There is Only an “Emergency Stop” Button in Google’s Second Generation Driverless Car

By Cory Bilton

Google Driverless Car

In case you haven’t been following along, Google is developing a driverless car. While other groups and researchers are also building autonomous vehicles, Google’s project has created the most buzz by far. Up until recently, Google’s driverless car was a Toyota Prius with various pieces of equipment, sensors and laser radar, mounted on it. Recently, Google unveiled the next generation of its driverless car, pictured above. This version is a specially built vehicle with all of the equipment incorporated right into the car itself. As I’ve blogged before, one of the puzzling questions is what is the legal duty of the occupant of a Google car?

The second-generation Google car has thrown a wrench in the works on this issue: the car has no steering wheel. There is, however, still an “emergency stop” button in this new version of Google car.

Despite all the excitement being showered on the autonomous vehicle, it remains unclear what is expected of the driverless car occupant. No steering wheel means that the occupant can’t take over operation of the car. No can do. If you are a control freak, you are out of luck. But by having an emergency stop button, there is still some element of control. So one of the practical questions remains whether the occupant is supposed to pay attention or not? The Google promotional video on Youtube featured some occupants that appear to be blind. So Google doesn’t intend the occupants to be an active participant in the vehicle’s operation. In fact, if the occupant is expected to pay attention, it defeats one of the main selling points of the autonomous vehicle—that you can read the paper, eat dinner, or spend quality time with your kids instead of paying attention to driving.

Loving to Hate Driving

Somewhere in the discussion about autonomous vehicles, we have to decide how involved we want to be while driving. The lure of being able to jump in the car and have the car effortlessly carry us to our destination is appealing. Articles about driverless cars inevitably highlight the promise of new found freedom of devoting our commute to more productive pursuits. But maybe we just love to hate driving. If I really want to devote my time in the car to reading the newspaper, I won’t be able to prevent harm by hitting an emergency stop button when I think the car is making a mistake. These next generation driverless cars may be controllable in more ways than just an emergency stop button, too. As elaborated recently in the Washington Post, engineers working on driverless cars are trying to create driverless vehicles that behave differently, depending on the occupant’s preferences. The idea is that if you prefer to drive more aggressively, you can tell the driverless car to drive more aggressively, too. The open question is how much vehicle control are people willing to give up.

The Liability Crossroads for Autonomous Vehicles

Whether we make a deliberate decision, or an passive one, we will have decided the role that autonomous vehicle occupants will play before the cars become available for the public. Either autonomous vehicle occupants are expected to be passive passengers or they are active authorities.  If occupants are passive passengers, they are unlikely to be legally responsible for any harm caused by the vehicle. But if occupants are expected to remain vigilant and save the car from mistakes, the human remains in the driver seat, legally speaking.  The next generation Google Car removes some elements of control, but retains others.

Maybe autonomous vehicles development will continue on the path which will allow them to operate on a continuum of human involvement. For some of us, maybe we want the car to operate completely by itself. Maybe you want to put your child in the car destined for soccer practice and let the vehicle accomplish the task. But other occupants may want to retain some control of the vehicle, or a lot of control, even. Maybe we would prefer telling the car to drive more aggressively or take over control in certain circumstances. Until we know what kind of control an occupant retains in an autonomous vehicle, we can only continue guessing who will be liable when something goes wrong and autonomous vehicles cause injury or damage.