Improved driver modeling for human-in-the-loop vehicular control
DOI: 10.1109/icra.2015.7139410
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 challenge of developing provably safe human-in-the-loop vehicular control systems by creating accurate, individualized models of driver behavior. The authors argue that traditional safety methods, such as calculating vehicle reachable sets, are insufficient for high-speed driving because they treat the human driver as a disturbance rather than an active agent whose actions depend on mental state and context. To mitigate the high rate of human-error-related accidents, the study proposes a data-driven model that predicts likely driver trajectories based on discrete mental states (attentive, partially attentive, distracted) and environmental context. The methodology employs a human-in-the-loop testbed consisting of a Force Dynamics CR401 motion platform simulator integrated with PreScan software. Thirteen subjects performed driving tasks in highway and intersection scenarios while their mental states were monitored via a distraction protocol involving text messaging. The model uses a hierarchical algorithm to parse data into distinct driver modes based on mental state, road context, and surrounding obstacles identified through k-means clustering. This approach estimates an empirical probability distribution of future trajectories, allowing the system to predict the set of states the driver is likely to visit over time horizons ranging from 0.5 to 5 seconds. Results demonstrate that the proposed model significantly outperforms standard reachable set methods in precision while maintaining high accuracy. In highway scenarios, the model achieved accuracy rates above 90% for time horizons up to 1.5 seconds, with precision improving as the number of environmental clusters increased. For intersection scenarios, the model maintained high accuracy (approximately 97% at 1 second) and high precision (up to 96% at 1 second) over longer horizons of up to 5 seconds. The study highlights a trade-off where increasing the granularity of environmental clusters improves prediction precision but slightly reduces accuracy due to smaller dataset subsets. The significance of this work lies in its application to semi-autonomous control frameworks, specifically switched and augmented control systems. By providing a probabilistic assessment of driver safety, the model enables intervention algorithms to act only when the likelihood of collision exceeds a defined threshold, thereby minimizing unnecessary system intrusiveness. The authors conclude that this context-aware, individualized modeling approach allows for robust safety guarantees in human-machine interaction, paving the way for safer autonomous vehicle integration and providing a foundation for future research into more complex driving conditions.
Key finding
The proposed context-aware driver model significantly improves prediction precision over standard reachable set methods while maintaining high accuracy, enabling more precise and less invasive semi-autonomous vehicle control.
Methodology
simulator
Sample size: 13
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: tool software
- Theoretical Contribution: computational model, theory or model