Pedestrians’ Crossing Decisions While Interacting with Automated Vehicles – Insights from a Longitudinal Study

Joisten, Philip; Theobald, Nina; Abendroth, Bettina · 2022 · Crossref

DOI: 10.54941/ahfe1002445

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates how pedestrians’ perceived safety and crossing decisions evolve with repeated exposure to automated vehicles (AVs) equipped with external human-machine interfaces (eHMIs). While prior research has largely relied on single-session experiments or video simulations, this longitudinal field study addresses the gap in understanding long-term effects of AV interaction. The primary research question examines how pedestrians’ feeling of safety changes as their experience interacting with AVs and eHMIs increases. The researchers conducted a mixed 3 × 2 × 2 quasi-experimental study in a controlled field environment using a Wizard-of-Oz approach. Twenty-one participants (aged 21–26) were randomly assigned to two groups based on eHMI type: Group A received an eHMI displaying only automation status (ADS-S), while Group B received an eHMI displaying status, perception, and yielding intention (ADS-SPI). Participants completed three sessions over five days. In each session, they faced scenarios involving a conventional vehicle (CV) and an AV (simulated by a hidden driver) that either yielded or did not yield. The dependent variable, "feeling of safety," was measured continuously using a handheld device; participants pressed a button when they felt safe to cross and released it when they did not. Results for nonyielding vehicles showed that perceived safety decreased linearly as the vehicle approached, with no significant effects from experience, vehicle type, or eHMI type. For yielding vehicles, participants initially felt safer crossing in front of the AV than the CV. As experience increased, perceived safety for the CV condition improved significantly, converging with the AV condition for Group A. However, for Group B (ADS-SPI), perceived safety at close distances (<10 m) increased with experience, reaching 100% in the final session. Statistical analysis revealed a significant interaction between experience and vehicle type, but no significant main effect for eHMI type overall. The study concludes that while eHMIs generally increase perceived safety, their specific impact depends on the information displayed and the pedestrian’s experience. Dynamic eHMIs that communicate perception and intention (ADS-SPI) are particularly effective at increasing perceived safety at close distances, where interactions are most ambiguous. This suggests that for successful AV integration, eHMIs should provide explicit intent information to support pedestrians’ developing mental models and trust over time.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• vru facing ehmi
• ehmi external hmi
• pedestrian behavior perception
• driver vru interaction
• acceptance adoption