Processing of complex traffic scenes for effective steering and collision avoidance: a perspective, from research on locomotion
DOI: 10.3389/fpsyg.2024.1347309
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This perspective paper evaluates how behavioral research into human steering and collision avoidance can inform the development of sensor-based autonomous vehicles (AVs), particularly regarding their performance in complex urban environments. The author argues that while AVs excel in simple, grid-type road networks, they face significant challenges when navigating historic inner-city streets characterized by irregular geometry, narrow passages, and interactions with vulnerable road users. The central research question is whether AV control systems must more closely mimic human visual processing and develop a "theory of road users" to attribute intent to other drivers, cyclists, and pedestrians. The paper synthesizes existing behavioral research on human locomotion, focusing on the processing of optic flow, optical looming, and the role of the mobile gaze system. It contrasts human perceptuo-motor capabilities with current AV sensing technologies, such as fixed-axis cameras and LIDAR. The analysis highlights that human drivers utilize dynamic inputs, including gaze angle and egocentric visual direction, to navigate complex trajectories, whereas AVs rely on static geometric data and direct internal state monitoring. The author examines specific scenarios, such as detecting cyclists with heterogeneous shapes and interpreting subtle behavioral cues from pedestrians and other drivers at pinch-points where formal right-of-way rules are ambiguous. Key findings indicate that current AV systems are insufficient for complex urban scenes because they lack the ability to seamlessly integrate object recognition with collision detection and intent attribution. While AVs can effectively manage lane-keeping and collision avoidance in simple environments, they struggle with the "discord" between retinal flow and optic flow that humans resolve through mobile gaze. The paper identifies that humans use high-resolution foveal vision to rapidly resolve ambiguities and detect subtle signals, such as a cyclist’s rearward glance or a pedestrian’s body language, which static machine vision systems cannot easily replicate. Furthermore, AVs often fail to interpret unwritten social conventions, such as yielding at narrow streets, leading to potential conflicts with human drivers who expect assertive or cooperative behavior. The significance of this work lies in its proposal for a "theory of road users" for AVs. The author suggests that to safely negotiate complex environments, AVs must move beyond simple hazard detection to attribute mental states and intentions to other road users. Guidelines are provided for developing this theory, emphasizing the need to recognize signaling features such as gaze patterns, body language, and approach kinematics. The paper concludes that without mimicking the cognitive and perceptual flexibility of human drivers, AVs risk being unduly delayed or causing dangerous situations in dense urban settings, highlighting a critical gap in current autonomous driving research.
Key finding
Autonomous vehicles require a 'theory of road users' that mimics human visual processing and intent attribution to safely navigate complex urban environments, as current systems struggle with the social and perceptual challenges of interacting with pedestrians and cyclists.
Methodology
review
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 scout_discovery on 2026-05-08.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | partial | scout | — | — | 2 | 2026-05-08 |
| archive | success | unpaywall | — | — | 1 | 2026-06-04 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | semantic_scholar | — | — | 2 | 2026-06-04 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- situational awareness
- anticipation
- ehmi external hmi
- mental model of traffic
- hazard perception
- driver vru interaction
- perception action locomotion
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).
- Theoretical Contribution: computational model, theory or model, conceptual framework