Happy Nice Mall

cruise control

Either way, should we rely on this new wave of firms to chase the frontrunners? Not surprisingly, Mo ElShenawy, vice president of engineering at Cruise, was not convinced. “The cutting-edge technology as it stands today isn’t enough to get us onto the stage where Cruise is,” he says.

Cruise is one of the most advanced driverless car companies in the world. It has been running a live robotaxis service in San Francisco since November. Their cars run in a limited space, but now anyone can greet a car with the Cruise app and have it pull to the curb when no one is inside. “We see a real spectrum of responses from our customers,” says ElShenawy. “Super exciting.”

Cruise built a massive virtual factory to support his software, with hundreds of engineers working in different parts of the pipeline. ElShenawy argues that the mainstream modular approach is an advantage because it allows the company to tweak new technology as it emerges.

He also rejects the idea that Cruise’s approach cannot be generalized to other cities. “We could have set off in a suburb years ago and that would have cornered us,” he says. “The reason we chose a complex urban environment like San Francisco, where we saw hundreds of thousands of cyclists and pedestrians and emergency vehicles and cars intercepting you, was very deliberate. It forces us to build something that scales easily.”

But before Cruise can drive in a new city, he must map its streets in centimeter-level detail. Most self-driving car companies use such high-resolution 3D maps. They provide the vehicle with additional information, along with the raw sensor data it receives as it moves, typically including clues such as the location of lane boundaries and traffic lights, or whether a particular street has a pavement.

These so-called HD maps are created by combining road data collected by cameras and lidar with satellite imagery. Hundreds of millions of kilometers of roads in the USA, Europe and Asia have been mapped this way. But road patterns change every day, which means map making is an endless process.

Many self-driving car companies use HD maps created and maintained by specialist firms, but Cruise makes his own. “We can recreate cities, all driving conditions, street patterns and everything,” says ElShenawy.

This gives Cruise an edge over mainstream rivals, but newcomers like Wayve and Autobrains have abandoned HD maps entirely. Wayve’s cars have GPS, but otherwise they just learn to read the road using sensor data. It may be more difficult, but it means they are not tied to any particular place.