January 03, 2013

The emergence of self-driving cars is imminent, making urgent the task of creating the appropriate legal infrastructure, according to legal scholar Bryant Walker Smith in a recent article at Slate. Self-driving cars were envisioned as far back as the 1930s, according to Walker Smith, but were always "just 20 years away" because they were conceived of as requiring special infrastructure, such as special roads embedded with magnets.

Probably the visionaries of the day were thinking of driverless cars as a kind of variation on locomotive trains — Walker Smith doesn't say, but he contrasts the old vision with the present reality:

Fast forward to today, and many of the modern concepts for such vehicles are intended to work with existing technologies. These supercomputers-on-wheels use a variety of onboard sensors— and, in some cases, stored maps or communications from other vehicles—to assist or even replace human drivers under specific conditions. And they have the potential to adapt to changes in existing infrastructure rather than requiring it to alter for them.

The problem is that the inherent autonomy of rail-less vehicles opens up a range of difficult questions. Walker Smith addresses the fundamental question of whether such vehicles can even be legal today. "For the United States, the short answer is that they probably can be," he comments. However, some of the key challenges will be addressing the impact of autonomous technology on the responsibility of drivers — assuming they are even in the picture. Says Walker Smith:

Must the "drivers" remain vigilant, their hands on the wheel and their eyes on the road? If not, what are they allowed to do inside, or outside, the vehicle? Under Nevada law, the person who tells a self-driving vehicle to drive becomes its driver. Unlike the driver of an ordinary vehicle, that person may send text messages. However, they may not "drive" drunk—even if sitting in a bar while the car is self-parking. Broadening the practical and economic appeal of self-driving vehicles may require releasing their human users from many of the current legal duties of driving.

Significant technical technical challenges must still be addressed before self-driving cars can become common, but that gives us time to deal with novel legal issues, such as "How do you ticket a robot? Who should pay? And can it play (or drive) by different rules of the road?" Of course, legal issues of responsibility will factor in the insurance dimension of driverless cars. It may be, as Celent's Donald Light has speculated, that self-driving cars will result in safer transportation — and a corresponding drop in automobile premium. But Walker Smith's piece suggests that the relevant technologies will also complicate the legal issues associated with telematis, or pay-as-you go insurance:

Data protection is a more pressing issue. Many cars and trucks available today already collect driving data through onboard sensors, computers, and cellular devices. But imagine taking a dozen smartphones, turning on all of their sensors and cameras, linking them to your social media accounts, and affixing them to the inside and outside of your vehicle. That is an understatement of a self-driving vehicle's potential data collection. Because consumer versions of such vehicles do not yet exist, we don't know what data will actually be collected or how it will be transmitted and used. However, legal issues related to disclosure, consent, and ownership will mix with important policy questions about the costs and benefits of data sharing. Indeed, some research vehicles in Germany already have privacy notices printed on their sides to warn other road users.

[For more on the potential impact of self-driving cars on the insurance industry, see Prepare for Deep Auto Insurance Premium Drop Scenario, Celent Report Advises .]

ABOUT THE AUTHOR
Anthony O'Donnell has covered technology in the insurance industry since 2000, when he joined the editorial staff of Insurance & Technology. As an editor and reporter for I&T and the InformationWeek ...