Please stop now, automated vehicle! – Passengers aim to avoid risk experiences in interactions with a crossing vulnerable road user at an urban junction

Is the automated vehicle “aware” of the pedestrian? Examining driving behavior adaptation as a cue to inform the passenger of a potential hazard

“…As the positive effect of deceleration obtained in the driving simulator study was small, it may also be argued that longitudinal adaptation might only be necessary in more hazardous situations where passengers doubt that the HAV will safely handle the situation. For example, in a space-sharing conflict where the passengers of an HAV approached a Tjunction with a crossing pedestrian, early deceleration proved to be very effective in reducing the risk perceived by the passengers (Stange et al, 2022). Applied to the driving situation studied in this paper, longitudinal adaptation might only be necessary when the pedestrian enters the road and the HAV has to brake (early) to avoid a collision.…”

“…A growing body of studies are currently investigating passengers' risk perception during automated driving (e.g., He et al, 2022;Li et al, 2019;Pascale et al, 2021;Stange et al, 2022). Stange et al (2022) examined passengers' risk perception when approaching a T-junction with crossing pedestrians and bicyclists in an automated vehicle in a multi-method study, including an online video study, a driving simulator study, and a Hu&ViL study. The results showed that passengers prefer defensive automated driving behaviors, including early braking when approaching the T-junction to avoid any risk in interactions with VRUs.…”

“…Participants rated perceived risk on an 8-point scale which consisted of four categories: 1 (harmless), 2 to 4 (unpleasant), 5 to 7 (dangerous), and 8 (situation unacceptable). For the unpleasant and dangerous categories, participants indicated their ratings as little, medium, or very (see also Stange et al, 2022). In addition to perceived risk, the passengers' trust in the automated driving system (I trusted the automated vehicle in the previous situation.)…”

Incorporating Human Factors into Scenario Languages for Automated Driving Systems

“…1) Risk perception: A very important area of research for ADSs is the perceived risk of the users. One such instance is investigating the acceptable distance for braking for an ADS at a pedestrian crossing [26]. Part of this study, multiple similar scenarios were used to empirically determine the acceptable braking times.…”