As we travel further into the unknown, the risks will increase.
IT DOESN’ T REALLY matter who I talk with—CE Os, friends, colleagues, random people I’m introduced to—when the subject turns to the future of the automobile, specifically autonomous vehicles, the conversation starts to crackle.
Almost every day there is another development, another story written, or an opinion offered (like this one).
Some of the latest piling up in my inbox: There’s a new fog-penetrating imaging system being developed by MIT researchers that could change car camera technology; a study by Ed Sappin, CEO of Sappin Global Strategies, that examines whether self-driving cars will kill the big automakers; and a survey from Ipsos that says one in four Americans would “never use” an autonomous vehicle.
Then there’s the type of news that really gets everyone’s dander up: when something goes haywire. Reports of the death in March of 49-year-old Elaine Herzberg, who was hit and killed by a self-driving Uber car in Tempe, Arizona, while crossing a dark road away from a crosswalk, was broadcast breathlessly around the world. It was the first known fatality involving a pedestrian and a vehicle in autonomous operation. Uber’s permit to operate its AV fleet in Arizona was subsequently revoked amid renewed calls for more regulation and a full-scale re-examination of our rush toward automotive autonomy.
The incident spurred us to have automotive technology reporter Doug Newcomb examine where this country stands in terms of state and federal government approaches to autonomous-technology testing. You can read his report beginning on page 98.
It was still a scalding hot-button topic at a dinner during the New York auto show I had with colleagues, including Newcomb and Wolfgang Ziebart, Jaguar Land Rover’s head of product development. There is a video of the crash, and after watching it, Ziebart questioned how the technology did not properly detect Herzberg and react. Anyone with even baseline knowledge of how autonomous systems operate— let alone a seasoned engineering mind like Ziebart—would arrive at the same question. I know I did.
But such an incident cannot lead to knee-jerk reactions. Something went wrong. A life was lost. Lives are lost all the time when cars driven by humans hit pedestrians. I’m not minimizing Herzberg’s death but rather reflecting on what it means. We can be afraid of technology and the unintended and sometimes truly unfortunate consequences that come along with it, but the benefits we can reap as a society could be tremendous and far-reaching.
That’s not to say the emerging autonomous vehicle landscape shouldn’t be regulated. As Newcomb’s piece reveals, what is happening now at the state and federal level is woefully inadequate. We need to have some real, topdown action to create a climate where the progress being made can be done safely and with some boundaries. Additionally, those developing the systems need to slow their driverless roll and better address the flaws and holes in their approaches, assess where infrastructure comes in and how cars will communicate with it, how to best navigate the legal implications, and other obstacles.
One way to deal with testing, at least in the short run, is geofencing (essentially virtual boundaries using GPS). If you as a non-autonomous driver or pedestrian enter a geofenced zone, you would know that some cars running within it are fully autonomous, heightening your awareness. It’s not hard to imagine a time soon where parts of Manhattan could be geofenced, where cabs and ridehailing vehicles within it would be driverless.
Most of the automaker executives I’ve talked with lately are confident they will be able to develop cars equipped with fully autonomous systems you’ll be able to turn on when you need/want them but turn off when you want to run free, to drive yourself. What they aren’t confident about is when. They all claim the technology has basically taken shape, but it’s those last miles into the unknown that will be the hardest.
Other than those one in four who say they will never let a car drive them, most people outside of the automotive beltway I talk with are intrigued about the prospect of autonomous cars. They want to be able to sleep in their car, work in their car. My parents are getting older. I worry about them traveling long distances, and they hate traffic with a passion. (Who doesn’t?) There’s nothing more I’d rather have them do than hit a button and have them whisked to my house quickly and safely without laying a hand on the wheel.
Some days I think that time is a long way away, 20 to 30 years at least. Others I think it’s closer. No matter what, there will continue to be risks and likely more sad incidents. We need to get better as a society at mitigating them.
Are you looking forward to the advent of autonomous cars? Or are you loathing the prospect of a self-driving future? Let us know at email@example.com. AM