Interaction of public autonomous vehicles with stochastic driving modes on Indian roads

Abstract

With advent of new technologies, the transportation sector is experiencing a strong influence towards an era of driverless vehicles. The influx of autonomous vehicles (AVs) can have an impact on mobility, and traffic dynamics will differ hugely from traditional human-driven vehicles in terms of conservative driving, increased intervals, low average travel speeds, etc. because of the pre-set algorithms on which the AVs work. Because of the substantial research done on AVs from both academia and industry over the last decade, AVs have shown incredible development. Despite the excitement around autonomous vehicles, the implementation of AVs as public transport is an emerging field of interest as it’s expected to improve network efficiency, safety and alleviate network congestion. Diverse scenarios are being studied and looked into in several European projects but seldom do they talk about a heterogeneous mix of AVs and other stochastic driving modes. Thus, there arises a need to study the transition phase especially in the Indian scenario. This paper tries to study the changes to the existing network by introducing Autonomous Buses (ABs) at different penetration rates on a realworld Indian urban road network using a microsimulation platform. ABs have significant potential on improving efficiency not just of the bus network and operations but also for the whole network but only when they are implemented at higher penetration rates. The study finds that the network becomes performs better only after ABs are introduced at about 70%. Further, when compared with a bus only lane, the network performs better even without the penetration of ABs, thus providing useful insights regarding the feasibility of ABs. Further, the study provides suggestions for improving the future microsimulation platforms at each stage of the research.