Open search

SAE level-4 vehicles

Chinese cities grant licenses to self-driving cars

Beijing, Guangzhou, Chongqing, and now also Changsha have issued plates to autonomous vehicles. These cars will be tested in the mentioned Chinese megacities under real traffic conditions.

In Changsha, 49 vehicles will drive soon autonomously (Photo: Apollo/Baidu)

Recently, Chinese authorities granted five companies with 49 licenses to allow road tests for autonomous vehicles (AV) in Changsha. Baidu secured 45 of the licenses, bringing its total number of licenses for road testing to more than 100, over half of the 183 licenses total granted in the People’s Republic. Two self-driving Daimler trucks and one commercial vehicle from the Inceptio startup got also permits. The megacity has equipped 135 km of roads with electronics navigating vehicles. Baidu’s Apollo vehicles are equipped with CAN and Ethernet in-vehicle networks. The connected networked ECUs (electronic control units) implements the ROS (Robot Operating System), which is an open-source operating system originally developed for industrial and service robots.

A few days earlier, Guangzhou issued 24 licenses to companies offering autonomous vehicles. This includes WeRide,, and AutoX. WeRide is a joint venture company with China Unicom, the world’s fourth largest mobile service provider. The company got 20 license plates. The Operation Center of Guangzhou Intelligent Connected Vehicle Demonstration Area is the official organization in charge of the road test and licensing in Guangzhou. By referring to test procedures for autonomous driving function of connected vehicle issued by the National Technical Committee of Auto Standardization, the Operation Center arranged 13 items covering 30 testing scenarios to be in line with the national standard. Guangzhou local government encourages and supports testing of autonomous vehicle with great efforts. "Both autonomous driving technologies and commercialization are equally important to us. In the future, WeRide will continue to apply for more licenses of road test and apply them onto different autonomous vehicle models aiming at providing safer and better smart mobility experience for citizens in the future." said Tony Han, CEO of WeRide.

WeRide ranks second in China with 20 road-test license plates in Guangzhou (Photo: WeRide)

In Shanghai, car plates for self-driving cars have been given already last year. Most of these self-driving cars are equipped with multiple CAN-based in-vehicle networks. Ethernet-based networks are used for navigation and sensor-fusion purposes. First reports said that the 56 vehicles in Beijing already drove more than 100 000 km on public streets in 2018. Based on data got from monitors installed in self-driving cars, there were described 23 situations that led to autonomous vehicle disengagements, requiring human drivers to take over. Baidu’s 45 Apollo self-driving cars run about 85000 km on Beijing’s streets. Besides the Baidu cars, there are also self-driving vehicles from Baic, BJEV, Daimler, Didi Chuxing, Nio,, and Tencent, under test in the Chinese capital.

Apollo Lite

Baidu, the Chinese Google, received its initial permit to test its self-driving cars in California from the state’s Department of Motor Vehicles in 2016 and has been road testing its fleet ever since. Beginning of this year, the company has launched the Apollo 3.5 open source driverless car platform. Recently, Baidu has unveiled the Apollo Lite, a vision-based autonomous driving solution that uses ten cameras. The vehicle achieves SAE level-4 autonomous driving, meaning the driver is not required to intervene with driving operations, and that autonomous travel can take place under specific conditions, such as in designated self-driving zones.

Without using Lidar technology, the Apollo Lite can detect objects about 200 meters away. Vehicles equipped with the system have completed Level-4 autonomous driving tests on some of Beijing’s open roads. The company said via its WeChat account, that Apollo Lite’s performance is comparable with that of the pure vision-based sensing solution of Mobileye.


Publish date