Tech

Self-Driving Roadway Safety Requires Industry Leadership

As details continue to emerge about the circumstances surrounding the fatal tragedy involving a pedestrian and a self-driving Uber in Arizona, this unfortunate incident raises important questions about the promise and current limitations of driverless car technology, and creates an opportunity to examine the challenge of roadway safety and how to best leverage this technology to save lives. It’s also critical that we keep the incident in the appropriate context. Today the United States experiences an estimated 40,000 traffic deaths per year. Autonomous vehicle technology is still the key to saving tens of thousands of lives.

Historically, the federal government has closely regulated automotive safety. This relationship has created public trust that, with only rare exceptions, cars are worthy of safeguarding our lives and those of our families. AV technology has enormous potential to save even more lives and let Americans travel with increased safety wherever they wish to go. And because of the unique and futuristic nature of AV technology, public trust will be even more important – and even more fragile.

Currently, much of the testing and development work for self-driving technology is conducted on public roads. As such testing becomes more widespread the debate around how to best deploy this innovation also has grown. For the tremendous potential that this lifesaving technology holds, for society to realize, a trusted framework that advances the safe, yet expeditious, nationwide deployment of this transformative technology is required.

A national safety framework is the only way to generate consumer trust in a technology that has so much promise but remains so unfamiliar to most of the population. However, due to how rapidly the technology is developing and the uncertainty that remains, the federal government is understandably not currently ready to write comprehensive safety regulations for AVs.

This situation is largely addressed by legislation that is pending before the U.S. Senate, which would create a customized, incremental approach to formulating AV regulations and establish mechanisms to incorporate input from AV technology developers.

Whether or not the legislation is passed – and it is deeply in the public interest for it to pass – now is the time for the broad array of stakeholders including automakers, technology companies, AV software developers, consumer–safety advocates, research institutions, regulators and legislators to start working together to create and advance a framework for AV safety. We do not need to – and should not – wait for legislation to pass, or for the National Highway Traffic Safety Administration to force the parties to do so.

Fortunately, the industry has started to take action.

A current example of such an effort is called the Responsibility Sensitive Safety model, made public by Mobileye, an Intel company. RSS is a model that enables AVs and human-driven vehicles to interact safely, offering insight and transparency into how and why the vehicle’s driving decisions are made. The building block for good decision-making is first and foremost a proven system of sensors that provides high-accuracy data on the environmental (detection and classification of vehicles, pedestrians, cyclists, lanes, etc.). Systems of this nature are already available in many Advanced Driver Assist Systems on vehicles today and would be able to provide a clear picture of the exact environment that preceded an incident.  But, additionally, it will be critical to provide transparency on why a decision was made and to clearly illustrate the parameters that the system used to evaluate the safety of that decision.

In incidents like the one in Arizona, RSS could help to break down the actions that took place and provide digestible data to help with the investigation, as well as to better the AV technology by learning from this event to prevent a recurrence. While RSS is certainly an important idea worth public discussion and engagement, our purpose is not to promote a particular approach to AV safety but to urge other companies and stakeholders to put forward proposals and actively play a role in this discussion. As more AV technology experts come forward with their ideas on how to promote public confidence, these ideas will be vetted in the public domain with input from industry, government and other stakeholders, and the best ideas will stand to prevail. By allowing this process take place in public view, both the debate and its eventual outcome will boost public confidence that companies are seriously grappling with how to best improve the safety of AVs.

We understand that this is not easy for industry; approaches to assure AV safety are often the result of significant investments that set companies apart from their competitors.

However, any lack of public confidence in AVs is a crisis that will damage the business prospects of all companies in this sector. At the same time, it will still be critical for any developed safety model to be technology-neutral, allowing industry to advance without stifling innovation or favoring one AV technology over another.

Today the industry and the consumers are at an inflection point at which broader discussion and collaboration on how best to ensure AV safety is necessary. The vast societal benefits of AVs are too important for this collaborative effort to be delayed any longer.

Cuneyt Oge is the past president of the Society of Automotive Engineers and served on SAFE’s Commission on Autonomous Vehicle Testing and Safety. Robbie Diamond is the CEO and president of Securing America’s Future Energy, an organization committed to advancing the mobility revolution to strengthen economic and national security.

Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.

Morning Consult