Identification of Human Driver Critical Behaviors and Related Reliability Evaluation in Real Time
DOI: 10.3850/978-981-18-5183-4_s06-09-254-cd
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the critical need for real-time evaluation of human driver reliability, motivated by the increasing proportion of human-related accidents in traffic and the evolving role of drivers in automated vehicles. While automation shifts drivers from active control to passive monitoring, human error remains a primary safety concern. Existing Human Reliability Analysis (HRA) methods are largely static and retrospective, failing to capture the dynamic, situated context of driving where performance changes on a second-timescale. The authors aim to bridge this gap by identifying critical driver behaviors and evaluating reliability dynamically using a modified fuzzy-based Cognitive Reliability and Error Analysis Method (CREAM). The methodology employs a modified fuzzy-based CREAM approach to calculate a Human Performance Reliability Score (HPRS) in real time. To individualize the evaluation, the authors use the FN-DBSCAN (fuzzy neighborhood density-based spatial clustering of application with noise) algorithm, optimized by a genetic algorithm, to automatically generate membership functions for Common Performance Conditions (CPCs). These CPCs include factors such as ego-vehicle speed, time to collision (TTC), longitudinal and lateral acceleration, traffic density, and visibility. Data was collected from a driving simulator (SCANeRT M studio) featuring a two-way highway scenario with interacting vehicles. A specific dataset from a driver with eight years of experience was analyzed, focusing on a 120-second interval to generate membership functions and calculate continuous HPRS values. The results demonstrate that the approach successfully identifies dynamic fluctuations in driver reliability. Analysis of a specific critical incident revealed that the driver misjudged the decreasing TTC, leading to a delayed reaction and subsequent hard braking. This behavior caused significant fluctuations in lateral acceleration, indicating driver stress. The HPRS values dropped during this period, correlating with the critical situation. Using the Stanton and Salmon taxonomy, the error was classified as a "misjudgment" related to cognitive and decision-making errors. The study confirms that the method can detect these errors in real time, distinguishing between reliable performance and critical deviations caused by situational misinterpretation. The significance of this work lies in its contribution to third-generation HRA methods that account for dynamic contexts. By providing a continuous, data-driven measure of human reliability, the approach enables real-time supervision of interacting human drivers, which is vital for advanced driver assistance systems (ADAS) and automated vehicle safety. The authors conclude that this method offers a measurement-based understanding of human error mechanisms in situated contexts. Future work includes applying alternative clustering approaches, such as genetic-based membership function parameter estimation, and comparing control mode levels with the Skill-Rule-Knowledge framework to refine critical situation estimation.
Key finding
The modified fuzzy-based CREAM approach successfully identifies critical driving behaviors, such as time-to-collision misjudgment, by dynamically calculating human performance reliability scores from real-time driving data.
Methodology
simulator
Sample size: 1
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 author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 7 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-07 |
| chunk | success | chunk | — | — | 1 | 2026-06-07 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-07 |
| enrich | failed | — | — | — | 5 | 2026-07-02 |
| 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.
- human error taxonomy
- pre crash contributing factors
- automation surprise
- crash reconstruction hf
- naturalistic crash near crash
- 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).
- Methodological Resource: validation psychometrics
- Theoretical Contribution: computational model, theory or model