Data and Methods to Estimate Connected and Automated Vehicle Penetration Rates

About the report:

Automated driving systems are becoming increasingly prevalent on Virginia roadways.  These vehicles rely on radar, lidar, and machine vision to operate and may detect road markings, barriers, and other vehicles in ways that human drivers do not.  Vehicles may also leverage wireless communication to assist in driving, path planning, and communicating with roadside infrastructure.  Recent research has investigated the impact of an increasingly connected and automated vehicle (CAV) fleet on safety and capacity, but these estimates rely on accurate measurements of the volumes or proportions of vehicles on the road equipped with and using these technologies. 

The Virginia Department of Transportation (VDOT) does not currently have a way to estimate the volume of connected vehicles, automated vehicles, or CAVs operating on Virginia roadways.  The purpose of this study was to identify data required for VDOT to estimate accurately the proportion of vehicles equipped with and using vehicle automation technologies that may affect safety and operations.  The study also investigated practical ways to collect these data using both available data sources and proprietary and future data sources.  Using existing data sources, the study estimated that in 2022, 16% of the vehicle fleet was equipped with adaptive cruise control, 16% with automatic emergency braking, 22% with forward collision prevention, 8% with lane centering assist, 15% with lane departure prevention, and 25% with pedestrian automatic emergency braking.  These percentages were further adjusted to reflect observed driver activation rates. 

The study concluded that there are currently quality estimates of penetration of certain vehicle automation features available, and there are several methods to obtain reasonable estimates of other automation technologies.  No methods were found that could directly measure technology usage rates in actual on-road driving, although an upper estimate can be calculated based on observed rates of system activation for vehicles brought in for service. 

The study recommends continued dialogue with research partners to obtain access to aggregated or anonymized CAV penetration data in proprietary datasets.  The study also recommends continued monitoring of ongoing research in on-road CAV technology usage rates from either naturalistic driving studies, industry data, or federal data collection efforts.  Implementing these recommendations would benefit VDOT by providing more accurate data on the rate of CAVs in the vehicle fleet, allowing the development and calibration of existing empirical models of the effect of CAVs on traffic flow, capacity, safety, and infrastructure planning.

Authors