Abstract
This research is aimed at developing an in-vehicle voice-based interface. Integrating and generalizing those pros and cons of the few existing speech control systems and through user scenario analysis, operational process (including the voice dialogue design), panel interface design and voice commands collection in different operational process, a voice-based interface is developed which is called Talking Car. By conducting a driving simulation experiment, the driving performance of Talking Car and that of simulated LUXGEN Think+ system in radio and music functions were compared.Two cameras were set to capture eye movements and drivers response to collect users’ every single glance time, to measure the reaction time of users before stepping on the brakes when a “STOP!!!” message appears on the display, to determine times of lane change and also task completion time, while carrying out the driving task. The NASA-TLX and a subjective questionnaire were also applied to evaluate subjects’ mental workload.Regarding subjects’ response and workload, the Talking Car System required less glance time, less manual distraction occurred and the mental workload was less than that of using the LUXGEN Think+ System. Therefore, with the Talking Car System drivers are able to focus on their driving when they are controlling the AV System in-vehicle and so have a safer response towards road situations.