Like in many countries, the allure of self-driving cars—and the safety, savings and efficiencies they promise—is a powerful one. That’s why, earlier this year, the Chinese government’s top economic-policy agency set a goal to have half the new cars on China’s roads to be partially or fully autonomous by the end of 2020. If done right, a fleet of fully autonomous electric cars this large could be one of the most resource-efficient systems of transport the modern world has seen.
Welcome to our field guide on China’s electric-vehicle industry. Check out other parts of our deep dive here.
China’s technology and manufacturing capabilities still have a ways to go if the country is to meet the goals it has set for itself. China doesn’t have the domestic software talent that powers the world’s leading autonomous-driving companies like Waymo, Uber, and Tesla, all based in the US. So China’s self-driving startups are setting up research offices in Silicon Valley before taking the technology back home.
Baidu, the biggest search engine in China (and second-biggest in the world, following Google), opened an office in Sunnyvale, California in 2014, with 200 engineers and scientists working on autonomous driving. Some of Baidu’s alumni have gone on to start their own companies: Pony.Ai (founded in December 2016), and WeRide (formerly known as Jingchi, founded in April 2017), were both started by former Baidu executives based in California, before moving their headquarters to China.
To develop better self-driving software, these startups need data and lots of it. They gather information from real-world road tests, as well as computer simulations, and run them through machine-learning algorithms, which translate those data into models that map the complexities of driving on a road with hundreds of changing variables. That’s why one way to evaluate how far along autonomous-vehicle startups are in their maturation is to check how many miles their self-driving software have traveled.
The leader of the pack is Waymo, the self-driving arm of Google’s parent company Alphabet, which had accumulated 10 million self-driven miles as of mid-October. Chinese startups likely have a long way to catch up. California’s Department of Motor Vehicles requires self-driving car companies to report their mileage annually; through 2017, Baidu reported 2,000 miles in California. Neither Pony.ai nor WeRide, nor NIO, another Chinese company working on self-driving cars, have reported any mileage in California to date. Waymo has begun testing its self-driving vehicles in China, and Pony.ai and WeRide both say they plan to roll out self-driving taxis in China in the near future, but none have shared any test mileage in China.
In other words, it’s unclear where these companies’ technologies stand, but if they aren’t crowing about their high mileage, it’s likely that they aren’t nearly as far as they’d like to be.
The Chinese government has given large subsidies to lower the cost of electric cars (and thus boost demand for the cleaner form of transport) in recent years. Beijing has yet to announce any similar programs specifically for self-driving cars, and one major Chinese self-driving company told Quartz on background that it has never received any subsidies or tax breaks from the government.
But if you read the tea leaves right, you can see a shift coming. In 2017, three regulatory agencies of the Chinese government announced a plan to push the country’s auto industry into manufacturing cars that are cleaner, more efficient, and internet-connected. It was Beijing’s latest effort to advance a range of technologies—including big data, the Beidou satellite system (a Global Positioning System replacement), and battery charging—that could combine to make China the world leader in autonomous vehicles.
The monetary and strategic value of cracking fully autonomous driving is huge. It’s the kind of opportunity that smartphones presented to make Apple the first $1 trillion company and, in the process, all but shut out the old guard of Nokia, Blackberry, and Motorola. What’s even more important is that, if self-driving can be made to work on chaotic Asian streets, it can be made to work anywhere in the world.