A scientific approach searching for a protocol to design and improve safety of existing roads

Road safety is an issue that recently has deeply modified the design and the enhancement criteria of roads. However, neither in single Countries (with the main exception of the US) nor internationally, a unique operational, technical and regulatory protocol has been yet defined. The analysis carried out in this paper is an attempt to achieve an operational protocol valid in Italy and in other countries, where some procedures to improve road safety are not yet established. This paper critically analyze both the procedures of the Highway Safety Manual and the European Directive 2008/96/CE, implemented in Italy together with the Guidelines N. 35/2011. The methodologies show many similarities but also significant differences. In particular, the Guidelines are simply based on Observed Accidents that suffers, for short periods of observation, from the error due to regression to the mean and, for long periods, from the error due to the variations of the boundary conditions. The HSM, instead, focuses on the estimation of the Expected Average rate of accidents. The main difficulty of this method is represented by the development of the Safety Performance Functions and the calculation of Calibration Coefficients, related to the specific site; obviously variable according to the number of boundary conditions, including geography and driver behavior. The differences in choosing priority interventions are also analyzed: while the Guidelines use mainly the Benefit-Cost Analysis, the HSM provides further possibilities depending on the objective to be achieved. The work ends with an hypothesis of operational methodology, even showing an application.