Shared vs. dedicated lanes for automated vehicle deployment: A simulation-based assessment

Abstract

• In shared lanes (SL), CAVs reduce delay and emissions by 38% and 52%, respectively. • An optimum percentage of CAV dedicated lanes (DL) is calculated for each MPR level. • SLs perform better at low traffic volumes, while DLs perform better at high volumes. • Emissions increase slightly below 40% MPR but are quickly reduced at higher MPRs. • Delay and throughput improvements occur mostly at MPRs above 30%. Connected and automated vehicles (CAV) will change the transportation world at an increasingly substantial pace. To realize the full potentials of CAV technologies, we need to investigate the transition period during which CAVs will share the roads with manually driven vehicles (MDV). This study provides a detailed assessment of the impacts of CAVs on freeways in the near future, which could serve as a basis for transportation officials and authorities to compare the different approaches for CAV deployment. The analysis was performed by modeling a hypothetical freeway segment using microsimulation. Three measures of effectiveness were used to quantify the impacts of CAVs, namely, average delay, throughput, and greenhouse gas emissions. The results reveal that at the baseline scenario of shared lanes (SL) with medium traffic volume, replacing MDVs with CAVs reduces 28.4 seconds (38%) in average delay. Additionally, throughput increases by 30%, and over 6,900 gCO 2 equivalent emissions (52%) are saved. Following the baseline analysis, an optimum percentage of dedicated lanes (DL) is derived for each market penetration rate (MPR) during the transition period. Finally, a comparison between DLs and SLs is conducted across multiple traffic volume scenarios. The findings indicate that SLs generally perform better at light traffic volumes, while DLs surpass them in congested conditions. When CAVs are deployed, both delay and throughput follow steady improvement in performance that is more evident at higher MPRs, i.e., above 30%. On the other hand, emissions experience a significant reduction in efficiency at 0 to 40% MPR but quickly make up for these losses at higher MPRs.