BMW, A COMPANY that prides itself on building “the ultimate driving machine,” plans to start producing fully autonomous vehicles by 2021 for ridesharing programs. Think of it as Uber for people who don’t like people.
This is a surprising move, given that the company has said essentially nothing about technology that everyone from Google to General Motors to Tesla is racing to develop. And it marks a radical departure from the slow-and-steady approach of the mainstream automakers, who see the technology rolling out slowly over the next two decades.
Still, ze Germans see themselves surging ahead by relying upon help from Intel and Mobileye, an Israeli firm that dominates the market for the cameras that are key to active safety features like collision warning and lane-keeping. The trio aims to standardize self-driving technology. “It is the only way to make this crucial next step a reality,” Ziv Aviram, co-founder and CEO of Mobileye, said in a statement.
Friday’s announcement was long on promises and short on details. There’s no word on where BMW will deploy the cars, which ridesharing platform it will work with, or what role Intel plays in the partnership. And BMW is well behind the competition, which is led by Google. The company’s fleet of two dozen or so fully autonomous vehicles logs 10,000 to 15,000 miles each week and has covered 1.3 million miles in all.
Still, BMW has a key partner—Mobileye and its comprehensive cartographic capabilities. Mapping is the key to making these cars work. A self-driving car with detailed maps can dedicate far more computing power to identifying and addressing things like cyclists, pedestrians, and other cars, in real-time. That’swhy TomTom still exists, and why BMW, Daimler, and Volkswagen chipped in to buy mapping company Here last year. These companies, and Google, map everything using LIDAR-equipped cars to record everything from the height of traffic lights to the exact location of curbs to the centimeter.
Mobileye relies upon artificial intelligence. Its cameras and systems use machine learning to process information about their surroundings and make complex, nearly instantaneous driving decisions just like human drivers. The system is constantly learning and improving, and cameras are far cheaper than LIDAR—that spinning bucket thing on top of Google’s cars costs about $80,000 right now.
BMW says its autonomous car will debut in geo-fenced areas—thoroughly mapped sites with defined boundaries. That’s more than a safety move. It allows the company to proceed in an uncertain regulatory environment while waiting for consumers to grow comfortable with the idea of robots doing the driving. Geo-fencing is a “scalpel for carving away the tricky areas,” says Edwin Olson, a researcher with the University of Michigan who works on Toyota’s autonomous efforts.
In other words, a controlled test bed for the ultimate riding machine.