When Self-Driving Cars Don’t Actually Drive Themselves
I experienced my first ride in a self-driving car almost a decade ago. Like many others, I was awestruck by the machine’s ability to master a skill once exclusive to humans. But then, the journalist in me kicked in.
Over the past ten years, initially with Wired Magazine and now with The New York Times, I’ve reported on the fast-paced evolution of self-driving cars. During this period, it became increasingly apparent that these vehicles, while impressive and surprisingly agile, have not yet matched the cognitive power of the human brain. And that remains true today.
My reporting recently culminated in an article that highlights how driverless cars enlist human assistance. It reveals that although today’s robot taxis operate without drivers, some even without steering wheels, they still depend on the judgment of human technicians.
In April, my colleague Yiwen Lu and I visited a command center in Foster City, California, run by Zoox, a self-driving car company owned by Amazon. We discovered that, similar to other robot taxis, Zoox’s self-driving vehicles occasionally require assistance. This support comes from human technicians stationed in a spacious command center.
These technicians may intervene to assist a robot taxi hundreds of miles away. For example, if a Zoox car encounters an unfamiliar construction zone, it will trigger an alert in the command center.
Using a computer mouse, a technician can then draw a new route on a digital road map, guiding the car around the construction. This is one of the critical ways these vehicles still rely on human oversight to function effectively.
