Planning for Automated Vehicles with Human Trust
DOI: 10.1145/3561059
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 gap in automated vehicle route planning by incorporating human trust as a critical factor, arguing that existing methods focusing solely on distance, time, or energy fail to account for the psychological dynamics of human-automation interaction. The authors posit that trust influences driver takeover decisions and overall satisfaction, necessitating a planning approach that models trust as a partially observable state variable. To achieve this, the researchers formalize the human-vehicle interaction as a Partially Observable Markov Decision Process (POMDP). They developed data-driven models for trust dynamics and takeover decisions using data from an online user study with 100 participants on Amazon Mechanical Turk. Participants viewed driving videos depicting three incident types (pedestrian, obstacle, oncoming truck) and reported their trust levels and takeover intentions on a 7-point Likert scale. The authors modeled trust evolution as a linear Gaussian system and estimated parameters using Bayesian inference with Hamiltonian Monte Carlo sampling. They compared a "trust-based" model, where takeover decisions depend on dynamic trust levels, against a "trust-free" baseline where decisions rely only on incident type. The study evaluated the resulting optimal routes via human subject experiments with 22 participants in a driving simulator. Using the Approximate POMDP Planning Toolkit, the authors computed two distinct routes for a motivating example: a trust-based route (A-D-G-J-K) and a trust-free route (A-C-E-H-K). The trust-based route strategically ordered incidents to manage trust levels, whereas the trust-free route did not. Experimental results indicated that participants following the trust-based route reported more positive responses in post-driving surveys compared to those on the baseline route. Additionally, the trust-based takeover model demonstrated superior fit to the collected data (log-likelihood of -359.37) compared to the trust-free model (-446.83), confirming that accounting for trust improves prediction accuracy. The significance of this work lies in demonstrating that route planning can be optimized for human factors, specifically trust, to enhance user satisfaction and safety. The authors also analyzed trade-offs between multiple objectives, such as trust, distance, and energy consumption, via multi-objective optimization. The paper concludes by identifying open issues and implications for real-world deployment, suggesting that integrating cognitive trust models into automated vehicle systems can lead to more effective and acceptable automation experiences.
Key finding
Participants who followed a trust-based route in a driving simulator reported more positive responses in post-driving surveys compared to those who followed a trust-free baseline route.
Methodology
mixed_methods
Sample size: 122
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 | — | — | 11 | 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).
- Theoretical Contribution: computational model