Cyclists’ Crossing Intentions When Interacting with Automated Vehicles: A Virtual Reality Study

Abstract

Most of cyclists’ fatalities originate from collisions with motorized vehicles. It is expected that automated vehicles (AV) will be safer than human-driven vehicles, but this depends on the nature of interactions between non-automated road users, among them cyclists. Little research on the interactions between cyclists and AVs exists. This study aims to determine the main factors influencing cyclists’ crossing intentions when interacting with an automated vehicle as compared to a conventional vehicle (CV) using a 360° video-based virtual reality (VR) method. The considered factors in this study included vehicle type, gap size between cyclist and vehicle, vehicle speed, and right of way. Each factor had two levels. In addition, cyclist’s self-reported behavior and trust in automated vehicles were also measured. Forty-seven participants experienced 16 different crossing scenarios in a repeated measures study using VR. These scenarios are the result of combinations of the studied factors at different levels. In total, the experiment lasted 60 min. The results show that the gap size and the right of way were the primary factors affecting the crossing intentions of the individuals. The vehicle type and vehicle speed did not have a significant effect on the crossing intentions. Finally, the 360° video-based VR method scored relatively high as a research method and comparable with the results of a previous study investigating pedestrians’ crossing intentions confirming its suitability as a research methodology to study cyclists’ crossing intentions.

Blinded windows and empty driver seats: The effects of automated vehicle characteristics on cyclists’ decision‐making

“…Our findings are in line with a simulator study by TRL [49], which reported that AV distinguishability had no significant effect on gap acceptance, a video‐based experiment by Dey et al. [50], which concluded that knowledge of a vehicle driving mode does not play a significant role in pedestrians’ crossing behaviours, and a study using a head‐mounted display which found that vehicle type did not have a significant impact on cyclists’ crossing intentions (Nuñez Velasco et al [51]). A possible reason for the lack of effect is that stickers and a lidar are hard to detect, as Experiment 2 showed.…”

Innovations in Transport

“…Despite that the vehicle type affected the perceived risk, no significant difference was found in the crossing intention. However, pedestrians who did recognize the vehicle as an AV had, overall, lower intentions to cross (Nuñez Velasco et al, 2019). For cyclists, the gap size and the right of way were found to be the primary factors affecting the crossing intentions of the individuals.…”

Automated vehicles that communicate implicitly: examining the use of lateral position within the lane

“…Moore et al (2019) made a case against eHMIs and reported that implicit cues are dominant; they reported no significant differences in interaction quality for pedestrian-vehicle encounters at a crosswalk, with and without a driver. Several other studies concur that implicit communication is a more important cue than explicit signs, gestures, or other vehicle features (Clamann, Aubert, and Cummings 2017;Dey et al 2019;Lee et al 2020;Nuñez Velasco et al 2020;Rothenb€ ucher et al 2016).…”