Pilot performance with mid-flight plan-based display changes

The paper extends previous experimental results from plan based display design. These results suggested that performance variability depends on an interaction between the methods (or plans) in use and the display with which they are paired. This result implies that displays should change as plans change. To examine the performance consequences of automated display change, the present experiment incorporated four different displays: 1) Cruise-in which speed is controlled using the autopilot; 2) Descent 1-a slight modification of the cruise display, designed to facilitate descent with manual speed control; 3) Descent 2-a major modification of the cruise display, designed to maximize the compatibility of the display with manual speed control; 4) Baseline-provides all of the tailoring for cruise and descent simultaneously. Specific pairings of displays and flight phase allow us to examine two hypotheses about uncommanded display changes in flight: 1) changes that optimize a display for current plans will facilitate performance, and 2) large differences in temporally adjacent displays will not be disruptive when they increase the compatibility between plans and display. Preliminary results obtained from five commercial pilots are generally consistent with these hypotheses