Automated Flight Plan Filing by Simulated Voice Recognition

摘要

The Systems Research and Development Service (SRDS) asked the Federal Aviation Administration (FAA) Technical Center to develop and demonstrate the capability of automatic flight plan filing by computerized word recognition. To accomplish a prototype capability, it was necessary to begin gathering human factors data to help determine the correct technical approach and design concept. The tests reported herein centered on two questions: Would pilots use this method of flight plan entry. Which of the three protocols simulated do pilots prefer most. Because of the limitations inherent in this experiment, the answers to these questions are tentative at this time. Given the assumption that the actual utterance recognition device (URD) performs at a level comparable to the simulation used in this experiment, it is entirely reasonable to conclude that general aviation pilots would (1) elect to file flight plans by computerized word recognition using a real computer and (2) prefer 'no prompt' or a 'word echo-back prompt' over the 'cue-tone prompt.' The results show that the preferred protocol was the 'no prompt' method. (Author)