Content area
Midsize cities frequently experience significant highway traffic congestion but lack the financial resources to implement extensive rail-based transit solutions. This quantitative, experimental study evaluated the impact of using an existing highway lane exclusively for autonomous vehicle use within the traffic network of Sioux Falls, South Dakota, to demonstrate an autonomous vehicle transit system. The research addressed two primary questions: first, the extent to which a dedicated autonomous vehicle lane would increase traffic flow; and second, how varying ratios of autonomous to human-driven vehicles would affect overall traffic congestion. Guided by traffic flow theory and utilizing car-following and macroscopic flow models, this study utilized traffic simulation software for detailed analysis. Real-world traffic data informed the throughput of the Sioux Falls interstate network simulations, measuring congestion indicators such as average speed and time loss at varying autonomous vehicle ratios. Results showed that dedicating a lane to autonomous vehicles negatively impacted traffic flow at lower autonomous vehicle ratios due to insufficient utilization. However, higher autonomous vehicle ratios significantly improved speed and reduced time loss, indicating decreased congestion and enhanced efficiency. The study concluded that a midsize-city could effectively reduce congestion through appropriately timed infrastructure adjustments, like dedicated autonomous vehicle lanes, provided autonomous vehicle adoption was sufficiently high. It highlighted the importance of simulation-based planning in anticipating impacts of new transportation technologies. Future research should explore transitional infrastructure strategies, reviewing additional traffic flow variables, and statistical analyses suitable for non-normal data distributions.