Robot-like in-vehicle agent for a level 3 automated vehicle with emotions

Zeng, X., Alam, M. S., Bazilinskyy, P.

Submitted for publication.

ABSTRACT In-vehicle agents (IVAs) have emerged as a transformative innovation for intelligent transportation systems. This paper presents the development and evaluation of a robot-like IVA prototype with emotional feedback capabilities for SAE Level 3 automated vehicles. A user study assessed emotional interactions between drivers and IVA. The results showed that emotional feedback and driver working status did not have a significant effect on average workload or acceptance (usefulness and satisfaction). However, emotional feedback influenced physical and temporal demands, and its interaction with working status significantly affected the overall workload. Voice communication remained the main interaction mode, especially when drivers were engaged in other tasks. The study highlighted the challenges of accurately detecting emotions through facial recognition in automated driving scenarios, emphasised the need to consider physical conditions such as fatigue and stress, and insight into the participants' perspectives towards the IVA robot.