Segment level safety analysis using lane-changing behavior and driving volatility features from connected vehicle trajectories
DOI: 10.1038/s41598-025-31303-8
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 limitation of existing urban arterial safety analyses, which predominantly rely on macro-level infrastructure and traffic features while ignoring micro-level risky driving behaviors. The authors aim to develop a directional-level segment crash analysis method that integrates high-resolution Connected Vehicle (CV) trajectory data to capture specific driving dynamics, particularly lane-changing behavior and driving volatility. By shifting from traditional bi-directional segments to unidirectional units, the research seeks to quantify the heterogeneous impacts of these microscopic behaviors on distinct crash types, specifically rear-end (RE) and sideswipe (SW) crashes. The methodology employs a three-step framework to extract lane-changing behaviors from raw CV trajectories. First, spatial matching maps GPS coordinates onto road centerlines using a Road Cartesian Projection system. Second, a Constrained Gaussian Mixture Method (CGMM) identifies lane boundaries and assigns lane IDs by modeling lateral distance distributions, constrained by known lane widths and uniform variance assumptions. Third, lane changes are detected by monitoring lane ID transitions, applying strict lateral distance and time thresholds to mitigate GPS positioning drift. Additionally, the study extracts other micro-level features, including traffic volume, aggressive speeding, and eight categories of risky driving behaviors defined by combinations of hard acceleration/braking and lane-changing maneuvers. Statistical modeling was conducted using data from Hillsborough County. A Bivariate Hierarchical Negative Binomial model was utilized to jointly estimate the impacts of driving behavior features on RE and SW crash frequencies, accounting for unobserved heterogeneity between crash types. Furthermore, a Hierarchical Zero-Inflated Poisson model was employed to identify contributors to speeding crashes. The empirical results highlight several critical safety relationships: segments with a high proportion of free-flow trajectories experience fewer RE and SW crashes. Driving fluctuations in stop-and-go traffic are positively associated with RE crash frequency. Risky right lane changes coupled with hard accelerations are significantly linked to SW crashes, while aggressive speeding behavior is highly correlated with speeding crashes. The significance of this work lies in its demonstration that microscopic driving behaviors, particularly lane-changing dynamics and driving volatility, are critical determinants of segment safety that are often overlooked in traditional macroscopic analyses. By validating the use of CV data to extract these features and modeling their specific impacts on different crash mechanisms, the study provides a more granular understanding of crash causation. This approach offers a foundation for developing targeted safety countermeasures that address specific risky behaviors rather than relying solely on infrastructure modifications or aggregated traffic metrics.
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 | 2026-06-18 |
| archive | success | unpaywall | — | — | 1 | 2026-06-25 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-18 |
| chunk | success | chunk | — | — | 1 | 2026-06-18 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-18 |
| promote | success | — | — | — | 1 | 2026-06-18 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-18 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- lane changing
- incidence prevalence
- naturalistic crash near crash
- sex gender
- telematics crash prediction
- urban rural setting
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).
- Empirical Findings: crash risk outcomes, observational prevalence
- Methodological Resource: dataset resource