Construction Work Zone Safety: Spatio-Temporal Analysis of Construction Work Zone Crashes
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This study addresses the critical safety risks associated with construction work zones, which account for significant fatalities and injuries annually in the United States. Motivated by the limitations of traditional statistical models in capturing the complex, dynamic nature of work zone crashes, the research aims to enhance the prediction of crash severity using advanced machine learning techniques. The study specifically investigates factors influencing crash severity, including driver behavior, work zone characteristics, contract types, and the effectiveness of various safety countermeasures. It also seeks to evaluate the state-of-the-practice among Department of Transportation (DOT) agencies regarding work zone safety management. The methodology involved a comprehensive analysis of crash and work zone data from Utah, alongside a broad literature review and a national survey of DOTs. The researchers cleaned and preprocessed the data, addressing missing values, outliers, and feature encoding. They developed a machine learning framework comprising both deterministic models (Decision Trees, Random Forest, Support Vector Machines, Neural Networks) and probabilistic models to predict crash severity. Feature importance analysis was conducted to identify key predictors. Additionally, the study analyzed the impact of specific variables, such as the presence of longitudinal rumble strips, different contract types (e.g., CMGC vs. design-build), and traffic countermeasures. A survey distributed to all DOTs yielded 24 responses from 22 states, providing insights into current practices and satisfaction levels with various safety interventions. Key findings indicate that machine learning models effectively capture the complex interplay of factors affecting crash severity. The analysis revealed that Construction Manager/General Contractor (CMGC) contracts are associated with a notably higher frequency of crashes and a higher crash rate per 100 million vehicle miles traveled compared to design-build or design-bid-build contracts. Regarding countermeasures, longitudinal rumble strips, while effective for general roadway departure crashes, showed less significant impact within work zones, suggesting a need for further investigation into their application in these areas. The survey highlighted varying levels of satisfaction with different countermeasures across states. Feature importance analysis identified specific work zone attributes, such as weather conditions, road geometry, and traffic characteristics, as critical determinants of crash severity. The Random Forest model demonstrated strong performance in predicting severity outcomes. The significance of this research lies in its provision of evidence-based recommendations for improving work zone safety. By identifying high-risk contract types and evaluating the efficacy of specific countermeasures, the study offers actionable insights for transportation agencies. The findings underscore the importance of tailoring safety strategies to specific work zone configurations and contract specifications. The integration of machine learning approaches provides a robust tool for predicting crash severity, potentially aiding in the development of targeted interventions. The study concludes by suggesting future research directions, including the incorporation of real-time data and further exploration of emerging technologies like smart work zones to enhance prediction accuracy and overall road safety.
Key finding
Construction Manager/General Contractor (CMGC) contracts exhibit a notably higher crash rate per 100 million vehicle miles traveled compared to design-build or design-bid-build contracts, and longitudinal rumble strips appear less effective at reducing crashes within work zones than in general roadway contexts.
Methodology
mixed_methods
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 bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 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.
- work zones
- incidence prevalence
- naturalistic crash near crash
- emergency work zone conspicuity
- causation analyses
- regulatory evaluation
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).
- Applied Guidance: countermeasure evaluation
- Empirical Findings: crash risk outcomes
- Methodological Resource: dataset resource