Toward the development of a hybrid approach to speed estimation in urban and rural areas

Given the strong relationship between road accident and traffic speed, the evaluation and prediction of this latter have always been considered as a critical issue for road safety analysis and for the evaluation of road network safety improvements. Prediction models developed to date mainly focused on spot speed in a rural environment or on running speed in an urban one. Very few analyze the speed estimation in "transition" areas. The objective of this paper is to develop a generalized speed estimation model able to predict mean speed in urban, rural, and "transition" environment as a function of road layout characteristics. It is believed that the proposed estimation tool can be effectively employed by road engineers in the road safety design and retrofitting stage.

The basic idea of the paper is to shed some light on this issue by making use of a hybrid estimation approach able to combine the information gathered from both previously mentioned models within a generalized speed adaptation framework that reflects road user behavior. The calibration and validation of the generalized estimation model have been carried out following a collection of Floating Car Data (FCD) on several candidate sites.

Preliminary results seem to indicate that the methodology proposed may be effective in estimating the spot speed in two-lane rural and urban arterials.

FCD data can be useful to develop more efficient estimation models to better manage the safety of urban and rural roads.