Date of Award
Master of Science In Aviation Human Factors
John E. Deaton
Kenneth E. Crooks
Lars R. Jones
There is a modern revolution of pilot interface designs as technological advancements force industrial progress in general aviation (GA). Recent accident statistics and safety reports indicate that computerized flight decks in Technically Advanced Aircraft (TAA) present novel safety hazards. Little empirical data have been used to measure the usability of integrated flight decks due to the complexity and time-consuming nature of flight data assessment. The objectives of this thesis are (1) to describe the human factors issues that are induced by the design of the integrated flight deck, (2) to demonstrate the evaluative capabilities of a TAA flight simulator, and (3) to validate the performance of the TAA flight simulator using an experimental approach. The theoretical and practical applications of a data-driven, evaluative-simulation technique, which measures pilot performance and the usability of integrated flight decks, are discussed. Two experiments are presented to validate the functionality of the TAA flight simulator. In the first experiment, 24 pilots were tested under conditions with and without the flight director while hand-flying two basic attitude instrument maneuvers. Each pilot flew the four possible trials in a different order. Pilot performance was quantified using five objective measures
(standard deviation, root mean square error, number of deviations, time outside tolerance and mean time to exceed tolerance) for four flight parameters (altitude, heading, bank angle and vertical speed). Pilot dwell performance was analyzed using the total dwell time in four areas of interest in the flight deck (primary and secondary display, center console and outside view). In the second experiment, the same 24 pilots were tested to compare the usability of a Flight Management System (FMS) with and without the autopilot engaged. In one trial, the pilots flew a scenario with the autopilot while performing in-flight FMS tasks. Pilot IFD operation performance during in-flight FMS tasks was evaluated using the number of operations on the secondary display and keyboard, as well as the total time to complete the prescribed tasks. Pilot performance was measured using the five objective measures for two flight parameters (altitude and heading). In the other trial of the second experiment, the pilots flew the scenario without the autopilot and also performed in-flight FMS tasks. Results for Experiment 1 showed that altitude and heading contained less variability than vertical speed and bank angle. TD was the most sensitive derived performance measure; whereas, MTE and ND were the least sensitive of the derived performance measures. Within-subjects MANOVAs demonstrated significant differences for pilot performance on heading, vertical speed and bank angle across the flight director states. Generally, pilot performance is improved by the use of the flight director. The MANOVAs for dwell times did not reveal any significant changes across the flight director states. The results for Experiment 2 showed that altitude and heading were appropriate flight parameters for demonstrating the effect of autopilot
usage on pilot performance and IFD operation performance. All derived pilot performance measures yielded statistically significant results. Also, the number of operations on the secondary IFD and the task time yielded statistically significant results. However, the number of operations on the keyboard failed to reveal significant differences across the autopilot states. Overall, pilot performance was improved by the use of the autopilot. The evaluative-simulation technique has been used to determine whether or not the flight director, autopilot, and FMS achieve their purposes, to augment pilot performance. Evaluative simulations can be introduced during the system design-phase, then implemented during the system development-phase when prototype technology becomes available.
Weeks, Troy Ricardo, "Simulation for Evaluating the Usability of Integrated Flight Decks" (2005). Theses and Dissertations. 36.