No One at the Wheel
Print Issue: February 2017
Jeffrey Zients, director of the U.S. National Economic Council, has some advice for workers who are worn out by their daily car commute: simply take your hands off the wheel and turn your attention to something other than driving. “Your commute becomes restful or productive, instead of frustrating and exhausting,” Zients said at a recent press conference.
Of course, Zients’ vision assumes that the commuter is in a driverless car—or in industry parlance, a highly automated vehicle (HAV). A few months ago, the development of such driverless cars received a jumpstart from U.S. officials, who released new guidelines for operating the vehicles while promoting the government’s position that American highways will be safer when more cars are machine-driven.
“Too many people die on our roads—35,200 last year alone–with 94 percent of those the result of human error or choice. Automated vehicles have the potential to save tens of thousands of lives each year,” U.S. President Barack Obama wrote in a Pittsburgh Post-Gazette op-ed article about the new guidelines. “And right now, for too many senior citizens and Americans with disabilities, driving isn’t an option. Automated vehicles could change their lives.”
Global consulting firm McKinsey & Company has predicted that consumers will begin to adopt driverless cars starting in 2020—and that their popularity will overtake conventional cars by 2050.
But not everyone shares the government’s rosy view about these developments. In some quarters, security and safety concerns about driverless cars abound. Those that are concerned argue that the cars themselves, and the roads they will drive on, will both be too vulnerable once automated vehicles become more common.
Despite these concerns, industry is speeding forward, and carmakers are vying to enter the driverless car market first. Tesla has already sold tens of thousands of cars with a self-driving feature known as Autopilot. The company says it aims to be the first to put a fully driverless car on the road, although it hasn’t set a specific date.
Both the Ford Motor Company and Nissan have said they plan to release driverless car models within the next five years. Driverless taxis may come sooner. General Motors Company, working with taxi service startup company Lyft, said it plans to start testing a fleet of driverless taxis soon.
Internationally, the NuTonomy company has said it will provide self-driving taxi services in Singapore by 2018, and expand to 10 cities around the world by 2020. And Nissan expects to release a feature called SuperCruise that will allow for hands-free highway driving.
Given this frenzy of market activity, Zients and U.S. Secretary of Transportation Anthony Foxx released the new U.S. federal guidelines at a press conference last September. The guidelines represent best practice guidance rather than rulemaking, and they outline the government’s expectations in terms of safety and how the new technologies should be regulated.
The guidelines are broken up into four main areas. The first part is a 15-point safety standard for the design and development of autonomous vehicles. The second part is guidance for states developing their own driverless car policies. The third consists of information on how current regulations can be applied to driverless cars. The fourth is a discussion of specific new regulatory tools and authorities that transportation officials believe might be needed for proper development of driverless cars.
The safety standards address questions such as: How will driverless cars react if their technology fails? How will occupants be protected in crashes? What measures should be put in place to preserve passenger privacy?
Also included is guidance on how automakers should approach cybersecurity issues in driverless vehicles. U.S. federal officials encourage carmakers “to design their HAV systems following established best practices for cyber physical vehicle systems.” The guidance calls on manufacturers to use best practice principles published by U.S. agencies and organizations, such as the National Institute for Standards and Technology, the Alliance of Automobile Manufacturers, and the Automotive Information Sharing and Analysis Center.
“The identification, protection, detection, response, and recovery functions should be used to enable risk management decisions, address risks and threats, and enable quick response to and learning from cybersecurity events,” the guidance reads.
The guidance adds, however, that “this is an evolving area and more research is necessary before proposing a regulatory standard.” And the view that more research is necessary is shared by many, including those who argue that driverless cars have a long way to go before security and safety concerns are satisfied.
Cybersecurity is the biggest concern for companies now evaluating risk in the developing driverless car industry, according to a recent survey conducted by Munich Re, the German reinsurance and risk management firm.
In the study, 55 percent of corporate risk managers surveyed named cybersecurity as their top concern regarding driverless cars. In the cybersecurity category, respondents said they believed the greatest threats were auto theft by an unknown individual hacking into and overtaking vehicle data systems (42 percent) and the failure of smart road infrastructure technologies (36 percent).
Researchers have demonstrated how a hacker can remotely take over the brakes, engine, or other components of a standard car. The attack surface for a driverless car is even larger, experts say, because it contains extra computers, sensors, and more extensive Internet connectivity.
There are also security and safety concerns regarding the roads that the driverless cars will travel on, says Howard Jennings, managing director of Mobility Lab, a transportation research firm.
Early testing shows that driverless cars will be able to drive with less space between them compared with conventional cars. But in areas that are meant to be village-type developments with many pedestrians, and with densely packed cars driving down narrow streets, this feature could create safety hazards. “We could have an unintended consequence here,” Jennings says.
Related to this issue is what some call the Waze effect, named after the community-based traffic application. When suggesting alternate routes, the Waze app sometimes sends many drivers down the same small street, causing logjams on narrow roads. Driverless cars could wind up doing this as well.
Finally, Jennings says that people make transportation choices based partially on the “hassle factor”—they take public transportation downtown because they think parking will be a problem, for example. If driverless cars make car commuting less stressful and take the hassle out of parking, many people may choose them over public transportation. This could put a huge unanticipated strain on road networks, causing infrastructure safety issues due to overuse.
Finally, some fear that these significant concerns are not being addressed quickly enough, given that driverless cars for consumer purchase may be only a few years down the road.
“It’s no longer a matter of if, but rather when the time will come for the widescale adoption of automated vehicles,” said Munich Re President and CEO Tony Kuczinski when the survey was issued. “The timeline for adoption may be sooner than many realize.”