Emotion-matched in-vehicle assistants: an exploration in regulating drivers' incidental emotions and enhancing takeover performance and situational awareness

AbstractIn conditional autonomous driving, drivers are prone to experiencing incidental emotions during non-driving tasks, which are unrelated to driving decisions but may have negative effects on drivers' takeover performance. Voice assistants have been shown to regulate incidental emotions. Therefore, this study endowed the in-vehicle assistant with emotional voice capability, matching the emotional state of the driver in conditional autonomous driving, in order to explore the effect of regulating incidental emotions. Forty participants were enrolled in a driving simulation experiment. During the autonomous driving process, participants were induced with two incidental emotions (happy or sad) and underwent experiments combining two emotional states of the driver and two emotional states of the in-vehicle assistant. The purpose was to examine the participants' takeover performance in autonomous driving, eye-tracking data, situation awareness, and to reveal the preferences of participants in different groups regarding the in-vehicle emotional voice assistant and their trust in the automated system. The results showed that in-vehicle emotional voice assistants that matched drivers' incidental emotions exhibited better effects on emotional regulation, as evidenced by shorter takeover response time, smaller standard deviation of lane position, smaller maximum lateral acceleration, and better situation awareness capabilities. Drivers also prefer in-vehicle emotional voice assista...
Source: Cognition, Technology and Work - Category: Information Technology Source Type: research