Automated Driving Systems’ Communication of Intent with Shared Road Users
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This report addresses the challenge of communication between Automated Driving Systems (ADS) and shared road users, specifically drivers, pedestrians, and bicyclists. The introduction of high-level automation removes the human driver from the dynamic driving task, potentially disrupting informal cues—such as eye contact, gestures, and vehicle movement—that currently facilitate safe interactions. The research aimed to identify the specific cues road users rely on to predict intent and to develop preliminary laboratory testing procedures for assessing external human-machine interfaces (eHMI) designed to replace these lost signals. The study comprised three distinct phases. Study 1 involved structured interviews with ten driving evaluation experts to characterize communication cues in nine roadway scenarios. Experts categorized cues by source (driver, vehicle, other users), type (explicit vs. implicit), and modality (visual vs. audible), while also rating their rarity and detectability in daylight and darkness. Study 2 utilized field observations with 40 participants (12 drivers, 16 pedestrians, 12 bicyclists) who performed think-aloud commentary while navigating public roadways. Participants verbally reported the cues they used to anticipate vehicle actions, which were coded into driver-related, vehicle-movement, and vehicle-signaling categories. Study 3 employed a laboratory method to test six prototype eHMI designs (including an ADS beacon, light bars, simulated eyes, yielding text, and pedestrian symbols) using video stimuli of a vehicle approaching an unsignalized crosswalk. Participants rated their confidence in the vehicle’s intent under naïve and informed conditions. Results from Study 1 indicated that driver behavioral cues were most prominent in low-speed scenarios like parking lots and four-way stops, while vehicle-movement cues dominated in congested pedestrian areas. Experts noted that visual cues were significantly harder to detect in darkness. Study 2 found that across all user groups and scenarios, participants relied most heavily on vehicle-movement cues. However, driver-related cues, such as eye contact, were frequently used by pedestrians and bicyclists in close-proximity situations with immediate safety implications. Study 3 demonstrated that participants generally recognized eHMIs as indicators of automated driving. Even without prior explanation, participants perceived that the vehicle had recognized their presence if the interface signaled yielding. The laboratory method showed promise in differentiating between eHMI designs for simple crosswalk scenarios, though the authors caution against generalizing these findings to complex environments. The significance of this work lies in establishing a baseline for understanding the information needs of shared road users in an automated future. By identifying which cues are most critical for safety and developing a testing protocol for eHMIs, the report provides a foundation for designing communication systems that mitigate the risks associated with the loss of human-to-human interaction. The findings suggest that while vehicle movement remains a primary cue, explicit signaling systems are necessary to convey awareness and intent, particularly in vulnerable-user scenarios.
Key finding
Participants recognized automated vehicle intent and perceived yielding behaviors as acknowledging their presence, while field observations revealed that shared road users rely most on vehicle-movement cues to predict driver intent.
Methodology
mixed_methods
Sample size: 40
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. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 2026-06-11 |
| verify | partial | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified_with_issues.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- ehmi external hmi
- signaling behavior
- vru facing ehmi
- driver vru interaction
- automation surprise
- situational awareness
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Applied Guidance: design guidelines
- Methodological Resource: tool software
- Theoretical Contribution: conceptual framework