Dynamic Planning of Vertical Road Signs Maintenance: Statistical Evidence

Abstract

This article is the second of two papers on a topic gaining ground in traffic and transport engineering, namely Dynamic Planning of Vertical Road Signs Maintenance. In particular, this second paper describes the development of the forecasting model and then the analysis of data and the statistical evidence. In detail, starting from the data, a comparison was made between the values measured and those prescribed by the reference standard (UNI EN 12899-1: 2008) according to the category of retroreflective film. Analysing the data macroscopically, it emerged that 75% of the signs met the minimum retroreflective requirements, in compliance with 25% did not respect them. Subsequently, a suitable database was selected for the development of the forecast model, excluding from the sample the vertical signs still under guarantee and proceeding with the correlation between the age of the sign and the retroreflection value found. A statistical analysis of the observed data was then carried out. In detail, the following were carried out: the two-sample t-test, used to verify the statistical significance of the average age difference relating to signs that meet or do not meet the minimum retroreflective requirements; the discriminant analysis with cross-validation, used for the development of the forecasting model; the linear regression, used to evaluate the variation of the retroreflective coefficient as a function of the age of the sign. The summary of the results and the insights for further research developments are made in the conclusions.