Impact of Connected Automated Buses in a Mixed Fleet Scenario With Connected Automated Cars

Considering the significant increase in transportation (R. Abe, 2019) demand in the last few decades and its consequences on society and the environment, the development of a sustainable transport system is vital. The relevance of public transport networks is expanding as present-day environmental, financial and resource pressures bear upon cities, providing the impetus for new sustainable and economical modes of mobility. Promoting public transport and introducing connectivity and automation are crucial for effectuating a sustainable transport system. Buses can improve the network performance, but curbside bus stops (CBS) elevate the side friction which cause substantial delays to other vehicles, urging imprudent forced lane changes and affecting safety. The introduction of connectivity and automation in buses may improve performance along with safety, and this study attempts to investigate the repercussions of connected automated buses (CABs) in a mixed fleet with connected automated vehicles (CAVs), through a microsimulation modelling exercise. Results indicate that CABs can reduce network-level travel time by 7.9%, 6.4%, 2.1%, 1.3%, 1.8%, 1.2% and 2.27% for 0%, 10%, 30%, 50%, 70%, 90% and 100% CAV penetration, when compared to the network with Human-driven buses (HBs) with the same CAV penetration. Furthermore, CABs also improve the performance of link containing bus stops by reducing standstill times (timespan when buses have zero velocity on the bus stop link). Results exhibit that CABs can decrease standstill times by 12.5%, 13.3%, 12.6%, 15%, 12.9%, 16.7% and 19.9% for 0%, 10%, 30%, 50%, 70%, 90% and 100% CAV penetration, when compared to the network with HBs with identical CAV penetration. Additionally, the number of lane changes in these links were reduced by 28.2% at 100% CAV penetration, indicating synchronous driving and reduced forced lane changes. The crash rate indicates that critical events involving automated vehicle entities and Manual Vehicles (MV) can increase as the MVs cannot adapt promptly to automated driving behaviour. This is supported by the non- parametric distribution of post-encroachment time (PET), time- to-collision (TTC) and relative speed (DeltaS), indicating that the conflict/collision typology’s severity can increase at intermediate CAV penetrations, but plunge significantly at 100% penetration when compared to the current scenario. Lastly, the limitation of tools in flagging conflicts involving CAVs is discussed, and future scrutiny is proposed.