Application of AV Dedicated Lanes to Dynamic Traffic Flow Analysis in Mixed Human-Autonomous Vehicles

This study investigates the safety and operational implications of integrating Autonomous Vehicles (AVs) into existing infrastructure, specifically focusing on the transitional phase where AVs coexist with Human-Driven Vehicles (HDVs). Using the Jakarta Harbour Tollroad as a case study, a VISSIM 2022 microsimulation model was calibrated to analyze three lane configurations—Mixed Traffic, 1 Dedicated Lane (1 DL), and 2 Dedicated Lanes (2 DL) across Market Penetration Rates (MPR) ranging from 10% to 90%. The simulation results reveal that while Mixed Traffic environments exhibit dangerous instability and negligible efficiency gains at low adoption rates, the implementation of dedicated lanes significantly alters performance outcomes. The study identifies the 1 Dedicated Lane configuration as the optimal strategy for the transition phase (MPR 10%-60%), where it functions as a "sweet spot" that simultaneously reduces fatal rear-end conflicts by over 1,600 occurrences and improves Space Mean Speed by nearly 13% compared to the baseline. Conversely, the 2 Dedicated Lanes strategy proves detrimental at low penetration rates, causing severe "capacity waste" and congestion, becoming viable only when AV adoption reaches a maturity threshold of 70%. Consequently, this research proposes a dynamic implementation roadmap, recommending a phased infrastructure policy that prioritizes a single dedicated lane to ensure safety during the early transition before expanding capacity as the market matures