The objective of this study was to evaluate a speech-based system and assess general task and on-road performance in terms of vehicle control and subjective measures. There were 24 participants in this simulator study that ranged from 18 to 55 years of age. Participants were required to send a variety of text messages using either a handheld device or the speech-based system while operating the simulator. Participants drove a 3.5 km (2.2 mi) closed track with two traffic lanes that form a continuous loop and were instructed to maintain the speed of 72 km/h (45 mph).
Performance was measured on three subjective scales: mental demand, frustration level, and situational awareness. Results indicated that driving performance measures, glance behaviour, and subjective ratings were all significantly degraded during manual texting tasks as compared with similar tasks that were completed using voice commands. Performance during speech-based destination tasks was similar to that observed during speech-based texting. Participants generally believed that tasks requiring the use of the speech-based interface could be accomplished safely, and the majority of participants (83%) indicated their preference to have a speech-based interface available on their next vehicle. Results demonstrated that a greater level of difficulty was associated with texting via a handheld device compared to using the speech-based interface. The speech-based interface used in this study appeared easy to use and this was deemed to be an essential feature for implementation in a vehicle.
Neurauter, M., Hankey, J., Schalk, T., & Wallace, G. (2012). Outbound texting: Comparison of speech-based approach and handheld touch-screen equivalent. Transportation Research Record: Journal of the Transportation Research Board, (2321), 23-30.