Operational Analysis of a Single-Lane Roundabout with a Mix of Driverless Vehicles

Abstract

Driverless cars (DLCs) are vehicles capable of managing several driving tasks by themselves under different degrees of human intervention. DLCs are able to maintain shorter headways in entering a roundabout and circulating in the roundabout. The Highway Capacity Manual (HCM) provides a regression-based gap acceptance model for estimating the capacity of roundabouts. The gap acceptance model captures driver behavior, which is characterized by headways. As automotive manufacturers and technology companies develop driverless capabilities for roadway vehicles, DLCs at various traffic compositions will likely have different effects on the traffic operation of roundabouts. This paper presents the operational analysis for a single-lane roundabout with driverless vehicles under different traffic compositions. Considering the uncertainty of the technological capabilities and the legal and behavioral acceptability of DLCs, sensitivity analyses were conducted. The simulation results for one sample case study show that when the roundabout operates at a Level of Service B (LOS B) or better, DLCs will not have an effect on either the capacity or the control delay. A reduction in control delay will occur when DLCs are set with shorter headways (as assumed for the connected DLC) under low traffic composition (1% to 10%) on a single-lane roundabout operating near capacity. When the share is below 1%, the effect of DLCs is very small. When the LOS of the roundabout approaches E, more reduction in delay is expected when the portion of DLCs in traffic increases.