Impact of an Autonomous Shuttle Service on Urban Road Capacity: Experiments by Microscopic Traffic Simulation

TL;DR

This study uses microscopic traffic simulation calibrated with real-world data to evaluate how autonomous shuttle services affect urban road capacity, revealing optimal shuttle speeds and frequency for improved traffic flow.

Contribution

It introduces a calibration method for autonomous shuttles in traffic simulation and assesses their impact on urban road capacity using real-world data.

Findings

Increasing shuttle frequency raises delays and reduces speeds.

Optimal shuttle speeds improve traffic conditions during peak and off-peak hours.

Simulation results inform policy and infrastructure planning for autonomous shuttles.

Abstract

Autonomous vehicles are expected to transform transportation systems with rapid technological advancement. Human mobility would become more accessible and safer with the emergence of driverless vehicles. To this end, autonomous shuttle services are currently introduced in different urban conditions throughout the world. As a result, studies are needed to assess the safety and mobility performance of such autonomous shuttle services. However, calibrating the movement of autonomous shuttles in a simulation environment has been a difficult task due to the absence of any real-world data. This study aims to calibrate autonomous shuttles in a microscopic traffic simulation model and consequently assess the impact of the shuttle service on urban road capacity through simulation experiments. For this analysis, a prototype of an operational shuttle system at Lake Nona, Orlando, Florida is emulated in a microscopic traffic simulator during different times of the day. The movements of autonomous vehicles are calibrated using real-world trajectory data which help replicate the driving behavior of the shuttle in the simulation. The analysis reveals that with increasing frequency of the shuttle service the delay time percentage of the shared road sections increases and traveling speed decreases. It is also found that increasing the speed of shuttles up to 5 mph during off-peak hours and 10 mph during peak hours will improve traffic conditions. The findings from this study will assist policymakers and transportation agencies to revise policies for deploying autonomous shuttles and for planning road infrastructures for shared road-use of autonomous shuttles and human driven vehicles.