Automatic speech recognition and text-to-speech technology can support a speech-baseddriver interface for in-vehicle computers. Data regarding cellular telephones suggest thatspeech-based interfaces may distract the drivers. This study uses a car following task toevaluate how a speech-based email system affects drivers’ response to a braking lead vehicle.A simple and a complex email system are compared in simple and complex drivingenvironments. The results show a 30% (300 msec) increase in reaction time when the speechbasedsystem is present. Subjective workload ratings and probe questions show that speechbasedinteraction introduces a significant cognitive load, which is highest for the complexemail system. These data show that speech-based email introduces a cognitive load that candegrade attention to the roadway, suggesting that indiscriminant use of speech-basedinterfaces for in-vehicle computers might compromise driving safety.
展开▼