An international review of challenges and opportunities in development and use of crash prediction models
DOI: 10.1186/s12544-018-0307-7
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This review paper addresses the significant gap between the state-of-the-art in crash prediction models (CPMs) and their state-of-the-practice application by road safety practitioners. While CPMs have become fundamental scientific tools for road safety management over the last decade, their practical uptake remains limited. The authors aim to bridge this divide by reviewing international experiences in developing and applying CPMs, specifically focusing on crash frequency estimation for road segments and intersections. The objective is to enhance practitioner understanding of why and how to utilize CPMs to improve road safety outcomes. The methodology involved a comprehensive review of both scientific literature and practice-oriented reports from road agencies and institutes, limited to English-language sources. The review synthesized information across eight hierarchical steps of CPM development and application: data collection, network segmentation, variable selection, model function specification, validation, network screening, development of crash modification factors (CMFs), and the use of CPM tools. The analysis focused on typical road settings, excluding macro-level planning or specific vulnerable user models, to address the most frequent practical applications such as network screening and safety impact assessment. The findings highlight that developing CPMs is complex due to numerous analytical choices lacking definite guidance, leading to diverse modeling approaches. The review details specific challenges, such as the need for adequate sample sizes (often 30–50 locations with at least 100 crashes per year) and the handling of data biases like underreporting and spatial correlation. It notes that while negative binomial models are standard for handling overdispersion, practitioners often struggle with variable selection and model validation. Furthermore, the paper identifies that while empirical Bayes methods improve network screening by reducing random statistical variation, the transferability of models across jurisdictions is limited, necessitating local model development. The review also critiques the derivation of CMFs from cross-sectional studies due to potential bias-by-selection, though acknowledges their common use when before-after data is scarce. The significance of this work lies in its identification of the need for CPM solutions that are both scientifically sound and practically feasible. The authors conclude that overcoming the identified challenges requires increased communication between researchers and practitioners. By addressing issues such as model validation, appropriate segmentation, and the integration of observed data with predicted frequencies, the field can move toward greater adoption of CPMs. This shift would enable more rational, data-driven road safety management, allowing agencies to effectively screen networks, assess treatment impacts, and prioritize safety improvements.
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 | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | canonical_url | — | — | 1 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| enrich | success | openalex | — | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| 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.
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