Driving Performance, Alignment Consistency, and Road Safety Real-World Experiment

Road infrastructure improvement is one of the three pillars in the European Union policy aimed at halving the number of fatalities in road accidents by 2010. Increased knowledge and experience have proved that highway consistency analysis is an effective tool in the evaluation of road performance in terms of safety. Recent studies have attempted to provide information relating to the geometric consistency of road alignment by using different approaches based on geometric relation design, differential between design and operating speed, driving performance with instrumented vehicles, and human workload evaluation. This paper presents the results of a naturalistic driving experiment carried out with the University of Catania's Driving Instrumented Vehicle Acquisition System-instrumented vehicle. Representative variables for describing alignment consistency were defined and collected by direct measurement during real-world driving tests. From the seven parameters taken into consideration, the maximum driving speed differential between two successive elements and between the average section speed and the minimum single element speed were chosen as driving performance indicators (DPIs), since they are not correlated and agree statistically with the accident history. On the basis of actual driving behavior by means of DPIs, design inconsistencies were identified along different test courses, and their agreement with safety evaluation criteria and relation design parameters was verified. Comparison analyses among DPIs, speed consistency variables, and relation design parameters have evidenced the effectiveness of the indirect measurement of design consistency and the possibility of defining threshold values for the identification of those elements characterized by acceptable (good), reasonable (fair), or intolerable (poor) alignment inconsistencies.