No abstract

Several studies have developed operating speed prediction models. Most of the models are based on spot speed data, collected by radar guns, pavement sensors, and similar mechanisms. Unfortunately, these data collection methods force the users to assume some invalid assumptions in driver behavior modeling: constant operating speed throughout horizontal curves and occurrence of acceleration and deceleration only on tangents. In this study, an instrumented vehicle with GPS continuous speed tracking was used to analyze driver behavior in terms of speed choice and deceleration or acceleration performance and to develop operating speed prediction models. The data used in the study were from a field experiment conducted in Italy on the rural motorway A16 (Naples–Avellino). Models were developed to predict operating speed in curves and tangents, deceleration and acceleration rates to be used in the operating speed profiles, starting and ending points of constant operating speed in a curve, 85th percentile of the deceleration and acceleration rates of individual drivers, and 85th percentile of the individual drivers’ maximum speed reduction in the tangent-to-curve transition. The study results showed that (a) the drivers’ speed was not constant along curves, (b) the individual drivers’ maximum speed reduction was greater than the operating speed difference in the tangent-to-curve transition, and (c) the deceleration and acceleration rates experienced by individual drivers were greater than the deceleration and acceleration rates used to draw operating speed profiles.

No abstract

Because speeding is one of the most significant contributing factors to fatal crashes, most road agencies attempt to achieve the right operating speed by imposing speed limits. Speed limit violations are prevalent, even on motorways with speed cameras. A problem with speed camera enforcement is that some motorists brake before passing a camera location and then exceed the speed limit after passing. This sudden braking can cause dangerous situations, crashes, and traffic jams. Furthermore, safe operating speed is not achieved where there are no cameras, especially where enforcement is overt, as in Italy. A new technique to overcome these problems is an automated section speed enforcement system, Safety Tutor. Unlike conventional speed meters, which measure vehicle speed at one point, the new technique determines average speed over a long distance. This study evaluated the safety effectiveness of the Safety Tutor system installed on Italian Motorway A1 Milan–Naples in 2007. An empirical Bayes observational before-and-after study was performed. The estimate of the total crash reduction is 31.2%, with a lower 95% confidence limit of 24.3%. The greatest crash reductions were observed for severe crashes and crashes at curves. Reduction was 55.6% for severe crashes, 26.6% for nonsevere crashes, 43.4% at curves, and 28.4% at tangents. However, the system's effectiveness decreased over time. The crash reduction was 39.4% in the first semester after the system's activation and 18.7% in the fifth semester after activation. Results strongly support activation of the new automated system owing to highly significant and substantial safety effects. The decrease in system effectiveness over time may be correctible with higher levels of enforcement.

Crash data collection is crucial for road safety improvement, but Italy is considerably behind the best international practices. To help to bridge this gap, a critical review of international crash databases was carried out and recommendations for improvement of the Italian police crash data collection and the national crash database were formulated. Main issues identified in the research are related to the database access procedures, the crash report, the crash location, the crash classification, and the severity classification