Pilot perceptions of airspace complexity. Part 2

Abstract

One of the distinguishing characteristics of the "free flight" concept is the sharing of aircraft separation responsibility between the pilot and the air traffic controller. In order to fulfill this responsibility, pilots will need to reliably detect and resolve conflicts as air traffic controllers do today. While there has been research on how air traffic controllers do this task, little is known about how pilots might do it. In particular, how pilots conceptualize the surrounding airspace and the relationships between their own aircraft and other aircraft may influence their decision making and the kinds of decision support they may need. In this study, fourteen pilots were given a series of short conflict scenarios that they had to resolve with and without the assistance of a cockpit display of traffic information that included an embedded conflict avoidance decision aid. Pilot success at resolving conflicts with and without the aid was measured. Pilots were also asked to rate airspace complexity, task difficulty, and aid acceptability. It was found that pilots performed significantly better with the aid, and that they rated the usability and value of the aid highly. A neural network model was then used to associate the measures of airspace complexity, derived from the spatio-temporal relationships between the pilot's ownship and other aircraft in the vicinity, to the pilot ratings of airspace complexity. The resulting analysis revealed the combinations of complexity measures that likely influence pilot perception of airspace complexity. Based on these results, a set containing eleven factors that appear to be most influential has been identified. Understanding of these factors may help designers of future air traffic management systems design conflict decision aids from a pilot's perspective that are likely to be more predictable and acceptable to pilots.