In any case, should we expect this new wave of firms to go after the leaders? Not surprisingly, Mo El Shenawy, executive vice president of engineering at Cruise, is not convinced. “The level of technology that exists today is not enough to bring us to the level that Cruz is at,” he says.
Cruise is one of the most advanced self-driving car companies in the world. Since November, a robot taxi service has been operating in San Francisco. His cars operate in a limited area, but now anyone can catch a car with the Cruise app and force it to the curb when no one is inside. “We see a wide range of reactions from our customers,” says El Shenawy. “It’s super exciting.”
Cruise has built a huge virtual factory to back up its software, and hundreds of engineers are working on different parts of the pipeline. ElShenawi argues that the core modular approach is an advantage because it allows the company to replace new technologies as they become available.
He also dismisses the idea that Cruise’s approach cannot be extended to other cities. “We could have started in the suburbs a few years ago and it would have driven us into a corner,” he says. “The reason we chose a challenging urban environment like San Francisco where we see hundreds of thousands of cyclists and walkers, ambulances and cars cutting you was very deliberate. It forces us to create something that scales easily.”
But before Cruise travels to a new city, it must first map its streets to centimeter accuracy. Most self-driving car companies use these high-definition 3D maps. They provide the vehicle with additional information on top of the raw sensor data it receives while driving, typically including clues such as the location of lane boundaries and traffic lights, or the presence of curbs on a particular section of the street.
These so-called HD maps are created by combining road data collected by cameras and lidar with satellite imagery. Hundreds of millions of miles of roads in the US, Europe and Asia have been mapped in this way. But road maps change every day, which means that the process of creating maps is endless.
Many self-driving car companies use HD maps made and maintained by specialist firms, but Cruise makes its own. “We can recreate cities — all the traffic conditions, street layouts and everything else,” says El Shenawy.
This gives Cruise an edge over its main competitors, but newcomers like Wayve and Autobrains have eschewed HD cards entirely. Wave’s cars have GPS, but otherwise they learn to read the road using only sensor data. It may be more difficult, but it means they are not tied to a specific location.