Trust-based route planning for automated vehicles
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 human trust considerations in automated vehicle route planning. While existing methods optimize for distance, time, or fuel, and some incorporate user profiles, none explicitly account for the driver’s trust in automation. The authors argue that trust is critical because distrust leads to frequent, unnecessary takeovers, while overtrust can cause catastrophic failures. To bridge this gap, the study presents the first trust-based route planning approach, modeling the human-vehicle interaction as a Partially Observable Markov Decision Process (POMDP) where trust is treated as a hidden mental state variable. The methodology involves two primary data collection phases. First, an online user study with 100 participants on Amazon Mechanical Turk was conducted to model trust dynamics and takeover decisions. Participants viewed driving videos depicting three incident types (pedestrians, obstacles, oncoming trucks) and reported their trust levels on a 7-point Likert scale and their willingness to take over control. This data was used to build a linear Gaussian model for trust evolution and data-driven models for predicting takeover decisions. Second, the authors formulated the route planning problem as a POMDP, incorporating vehicle position, incident type, vehicle capability, and trust as state variables. The reward function was designed to maximize safety and user experience, assigning higher rewards for successful autonomous handling of difficult tasks and penalizing failures. The proposed approach was evaluated through human subject experiments using a driving simulator with 22 participants. Participants were randomly assigned to either a trust-based route group or a baseline trust-free route group. The trust-based route utilized the POMDP framework to select paths that accounted for the driver’s evolving trust and predicted takeover likelihoods, whereas the baseline route ignored trust dynamics. The results demonstrated that participants navigating the trust-based route achieved higher cumulative POMDP rewards. Furthermore, these participants reported more positive responses in post-driving surveys compared to those on the trust-free route, indicating improved user experience. The significance of this work lies in establishing a formal framework for integrating human cognitive states, specifically trust, into automated vehicle navigation systems. By demonstrating that accounting for trust dynamics leads to better safety metrics and user satisfaction, the study highlights the importance of personalized, trust-aware planning in the development of Level 2 and Level 3 automated vehicles. This approach offers a pathway to mitigate the risks associated with both under-trust and over-trust, potentially accelerating the adoption of automated driving technologies by aligning system behavior with human psychological states.
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 | 2026-06-09 |
| extract | success | cached | — | — | 2 | 2026-06-09 |
| 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 | 1 | 2026-06-09 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 1 | 2026-06-09 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; 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