Virginia Tech report finds national crash rate for conventional vehicles higher than crash rate of self-driving cars
A new report, “Automated Vehicle Crash Rate Comparison Using Naturalistic Data,” performed by the Virginia Tech Transportation Institute and commissioned by Google, shows that the crash rates for self-driving cars are lower than the national crash rate of conventional cars.
Results show that when data is adjusted for unreported crashes and take into account accident severity, the national crash rate for conventional vehicles is higher than the crash rate of self-driving cars.
Up until now, comparisons based on existing data have been incomplete as requirements in each state for police reported crashes differ, and the majority of severe crashes that go unreported. Estimates of unreported rates of crashes have ranged from as little as 15.4% to as much as 59.7%. The result is that the current national crash rate is essentially a low estimate of the actual crash rate. Meanwhile, self-driving cars are required to report every crash, regardless of severity.
The report examines national crash data and data from naturalistic driving studies that closely monitors the on-road experience of 3,300 vehicles driving more than 34 million vehicle miles, to better estimate existing crash rates, and then compares the results to data from Google’s Self-Driving Car program.
Driving safety on public roads was examined in three ways. The total crash rates for the Self-Driving Car and the national population were compared to
Rates reported to the police. The researchers calculated crash rates from the Google Self-Driving Car project per million miles driven, broken down by severity level. These Self-Driving Car rates were compared to rates developed using national databases which draw upon police-reported crashes and rates estimated from the Second Strategic Highway Research Program (SHRP 2) Naturalistic Driving Study (NDS).
Crash rates for different types of roadways. SHRP 2 NDS data were then used to calculate crash rates for three levels of crash severity on different types of roads, broken down by the speed limit and geographic classification (termed “locality” in the study; e.g., urban road, interstate).
Scenarios that give rise to unreported crashes. SHRP 2 NDS data were again used to describe various scenarios related to crashes with no known police report. This analysis considered whether such factors as driver distraction or impairment were involved, or whether these crashes involved rear-end collisions or road departures.
Crashes within the SHRP 2 NDS dataset were ranked according to severity for the referenced event/incident type(s) based on the magnitude of vehicle dynamics (e.g., high Delta-V or acceleration); the presumed amount of property damage (less than or greater than $1,500, airbag deployment); knowledge of human injuries (often unknown in this dataset); and the level of risk posed to the drivers and other road users.
Google Self-Driving Car crashes were also analyzed using the methods developed for the SHRP 2 NDS in order to determine crash severity levels and fault (using these methods, none of the vehicles operating in autonomous mode were deemed at fault in crashes).
|SHRP 2 NDS and self-driving car crash rates per million miles at three levels of severity, Level 1 being the worst. Source: VTTI. Click to enlarge.|
Key findings include:
Adjusted for unreported crashes and accident severity (accidents that fall within the two highest severity levels), the national crash rate estimates of 4.2 crashes per million miles is higher than the crash rates for the Self-Driving Car operating in autonomous mode (3.2 per million miles).
The crash rate of conventional vehicles at all levels of severity is higher than the self-driving car crash rates, according to analysis of the Second Strategic Highway Research Program (SHRP 2) Naturalistic Driving Study.
Current data suggest that conventional vehicles may have higher rates of more severe crashes than self-driving cars, but given the small overall number of crashes for the self-driving car at these levels, there is insufficient data to draw this conclusion with strong confidence.
However, there is statistically-significant data that suggest less severe events may happen at significantly lower rates for self-driving cars than conventional vehicles.
When the Virginia Tech Transportation Institute, using methods developed for the Second Strategic Highway Research Program Naturalistic Driving Study, analyzed the Google Self-Driving Car events, none of the vehicles operating in autonomous mode were deemed at fault.
As self-driving cars continue to be tested and increase their exposure, the uncertainty in their event rates will decrease. This is particularly appropriate to vehicles intended for lower-speed use where less-severe events are the most likely to be encountered by the newer generation of the Self-Driving Car fleet.