Scholarly
    Communications Project


Document Type:PhD Dissertation
Name:Andrew William Gellatly
Email address:gellatly@ctr.vt.edu
URN:
Title:The Use of Speech Recognition Technology in Automotive Applications
Degree:Doctor of Philosophy
Department:Industrial and Systems Engineering
Committee Chair: Thomas A. Dingus
Chair's email:
Committee Members:Robert C. Williges
John G. Casali
Brian M. Kleiner
Raymond J. Kiefer
Keywords:Driver Behavior, Dual-task Performance, Decision Tools
Date of defense:March 28, 1997
Availability:Secure the entire work for patent or proprietary purposes.
After one year release worldwide only with written permission of the student and the advisory committee chair.

Abstract:

The research objectives were (1) to perform a detailed review of the literature on speech recognition technology and the attentional demands of driving; (2) to develop decision tools that assist designers of in-vehicle systems; (3) to experimentally examine automatic speech recognition (ASR) design parameters, input modalities, and driver ages; and (4) to provide human factors recommendations for the use of speech recognition technology in automotive applications. Two experiments were conducted to determine the effects of ASR design parameters, input modality, and age on driving performance, system usability, and driver preference/acceptance. Eye movement behavior, steering input behavior, speed maintenance behavior, reaction time to forward scene event, task completion time, and task completion errors when driving and performing in-vehicle tasks were measured. Driver preference/acceptance subjective data were also recorded. The results showed that ASR design parameters significantly affected measures of driving performance, system usability, and driver preference/acceptance. However, from a practical viewpoint, ASR design parameters had a nominal effect on driving performance. Differences measured in driving performance brought on by changes in ASR system design parameters were small enough that alternative ASR system designs could be considered without impacting driving performance. No benefits could be claimed for ASR systems improving driving safety/performance compared to current manual-control systems. Speech recognition system design demonstrated a moderate influence on the usability of in-vehicle tasks. Criteria such as task completion times and task completion errors were shown to be different between speech-input and manual-input control methods, and under different ASR design configurations. Therefore, trade-offs between ASR system designs, and between speech-input and manual-input systems, could be evaluated in terms of usability. Finally, ASR system design had a nominal effect on driver preference/acceptance. Further research is warranted to determine if long-term use of ASR systems with less than optimal design parameters would result in significantly lower values for driver preference/acceptance compared to data collected in this research effort. Human factors recommendations for the use of ASR technology in automotive applications are included. The recommendations are based on the empirical research and the literature review on speech recognition technology and the attentional demands of driving.

List of Attached Files

etd.pdf


The author grants to Virginia Tech or its agents the right to archive and display their thesis or dissertation in whole or in part in the University Libraries in all forms of media, now or hereafter known. The author retains all proprietary rights, such as patent rights. The author also retains the right to use in future works (such as articles or books) all or part of this thesis or dissertation.