Advocates’ Vice President Henry Jasny’s Statement to NHTSA on Autonomous Vehicles

  • October 20, 2017
150 150 Advocates for Highway and Auto Safety

Good morning. I am Henry Jasny, Vice President and General Counsel for Advocates for Highway and Auto Safety (Advocates).  Advocates is a coalition of public health, safety, and consumer organizations, insurers and insurance agents that promotes highway and auto safety through the adoption of safety laws, policies and regulations.  Advocates is a unique coalition dedicated to advancing safer vehicles, safer drivers, and safer roads.

Advocates has always enthusiastically championed technology and for good reason.  It is one of the most effective strategies for reducing deaths and injuries.  NHTSA has estimated that since 1960 more than 600,000 lives have been saved by motor vehicle safety technologies.

In 1991, Advocates led the coalition that succeeded in putting the airbag mandate in the ISTEA bill. As a result, by 1997, every new car sold in the United States was equipped with a front seat airbag and the lives saved have been significant.  Advocates continued to build on our success by promoting lifesaving technologies in other bills and regulatory proposals.  These efforts included electronic stability control, anti-lock brakes, rear view cameras and other important safety improvements to passenger vehicles, trucks, and motorcoaches.

According to the latest statistics from the National Highway Traffic Safety Administration (NHTSA), 37,461 people were killed on our nation’s roads in 2016.  This is an increase of over 5 percent from 2015.  Automated vehicle technology has the potential to significantly reduce this carnage.  However, NHTSA must not place exclusive focus during the next 10 years on the development of self-driving cars to the detriment of other safety advances that have already been shown to improve safety.  In fact, the agency is currently in breach of several mandated deadlines set by Congress in the 2012 Moving Ahead for Progress in the 21st Century Act (MAP-21) to issue regulations requiring rear seatbelt reminders, updating LATCH systems for child occupants and to improve occupant protection in motorcoaches.  The agency also continues to merely issue guidelines rather than regulations to govern in-vehicle and nomadic electronic devices that contribute to distracted driving, an issue that has long been identified by the safety community and the Department of Transportation (DOT) as a major public health epidemic.

Voluntary guidelines that can be ignored by the industry are completely inadequate to ensure that American families are not put at unreasonable risk during the testing and deployment of autonomous vehicle (AV) technology.  The optional safety self-assessment proposed in the NHTSA’s latest AV Policy illustrates the shortcomings of voluntary guidelines.  No matter how comprehensive the structure of the safety self-assessment may be, manufacturers can simply choose not to publish one or to provide superficial or incomplete information.  Advocates is pleased that Waymo has released the first safety self-assessment to the public that could be helpful to some consumers.  However, the two major problems with the document are that (1) Waymo has used the submission as an opportunity to craft a slick marketing brochure for the technology and the company’s product and, (2) the submission does not provide sufficient detail and technical information to allow the public to know the safety performance of their system.

Just yesterday my computer crashed but I walked away unscathed.  AVs are computers on wheels, so the industry and the agency bear a heavy burden to ensure crash safety in the brave new world of driverless cars.  The safety self-assessment must provide the public with an honest assessment of the technology’s limitations as well as its capabilities.  Certainly, in the near future not all crashes will be avoided, not all AVs will operate as designed, and the public and policy makers must be given an accurate picture of what lies ahead.

Advocates does not believe that manufacturers should use the safety self-assessment as a sales tool.  For example, Waymo touts the fact that it has conducted “more than 3.5 million miles of on-road driving experience.”  Testimony before Congress on this very issue has pointed out that New York City taxis accumulate approximately four million vehicle miles each day.  Also, the motor vehicle fatality rate is measured by 100 million vehicle miles traveled (VMT), so 3.5 million miles is a comparatively low level of VMT exposure.  While Advocates understands that accruing on-road VMT takes time, the 3.5 million miles statistic trumpeted by Waymo should not be presented to the public as an unqualified and rigorous mark of distinction for the technology.

Waymo’s submission also touts the fact that current motor vehicle crashes often involve human error as a factor.  But the company fails to mention that 100 percent of computer programming errors involve human error as a factor.  While we are endeavoring to improve safety we must never forget that we are replacing human driver error with human programming errors –mistakes that could have widespread unintended consequences.  The industry is telling the public to not trust people to operate a motor vehicle safely, but instead trust that people can engineer and program autonomous vehicles to operate flawlessly.  We know from defects and recalls that both people and complex systems are prone to errors and mistakes.  That is why Advocates prefers a phase-in approach to AV deployment to ensure public safety.

The safety self-assessments should provide enough information to consumers, researchers and the agency to properly evaluate the current state of the development of the technology.  For example, the Waymo document provides more specific information on the behavioral competencies and test scenarios, more than what is provided in many other subject areas of the document.  However even that level of detail still falls far short of providing enough information for researchers, or the public, to understand how the systems will perform in these scenarios and what would constitute a deviation from expected performance.

The Waymo self-certification addressed many of the recommended subject areas proposed by the NHTSA, but in a very general manner.  In some cases, critical details were lacking and in the end, the public is still left with no guarantee of safety.  For example, while Waymo provided a fairly clear explanation of the SAE level of their proposed system, when it came to cybersecurity the discussion was vague and talked about an approach “that builds upon best practices” and “includes aspects” of some industry standards.  Instead, Advocates recommends that companies identify the specific standards they have used and which they believe all of the industry should be held to.

Waymo also provides detailed numbers on current crash statistics, societal costs and mobility and quality of life impacts, but provides no specific context for the results of their “3.5 million miles” of testing in terms of crash rate or fatality rate in comparison to human driver performance.   Furthermore, Waymo did not quantify the results of their work which Waymo states was able to comprehensively analyze and evaluate the safety of the self-driving system   prior to operating their vehicles on public roads.  If the analysis was so complete, why are there no quantified results presented?  The public still has no guarantee that Waymo, or any other company, will meet a specific level of performance to ensure the protection of road users.

Finally, beyond the information in the safety self-assessments the agency must require explanatory information for each specific autonomous driving system to be provided at the point of sale.  Essential information should be on a label on the vehicle, clear, concise and uniform explanations and instructions must be provided at the dealership (and on the manufacturer’s website), as well as included in the vehicle’s owner manual.

In the end, the voluntary safety self-assessments may amount to little more than marketing materials dressed up as consumer information. Advocates supports the need for consumer information, but the type and quality of that information must be objective, specific and uniform to ensure that the public is able to evaluate the technology and even compare technology between companies. However, promotional materials and consumer information are no substitute for regulations to ensure public safety.

Thank you for the opportunity to provide comments today.