Autonomous electric vehicles (AEVs) will increasingly characterize future motorized traffic in urban areas. These vehicles are often silent and can be driverless. Therefore, understanding the interactions between these vehicles and vulnerable road users (VRUs) is crucial for traffic safety. This study explores the impact of an additional auditory alert system and external human-machine interfaces (eHMIs) on AEVs on cyclists' interaction behavior, comfort, and safety under varying vehicle–signal conditions and environmental noise levels. This study employed immersive Virtual Reality (VR) technology to simulate cycling across 18 traffic scenarios, using a within-subjects design with 40 participants. These scenarios included two vehicle types: autonomous conventional vehicles (ACVs) and autonomous electric vehicles (AEVs), two levels of ambient noise levels (quiet residential areas; noisier streets), and three levels of eHMI signal trigger distances (15, 20, 25 m). Data regarding cycling behavior (e.g., response distance, speed variation, gaze direction) and cyclists' perceptions (e.g., realism, simulator sickness, presence, trust, and perceived safety) were collected through VR and post-experiment questionnaires, respectively. We found that compared to vehicles equipped only with an eHMI, AEVs with auditory alerts significantly enhanced cyclists' perceptions and responses, with cyclists detecting the vehicles earlier and responding faster. Subjective feedback also indicated a stronger sense of safety and comfort, such as smoother deceleration that indicated natural responses and reduced stress. Participants also reported a more comfortable interaction experience with AEVs with auditory alerts. These findings provide evidence that equipping AEVs with additional auditory alerts improves the safety and comfort of cyclists in their interactions with AEVs in urban environments. Therefore, policymakers are advised to consider this when setting regulations for the deployment of AEVs in urban areas. ... Read more