Although road safety has improved in Italy over the last decade, the EU objective of halving the number of road deaths by 2020 has not been achieved yet. For pedestrians and cyclists, i.e., the most Vulnerable Road Users (VRUs), such trend is even worse: pedestrians’ road crashes have plateaued around 0%, and the one of cyclists has increased by 10%, compared to 2010. Therefore, VRUs road safety requires further commitment. Clustering techniques have been used in road safety analysis to investigate crash patterns. Latent Class Clustering (LCC) and K-Means clustering were mostly applied and specifically as a pre-processing stage for further statistical analysis. This study proposes a framework to detect crash clusters and related patterns and associate with them specific solutions from a predefined set of solutions, mainly based on infrastructural and environmental attributes. Specifically, starting from raw official statistical crash data, the framework adopts hierarchical clustering to structure pedestrians’ and cyclists’ road crashes in urban areas. The framework is applied to the Province of Brescia (Northern Italy) using crash data collected for the five-year period 2014-2018. Three clusters for pedestrians’ and five clusters for cyclists’ road crashes are identified and specific solutions to mitigate crash occurrences are suggested, accordingly. The provided insights and solutions may be useful as a decision-support tool for public administrators in improving road safety.

Identifying clusters and patterns of road crash involving pedestrians and cyclists. A case study on the Province of Brescia (IT)

Mor A.;
2022

Abstract

Although road safety has improved in Italy over the last decade, the EU objective of halving the number of road deaths by 2020 has not been achieved yet. For pedestrians and cyclists, i.e., the most Vulnerable Road Users (VRUs), such trend is even worse: pedestrians’ road crashes have plateaued around 0%, and the one of cyclists has increased by 10%, compared to 2010. Therefore, VRUs road safety requires further commitment. Clustering techniques have been used in road safety analysis to investigate crash patterns. Latent Class Clustering (LCC) and K-Means clustering were mostly applied and specifically as a pre-processing stage for further statistical analysis. This study proposes a framework to detect crash clusters and related patterns and associate with them specific solutions from a predefined set of solutions, mainly based on infrastructural and environmental attributes. Specifically, starting from raw official statistical crash data, the framework adopts hierarchical clustering to structure pedestrians’ and cyclists’ road crashes in urban areas. The framework is applied to the Province of Brescia (Northern Italy) using crash data collected for the five-year period 2014-2018. Three clusters for pedestrians’ and five clusters for cyclists’ road crashes are identified and specific solutions to mitigate crash occurrences are suggested, accordingly. The provided insights and solutions may be useful as a decision-support tool for public administrators in improving road safety.
Transportation Research Procedia
Vulnerable Road Users
Cluster analysis
Crash patterns
Decision-support tool
Solutions definition
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1203486
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact