Risk assessment and observation of driver with pedestrian using instantaneous heart rate and HRV
DOI: 10.54941/ahfe1003827
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 the relationship between driver risk assessment and physiological responses during interactions with pedestrians, aiming to improve driver models for autonomous vehicle development. While human drivers often outperform self-driving systems in complex collision avoidance scenarios, understanding their behavior requires accurate risk assessment metrics. Previous research has relied on subjective questionnaires or time-domain heart rate (HR) analysis, which suffers from individual variability and potential delays in physiological response. This paper addresses the gap in considering pedestrian dynamics by evaluating both time-domain HR and frequency-domain heart rate variability (HRV) to determine which method better correlates with driver behavior in safe versus unsafe traffic situations. The experimental design utilized a driving simulator built with Unreal Engine, featuring a two-lane road and a crosswalk. Nine participants, recruited based on valid licensage and health criteria, completed three scenarios each from a set of nine defined conditions. These scenarios varied pedestrian dynamics to create safe, unsafe, and dynamically changing risk environments, such as pedestrians crossing, stopping suddenly, or remaining stationary in the vehicle’s path. Participants were instructed to maintain 40 mph and brake to avoid collisions. Instantaneous HR and HRV data were collected using Polar H10 sensors. Data processing involved time-domain analysis of HR changes and frequency-domain analysis of HRV using Fast Fourier Transform (FFT) after spline interpolation, focusing on Very Low Frequency (VLF), Low Frequency (LF), and High Frequency (HF) bands to assess psychological stress and risk perception. Results indicated distinct differences between the two analytical methods. In the time domain, HR increases correlated directly with braking actions, showing that drivers’ physiological responses matched their vehicle dynamics in both safe and unsafe scenarios. For instance, participants who perceived risk and braked exhibited increased HR, while those who did not brake maintained steady HR. Conversely, frequency-domain HRV analysis revealed peaks in the HF region for both safe and unsafe scenarios, indicating perceived risk regardless of the actual danger level. This created a disconnect between HRV metrics and vehicle dynamics; for example, some participants showed high HF power (indicating stress/risk perception) in safe scenarios but did not brake. The authors attribute this discrepancy to individual risk thresholds, where drivers may perceive risk but choose not to act if the situation is deemed safe. The study concludes that time-domain HR analysis is a more reasonable and direct measure for assessing driver risk and predicting vehicle responses in pedestrian interactions. While frequency-domain HRV provides insight into the driver’s internal psychological state and risk perception, it cannot solely predict driving behavior due to individual risk thresholds. These findings suggest that future driver models for autonomous systems should incorporate expert driver data and account for individual risk parameters to enhance safety and comfort, moving beyond simple physiological proxies to more nuanced behavioral predictions.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-10 |
| archive | success | canonical_url | — | — | 1 | 2026-06-25 |
| extract | success | cached | — | — | 2 | 2026-06-25 |
| clean | success | clean | — | — | 1 | 2026-06-11 |
| chunk | success | chunk | — | — | 1 | 2026-06-11 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-11 |
| promote | success | — | — | — | 1 | 2026-06-10 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-25 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-11 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
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).
- Empirical Findings: physiological data
- Methodological Resource: tool software
- Theoretical Contribution: computational model