Keep Your Distance, Automated Vehicle! – Configuration of Automated Driving Behavior at an Urban Junction from a Cyclist’s Perspective

Abstract

In the future, cyclists will interact with highly automated vehicles (HAV) in urban mixed traffic, especially at junctions and intersections. Previous research focused on HAV interactions with pedestrians, whereas HAV interactions with cyclists are yet understudied. Therefore, we conducted two studies (video, cycling simulator) to examine how an HAV should behave in an interaction with cyclists. Specifically, we explored when cyclists want the HAV to start braking and how much perceived risk cyclists are willing to accept in HAV interactions. Results show that cyclists’ choice of braking onset of the HAV depended on the timing (ideal vs. last acceptable) and vehicle speed (30 km/h vs. 50 km/h). Subjective ratings showed that cyclists in both studies aim to avoid experiencing risk in HAV interactions but also accepted riskier interactions in the cycling simulator. The studies provide first insights into cyclists’ risk perception in an HAV interaction and contribute first ideas to the design of HAV driving behavior in the interaction with cyclists.