Document Type


Publication Date





Twenty-four pilots viewed dynamic encounters between the pilot’s “ownship” and an intruder aircraft on a 2-D simulated Cockpit Display of Traffic Information (CDTI) and estimated the point and time of closest approach. A three-level alert system provided a correct categorical estimate of the projected miss distance (MD) on 83% of the trials. The remaining 17% of alerts incorrectly predicted MD. The data of these pilots were compared with a matched “baseline” pilots, who viewed identical trials without the aid of automated alerts. Roughly half the pilots depended on and benefited from this automation, and others did not. Those who benefited did so when problems were difficult but not when they were easy. Furthermore, automation benefits were observed only when automation was correct, but automation costs were not observed when it was in error. While assisting miss distance prediction, the automation led to an underestimate of the time remaining till the point of closest approach.