Modeling Study of Variable Speed Limits in Italy Observed About a 4 Percent Reduction in Mean Travel Time with the System Active Across Multiple Lanes
Italian Simulation Study Evaluated the Impacts of Deploying Variable Speed Limits in a Real-World Setting
Padua-Mestre Expressway
Statistical and Clustering-Based Assessment of Variable Speed Limits Effects on Motorway Performance from Real-World Observations
Summary Information
Variable Speed Limit (VSL) systems are traffic management strategies designed to reduce congestion, improve safety, and smooth traffic flow. The success of these systems depends not only on robust control algorithms but also on driver compliance, which varies by whether limits are advisory or mandatory and by the level of enforcement. The study was set on a 14 km stretch of the multi-lane Padua-Mestre Expressway in Italy, a real-world corridor equipped with Variable Message Signs (VMS) that display speed limits and other warnings. The technology under examination was a non-mandatory VSL system triggered when upstream traffic densities exceeded critical thresholds. The control algorithm determined appropriate speed limits, applying them lane by lane via the VMS network. This study combined traditional statistical tests with clustering analysis techniques to evaluate the operational impact of VSL under real-world conditions.
METHODOLOGY
The study was conducted from January to December 2021 on a 14 km section of the Padua-Mestre Expressway in Italy, covering 246 days of VSL system activation across multiple locations. The evaluation focused on assessing the impact of this non-mandatory (advisory) VSL system on traffic performance, specifically examining effects on vehicle speeds, travel times, speed distributions, and flow conditions under both active and inactive VSL states, with special emphasis on recurring congestion scenarios.
Data collection was drawn from two main sources: VMS status data, which recorded timestamped speed limit messages displayed for each lane, and traffic count data from 13 video-equipped counting stations per carriageway, providing minute-by-minute traffic counts, arithmetic and harmonic mean speeds, headways, occupancy rates, and vehicle classifications.
Statistical tests were used to compare average speed, mean travel time, and mean travel distance between VSL-active and non-active days, and clustering analyses were used to classify speed and flow profiles and detect distinct traffic patterns. Additionally, speed variance analyses were conducted to evaluate the effect of VSL on intra-vehicle speed variance, a factor linked to crash risk.
FINDINGS
- The mean travel time per vehicle during periods of recurring congestion decreased by approximately 4 percent when the VSL system was active compared to when it was inactive.
- The mean distance traveled showed a significant reduction only for the left lane, where it decreased by about 1 percent (0.04 km) during VSL activation.
- Speed variability (standard deviation) decreased by about 12 to 20 percent in the most congested segments, signaling smoother and more uniform traffic flow during VSL activation.
- Compliance rates (percentage of drivers staying below the posted speed limit) improved by 2 to 8 percent across different lanes when the VSL was active, even though the system was advisory and overall compliance levels remained moderate.
