Investigating HMIs to Foster Communications between Conventional Vehicles and Autonomous Vehicles in Intersections

Abstract: In mixed traffic environments that involve conventional vehicles (CVs) and autonomous vehicles (AVs), it is crucial for CV drivers to maintain an appropriate level of situation awareness to ensure safe and efficient interactions with AVs. This study investigates how AV communication through human-machine interfaces (HMIs) affects CV drivers' situation awareness (SA) in mixed traffic environments, especially at intersections. Initially, we designed eight HMI concepts through a human-centered design process. The two highest-rated concepts were selected for implementation as external and internal HMIs (eHMIs and iHMIs). Subsequently, we designed a within-subjects experiment with three conditions, a control condition without any communication HMI, and two treatment conditions utilizing eHMIs and iHMIs as communication means. We investigated the effects of these conditions on 50 participants acting as CV drivers in a virtual environment (VR) driving simulator. Self-reported assessments and eye-tracking measures were employed to evaluate participants' situation awareness, trust, acceptance, and mental workload. Results indicated that the iHMI condition resulted in superior SA among participants and improved trust in AV compared to the control and eHMI conditions. Additionally, iHMI led to a comparatively lower increase in mental workload compared to the other two conditions. Our study contributes to the development of effective AV-CV communications and has the potential to inform the design of future AV systems.