Accident Data Analysis: an Experimental Investigation for a Rural, Multi-lane, Median-separated, Italian Road

Accident Data Analysis: an Experimental Investigation for a Rural, Multi-lane, Median-separated, Italian Road


F G Pratico, DIMET, University of Reggio Calabria, IT; R Vaiana, DIPITER, University of Calabria, IT



The goal of this paper is confined to the survey and analysis of accident data, road alignment, traffic data and environmental factors in order to identify the relevant variables that make a strong contribution towards a better understanding of accident circumstances and safety measures and policies.
The first section of the work deals with problem modeling. Geometry, traffic, weather factors are considered.
The next section concerns the survey of some of the above-mentioned factors such as alignment and the collection of accident data in order to construct an electronic data base; for six years, the accident numbers relative to about 400Km were collected and examined; the infrastructure is a rural, multilane, median-separated, Italian road, running from North to South in South Italy.
Given that, in the later section, analyses, interpretations and validations are performed.
The collected Accident data are statistically examined and different relative accident numbers are obtained. Among the different relative accident numbers also the following are used: Accident Rate (AR, accidents for million of vehicles and kilometer), Accident Cost Rate (ACR, in euro for one hundred of vehicles and per kilometer), Accident Density (AD, in accidents per kilometer and per year), Accident Cost Density (ACD, in euro per kilometer and per year).
In analyzing the problem, Authors bear in mind and study the relation between relative accident numbers and safety criteria (often based on road alignment congruence) as a basic support for a more comprehensive approach.
Alignment data are examined in relation to different method. In particular, design consistency, operating speed consistency, driving dynamic consistency are regarded as key-criteria and the influence of the previous elements on the i-th is well thought-out.
A clustering is performed by identifying homogeneous groups of sections and black areas are extracted and re-examined.
Accident data are compared to alignment, traffic and environmental factors, by considering different criteria and subsets.
In the light of the above-mentioned, safety measures and policies are identified and analyzed.
The practical application of the work concerns the proposal of an earlier factorial plan of the relevant variables, for the studied cases, and the corresponding safety measures and policies as handy reference for road engineers in safety analyses.


Association for European Transport