Human online reliability estimation applied to real driving maneuvers
DOI: 10.1109/cogsima49017.2020.9216153
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 lack of methods for estimating human driver reliability in real-time within dynamic, situated driving contexts. While human factors like fatigue significantly impact traffic safety, existing Human Reliability Analysis (HRA) approaches, such as the Cognitive Reliability and Error Analysis Method (CREAM), are primarily designed for static industrial settings and retrospective analysis. The authors aim to adapt CREAM for online application in driving, enabling assisted driving systems to monitor driver reliability and provide warnings or take over control when performance drops below safe thresholds. To achieve this, the authors modified the CREAM framework by replacing its original Common Performance Conditions (CPCs) with nine new CPCs specific to driving dynamics. These include the number of surrounding vehicles, time to collision (TTC), ego-vehicle speed, longitudinal and lateral acceleration, traffic density, number of available lanes, actual lane position, and visibility conditions. Each CPC is assigned levels that either improve or reduce performance reliability. The authors introduced a Human Performance Reliability Score (HPRS), calculated by summing the weighted effects of these CPCs. The HPRS maps to four control modes from the original CREAM model: strategic (highest reliability), tactical, opportunistic, and scrambled (lowest reliability). This allows for continuous, time-based evaluation of driver reliability. The approach was validated using a professional driving simulator (SCANeRT M studio) with six participants. Participants drove through four highway scenarios varying in traffic density and lane configurations, including maneuvers like lane changes and exits. Data on vehicle dynamics and surrounding traffic were collected to calculate HPRS in real-time. Results demonstrated that HPRS fluctuates dynamically in response to changing driving conditions. For instance, during overtaking maneuvers, HPRS decreased when the number of available lanes reduced and surrounding vehicles increased, then recovered as the maneuver completed. The study found that drivers generally maintained tactical or strategic reliability levels. Scenario analysis revealed that scenarios 1 and 4 had a stronger negative impact on reliability, causing drivers to spend more time in the tactical mode compared to scenarios 2 and 3. The order of scenarios did not significantly affect reliability, likely due to rest periods between trials. The significance of this work lies in providing the first method for online estimation of human reliability in situated driving contexts. By quantifying reliability through HPRS, the approach enables advanced driver assistance systems to assess whether a driver’s behavior matches the traffic situation. This capability supports safety interventions, such as issuing warnings when reliability drops to opportunistic levels or initiating takeover in scrambled modes. The study confirms the applicability of the modified CREAM approach for real-time monitoring, offering a foundation for integrating human reliability metrics into automated driving systems to enhance traffic safety.
Key finding
The modified CREAM approach with a new Human Performance Reliability Score successfully estimated dynamic human driver reliability in real-time based on situational driving data.
Methodology
simulator
Sample size: 6
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 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| 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.
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, tool software
- Theoretical Contribution: computational model