How Wayve’s driverless cars will tackle one of their biggest challenges yet

Figuring out why the model behaves the way it does tells Wayve what types of scenarios require additional help. Using a hyper-detailed simulation tool called PRISM-1 that can reconstruct 3D street scenes from video footage, the company can generate custom scenarios and run the model over and over again until it learns how to handle them. How much recycling might the model need? “I can’t tell you the amount. That’s part of our secret sauce,” Rus says. “But it’s a small amount.”

The self-driving vehicle industry is notorious for hype and over-promising. Over the past year, Cruise has fired hundreds after his cars caused mayhem and injuries on the streets of San Francisco. Tesla is facing a federal investigation after its driver assistance technology was blamed for several crashes, including a fatal collision with a pedestrian.

But the industry is still moving forward. Waymo has said it now makes 100,000 robotaxi rides a week in San Francisco, Los Angeles and Phoenix. In China, Baidu says it’s giving about 287,000 rides in a handful of cities, including Beijing and Wuhan. Undeterred by accusations that Tesla’s driver-assistance technology is unsafe, Elon Musk announced his Cybercab last week with a timetable that would put these driverless concept cars on the road by 2025.

What are we to make of it all? “Competition among robotaxi operators is heating up,” says Crijn Bouman, CEO and co-founder of Rocsys, a startup that makes charging stations for autonomous electric vehicles. “I think we’re close to their ChatGPT moment.”

“The technology, the business model and the consumer appetite are all there,” says Bouman. “The question is which operator will seize the opportunity and come out on top.”

Others are more skeptical. We need to be very clear about what we’re talking about when we talk about autonomous vehicles, says Saber Fallah, director of the Connected Autonomous Vehicle Research Lab at the University of Surrey, UK. Some of Baidu’s robotaxis still require a safety driver behind the wheel, for example. Cruise and Waymo have shown that a fully autonomous service is viable in certain locations. But it took years to train their vehicles to drive on specific streets, and extending the routes, safely, beyond existing neighborhoods will take time. “We’re not going to have robotaxis that can drive anywhere anytime soon,” Fallah says.

Fallah strongly believes that this will not happen until all human drivers surrender their licenses. For robotaxis to be safe, they must be the only vehicles on the road, he says. He believes that current driving models are not yet good enough to interact with the complex and subtle behaviors of humans. There are too many extreme cases, he says.

Wayve is betting its approach will win. In the US, it will start by testing what it calls an advanced driver assistance system, a Tesla-like technology. But unlike Tesla, Wayve plans to sell this technology to a wide range of existing automakers. The idea is to build on this base to achieve full autonomy in the coming years. “We’ll have access to scenarios that meet a lot of cars,” says Rus. “The road to fully autonomous driving is easier if you go level by level.”

But the cars are just the beginning, says Rus. What Wayve is building, he says, is an embodied model that could one day control many different kinds of machines, whether they have wheels, wings or legs.

“We’re an AI shop,” he says. “Driving is a milestone, but it’s also a step.”