data sources across the NAS. As an emergent property, safety of the NAS arises from interactions between many elements at different levels, ranging from those attributable to humans, technology, and the environment. Each component in the NAS needs to interact with other components, exchange resources and information, and operate under broad regulations to achieve overall system objectives (Harris & Stanton, 2010). Sometimes, incidents and accidents result from insufficient interaction (communication and coordination) between humans (e.g., pilot-controller). The content of communication provides value and supports understanding with a multitude of individual, group, team, and data sets within air traffic research. In addition, another dimension to communication with a potentially rich source of understanding is everything outside of its explicit meaning. Cooke and Gorman (2009) describe methods of communication flow between teams (considered to be a system) that have proven insightful. The first is a ratio of team members speech quantity, which can indicate the degree of influence one member has over others. Another is the communication required and passed score, or how much variation there is in actual team communication from expectations. Flow quantity represents how much speech each member of the team produces. Gorman et al.’s (2012) study applied discrete Recurrence Quantification Analysis (RQA) to team communication flow data in order to visualize and measure coordination dynamics of Unnamed Aircraft Vehicle (UAV) teams, both mixed teams (i.e., team members changed) and intact teams (i.e., team members stayed the same over successive experimental sessions). Interestingly, mixed teams were better able to adjust to unexpected perturbations; this ability was linked to team level coordination dynamics. That is, mixed teams adopted a globally stable pattern of communication while exhibiting strong temporal dependence (Gorman, Cooke, Amazeen, & Fouse, 2012). Similarly, Demir, Cooke, & Amazeen (2018) applied discrete RQA on human-robot interaction in an Urban Search and Rescue task and multivariate extension of RQA on human-synthetic team in a UAV task. They underline that metastable team coordination (not too stable nor too flexible) between team members is associated with the ability to successfully overcome novel events (i.e., team situation awareness) in a dynamic task environment. The current project addresses the question of how human factors related to air traffic control (ATC), specifically situation awareness and cognitive load, interact with other factors in the NAS to affect ATC performance and a result in a safe and effective NAS? One way to answer this question is focusing on ATC-pilot communication as a chief performance indicator. In the current study, we investigate the potential of dynamical systems perspectives to capture the differential dynamics of three cases between controller-pilot communication flow during incidents and accidents. Method One of the approaches for investigating interaction patterns between system components (in the controller-pilot case) and their change over time involves looking at communication flow using discrete Recurrence Plot (RP) and corresponding Recurrence Quantification Analysis (RQA), which quantifies how many recurrences with a certain length are present by multidimensional space (phase space) trajectory in a dynamical system (Marwan, Carmen Romano, Thiel, & Kurths, 2007). RP is the basis of discrete RQA (Eckmann, Kamphorst, & Ruelle, 1987), which is a visual tool for demonstrating a system’s recurrent structure in the phase space when a system revisits specific states or sequences of states within a region of phase space over a period of time. In the case of two or more systems, discrete RP displays the times wh