Since the concept of driverless cars first became a serious prospect, a lot of attention has been given to the possibility of their malfunction—if an autonomous vehicle damages property or even harms a human, who is at fault? And, given a worst-case scenario, how should a vehicle's software choose between whose lives it prioritizes, the passenger or the pedestrian? This last question even became the basis for the Moral Machine, an online platform created by the MIT Media Lab that essentially crowdsources public opinion on different variations of the classic trolley problem thought experiment.
However, all of these questions had been considered largely theoretical until last night when, as The New York Times reports, a woman was struck and killed by an autonomous vehicle in Tempe, Arizona. As a major component of many predictions of futuristic "smart cities," the development and testing of autonomous vehicles hold huge implications for urbanism (ArchDaily has previously covered predictions of major change by car manufacturers and researchers) meaning that this fatal event could have a ripple effect on the development of cities.
In what is believed to be the first case of an autonomous vehicle killing a pedestrian, the vehicle—which is owned by Uber and was part of their test fleet operating in Tempe, Pittsburgh, San Francisco and Toronto—killed a woman who was crossing a street outside of a designated crosswalk. At the time, the vehicle was in autonomous mode, though it did have a human "safety driver" at the wheel. (Update: As per a report by Slate, the woman killed in the accident was named as 49-year-old Elaine Herzberg. She was struck while crossing North Mill Avenue, an 8-lane road with only one crosswalk in almost 2 miles.)
According to The New York Times, Uber has stated that it is "fully cooperating" with the local authorities regarding the issue and has suspended testing of its self-driving cars in all four cities.
At this point, it is unclear exactly what effect the death of Herzberg will have on the rollout of driverless technology. As highlighted on Twitter by Streetsblog writer Angie Schmitt, Uber and its rival Waymo (operated by Google) have racked up a total of 6 million miles of travel with their combined testing of autonomous vehicles, making their one fatality a poor record compared to human drivers' record of about 1.25 fatalities for every 100 million miles traveled. This leaves autonomous vehicles with some work to do in order to prove that they are, as often promised, safer than human-controlled cars.
Even if driverless vehicles do eventually live up to their safety claims, the fact that a pedestrian has been killed is likely to create a sense of urgency among legislators to create rules around their introduction. And while, in this case, local law enforcement was quick to ascribe fault to the pedestrian, some experts have nevertheless highlighted that North Mill Avenue, a wide road without parked cars or other objects to obscure a driver's vision, should have been the last place this accident could have happened. Arizona State University urban planning professor David King, for example, told CityLab that "If there is any real-world scenario where it would be seemingly safe to operate in an automated mode, this should have been it. Something went seriously wrong."
While all of these issues pre-date the death of Elaine Herzberg, it is highly probable that they will now be given a higher priority in the minds of politicians, urbanists, and tech-evangelists. They may now be added to the list of issues that car website Jalopnik last year wrote could consign driverless vehicles to be ever on the horizon, but never a reality. Or they might just cause our driverless future—and its associated urban design—to be delayed by a year or two. Either way, the first death of a pedestrian at the hands of an autonomous vehicle is likely to be an event that becomes an important reference point for future urbanists.
News via The New York Times, Slate, CityLab.
Update: This article was updated on March 22nd to include commentary on the event's importance for urban design.
How Driverless Cars Could, Should - and Shouldn't - Reshape Our Cities
