Autonomous Vehicles Save Lives, But Barriers Still Abound

After years of rapid technological advances, autonomous vehicles are poised to revolutionize the motor vehicle industry. As society speeds toward a driverless future, policymakers are looking to implement regulatory mechanisms to maximize the public benefit. In “Preparing a Nation for Autonomous Vehicles: Opportunities, Barriers and Policy Recommendations,” Daniel J. Fagnant and Kara Kockelman study the barriers and benefits of mass autonomous vehicles adoption in the United States. The authors identify numerous potential economic benefits, including a reduction in traffic congestion, fuel use, and parking costs. However, the most consequential benefit would be the vast reduction in lives lost every year from motor vehicle collisions.

National Highway Traffic Safety Administration data indicates that nearly ninety percent of motor vehicle collisions are due to human error. If autonomous vehicles replaced human drivers, such accidents would be reduced or eliminated. The authors cite research suggesting that “self-driving cars” could potentially push the motor vehicle fatality rate down to one percent of current levels, making automobiles as safe as trains and commercial aircrafts. The degree to which the benefits of autonomous vehicles are realized depends upon how widely consumers adopt them. The authors posit that if autonomous vehicles capture ten percent of the motor vehicle market, over 200,000 crashes could be prevented and 1,100 lives saved each year; if they reach ninety percent market share, over four million crashes and nearly 22,000 traffic fatalities could be prevented each year.

The study notes that safety concerns are a possible barrier to adoption. Consumers may perceive driverless vehicles as inherently unsafe, regardless of the technology’s quality. Accidents involving autonomous vehicles also raise liability questions that regulators, insurers and courts have yet to clearly address. Both of these concerns make it difficult for regulators to determine exactly how safe autonomous vehicles ought to become to be certified for use on public roads. It is unclear whether autonomous vehicles that are only slightly safer than human drivers would be acceptable to regulators and consumers; this may delay their introduction onto public roads until manufacturers prove autonomous vehicles significantly outperform human drivers.

However, there are grave consequences to delaying their introduction. RAND Corporation researchers Nidhi Kalra and David Groves recently released a report stating that the total number of lives saved by the deployment of autonomous vehicles depends not only on their safety record and market share, but also on how quickly they are able to displace human drivers. The researchers’ model assumes that the software governing self-driving cars improves only through exposure to real-world conditions. Therefore, the rate at which these vehicles become safer scales with miles traveled—the sooner autonomous vehicles are allowed on public roads, the faster their safety performance will improve.

Kalra and Groves created a statistical model that enables users to explore the safety implications under hundreds of possible scenarios. Users can vary factors such as the year autonomous vehicles are commercially introduced, the rate at which they are adopted, how they change consumer driving patterns, and the rate at which vehicle safety improves. Their model estimates that early deployment of autonomous vehicles—even while they are only slightly safer than human-driven cars—results in fewer lives lost in the short-term in nearly every scenario, and fewer lives lost in the long run in every scenario. According to their model, the difference between introducing marginally safer autonomous vehicles in the near future and delaying ten years can be measured in hundreds of thousands of lives.

If the RAND model’s estimates are accurate, then regulators must ensure autonomous vehicles will have an open road to market as soon as they meet basic safety standards. Fagnant and Kockelman offer three policy recommendations for putting the country in the fast-lane to a driverless future: expand federal funding for research into how autonomous vehicles will affect the transportation system, develop uniform federal guidelines for certifying autonomous vehicles, and swiftly pass legislation addressing potential liability issues. Implementing these changes would streamline licensing, ease uncertainty among manufacturers and investors and save countless lives. Given the high stakes, policymakers may be incentivized to act sooner rather than later.

Article source: Fagnant, Daniel J, and Kara Kockelman. “Preparing a Nation for Autonomous Vehicles: Opportunities, Barriers and Policy Recommendations.” Transportation Research Part A: Policy and Practice, Volume 77. (2015): 167–181.

Featured photo: cc/(chombosan, photo ID: 864462052, from iStock by Getty Images)