Impact of Autonomous Vehicles on Control Delay & Safety: A Case Study of Signalized Tight Diamond Interchange at Executive Towers Business Bay, Dubai

Abstract

Autonomous Vehicles (AVs) promise to transform urban mobility by improving traffic flow and safety, but their actual impact under varied traffic and geometric conditions remains uncertain, warranting further study. This study evaluates the impacts of AVs on the operational and safety performance of a signalized tight diamond interchange at Executive Towers Business Bay, Dubai, under mixed traffic conditions. Three AV driving logics: aggressive, normal and cautious, were gradually introduced, replacing conventional cars while maintaining a constant mix of 2% heavy vehicles and 1% buses. A calibrated and validated traffic model was developed in PTV VISSIM using site-specific geometric and operational data, with maximum queue length used as the measure of effectiveness (MOE). Thirteen scenarios were simulated to evaluate varying AV penetration levels. Delay outputs were extracted from VISSIM, while vehicle trajectory files were analyzed in the Surrogate Safety Assessment Model (SSAM) using TTC thresholds of 1.5 and 1.0 seconds. Calibration yielded optimal values for VISSIM’s car-following parameters: average standstill distance (1.35 m), additive part of safety (0.75 m), and multiplicative part of safety (1.50 m). Results showed that at a demand level exceeding 5,000 veh/hr, AV-Aggressive at 100% penetration reduced average delay by 7.5% and total conflicts by 48.6% compared to conventional vehicles. In contrast, AV-Cautious increased delay by 90.6% and conflicts by 69.2%. AV-Normal caused a modest 3.5% increase in delay but reduced conflicts by 26.7%. Overall, Scenario 13, 100% AV-Aggressive—demonstrated the best operational and safety performance. These results highlight the critical role of AV driving logic in shaping interchange performance, with aggressive AV behavior at full penetration offering the most substantial improvements in delay reduction and conflict mitigation. This suggests that future AV integration strategies should consider behavior modeling as a key factor in optimizing traffic operations and safety in complex urban environments.