Disaggregate Road Accident Analysis for Safety Policy and Measures: Theoretical Aspects and Application

Disaggregate Road Accident Analysis for Safety Policy and Measures: Theoretical Aspects and Application


A Vitetta, F Russo, University Mediterranea of Reggio Calabria, IT



In this paper a method is proposed for analysing road incidents and defining safety policies and safety measures. The method uses the scenario incident approach proposed by Brenac and Megherby (1996) and Brenac et al. (1996) where some scenarios were defined with statistical approach. This approach was extended inside a theory and it is applied to calibrate the scenarios on an Italian extra-urban road considering real data.

The importance of the proposed paper consists in the proposal of a global method for specifying and calibrating the incident scenario and in scenario calibration with a quantitative approach.

The proposed method can be applied in the place where the frequency of incidents is greater than the corresponding value in other similar areas. This area will be called the study area. In this place a specific approach with in-depth investigations has to be applied for: incident detection; incident analysis; safety measures application; system monitoring.

The first step begins with incident detection measurement 'in time' and 'on the scene' in the study area. The 'in time' measurements are the microscopic traffic flow (counted or simulated) and ambient characteristics during the impact; the 'on the scene' measurements concern infrastructural characteristics, users and vehicles. The data have to be collected by an expert team and can be: macro or micro. The macro data are relative to socio-economic characteristics, supply and demand information, while the micro data are relative to vehicles, users and infrastructural characteristics useful for incident reconstruction.

The data relative to the incidents of the study area have to be analysed to identify the common events, incident factors relative to the users (human factors), infrastructure and vehicles and their reciprocal interactions. For the analysis two different methodological analyses can be used: macroscopic or microscopic. Macroscopic analysis concerns a large area (urban area, central business district, etc.) and estimates the probability of an incident occurring in relation to a set of variables that are generally macro. Microscopic analysis concerns an infrastructural element (road, junction, parking area, etc.) and defines common elements in the incidents in order to define safety measures for avoiding impacts in relation to a set of variables that are generally micro. With microscopic analysis the incident scenario can be generated.

After data analysis safety infrastructural or supply management measures have to be applied on the system. The safety measures are strictly connected with the common factors extracted from data analysis and in particular with incident scenario and may be of various types: user education, engineering measures, traffic organization, traffic control, enforcement. The last step of the method consists in monitoring the effect generated with the safety measure applied on the system in order to verify pre-defined targets.

The proposed method is applied on a freeway in the extra-urban area of Naples (Italy). A set of 63 road incidents was detected with an expert team in a study area and for each incident more than 500 microscopic measurements were made on the scene. The data were analysed with the microscopic approach and the proposed methodology is applied for calibrating incident scenarios for an extra-urban road in Italy. For each scenario the safety measures are defined and are communicated to infrastructural managers for application on the system.

h4. References

Brenac T., Megherby B. (1996), ?Diagnostic de sécurité routière sur une ville: intéret de l?analyse fine de procédures d?accidents tirées aléatoriement?, Researche Transports Sécurité, 52, 59-71.


Association for European Transport