Will self-driving cars bring us a safer future on the roads? Not everyone is convinced

Liz Lindqwister describes the first time she rode in a self-driving taxi as "kind of a surreal experience."

"You just get in the back seat. There's a robot voice that says, 'Please put on your seatbelt and get ready for the ride.' And it just goes," said Lindqwister, a data journalist for the San Francisco Standard, an online news organization.

Once it starts going, "it's just a car that moves eerily slowly and carefully throughout the streets."

But the reactions of people nearby seeing a car move with no one in the driver's seat was "kind of a wild experience," she told CBC Radio's Spark.

That experience is gradually becoming far less unusual, at least in California. In August, the state expanded permits for two autonomous vehicle companies: Cruise, owned by General Motors, and Waymo, owned by Google parent company Alphabet.

But after a handful of controversial incidents, Lindqwister said, you'll find more polarized opinions about robotaxis in the city than neutral ones.

"People feel very strongly about them both ways, both supporting and completely against them," she said.

In August — one day after California approved the expanded use of robotaxis — 10 of them ground to a halt on a busy San Francisco street, creating gridlock spanning several streets during a music festival.

Later that month, city officials claimed Cruise's cars delayed an ambulance carrying a critically injured patient on its way to the hospital, and the patient later died.

The cases have raised questions about how safe autonomous vehicles really are at this stage in their ongoing development, and what regulations and even city planning would be required before they can become a fully integrated element of city roads around the world.

Self-driving cars use a complex array of technology to function on city streets — from multidirectional cameras and LiDAR (Light Detection and Ranging), a laser scanning system that paints a digital map of the surrounding area, to artificial intelligence (AI) designed to make driving decisions on the fly.

Common situations such as stopping at a red light and staying in a lane may be relatively simple. But as anyone who's driven through a city's downtown core knows, a lot of other things may be happening at the same time.

"The world that we live in has sort of an infinite variety of objects. And, you know, people do crazy things sometimes," said Steven Waslander, director of the Toronto Robotics and AI Laboratory at the University of Toronto and an expert on autonomous vehicles.

"They'll carry weird objects or they'll make modifications to their vehicles. And these networks, if they've never seen those before, they don't necessarily know how to respond."