Estimating Driver Personality Traits From On-Road Driving Data
DOI: 10.1109/ACCESS.2023.3308819
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 addresses the challenge of developing personalized driving assistance systems by estimating drivers’ psychological characteristics from on-road driving data. The motivation stems from the high rate of traffic accidents involving older drivers, often linked to cognitive decline, and the inadequacy of "one-size-fits-all" assistance systems that fail to account for individual differences. The research aims to determine if machine learning and deep learning techniques can accurately estimate cognitive function, psychological driving style, and workload sensitivity using real-world driving behavioral data, thereby enabling adaptive feedback without burdensome manual testing. The methodology utilizes a dataset from 23 elderly drivers (aged 50–79) who completed driving sessions on public roads. Ground truth labels were established using neuropsychological tests (Trail Making Test, Maze test, Useful Field of View test) and self-reported questionnaires (Driving Style Questionnaire, Workload Sensitivity Questionnaire). The authors processed time-series data from nine in-vehicle sensors (e.g., steering angle, acceleration, brake pressure) and GPS data. A key methodological innovation involves segmenting driving data by road type (arterial roads vs. intersections) and further dividing these segments into various durations (e.g., 3s to 60s for arterial roads; first/second halves for intersections) to capture diverse behavioral contexts. Statistical features were extracted from these segments and fed into regression models, including Random Forest and Long Short-Term Memory (LSTM) networks, to predict the psychological traits. The results demonstrate that the proposed models can estimate cognitive function with moderate to high accuracy. Specifically, the model achieved Pearson correlation coefficients of 0.579 for TMT (B) scores and 0.708 for UFOV test scores. The study also found that psychological driving style and workload sensitivity could be estimated with high accuracy. However, the effectiveness of segmenting data by various durations was inconsistent; it improved accuracy for some characteristics but not for others. Additionally, the analysis identified specific sensor inputs and road types that were most critical for estimating cognitive function, highlighting that intersection data and specific sensor metrics provided distinct predictive value compared to arterial road data. The significance of this work lies in its validation that driver psychological traits, particularly cognitive function, can be inferred from routine driving behavior using machine learning. This finding supports the feasibility of creating adaptive driving assistance systems that tailor alerts and support to individual driver capabilities. By identifying which sensors and road contexts are most informative, the study provides a roadmap for efficient feature extraction in future autonomous or semi-autonomous vehicle systems, potentially enhancing safety for older drivers by mitigating risks associated with cognitive decline.
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 | DOAJ | — | — | 1 | 2026-06-16 |
| archive | success | unpaywall | — | — | 1 | 2026-06-25 |
| extract | success | cached | — | — | 2 | 2026-06-25 |
| clean | success | clean | — | — | 1 | 2026-06-16 |
| chunk | success | chunk | — | — | 1 | 2026-06-16 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-16 |
| promote | success | — | — | — | 1 | 2026-06-16 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-25 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-16 |
| 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.
- cognitive capacity variation
- personality driving
- telematics crash prediction
- mental demand
- situational awareness
- exposure measurement
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