Baidu Obtains License for Testing Autonomous Vehicles in Beijing

2020-01-02

Baidu announced this week that it has obtained 40 autonomous driving licenses, allowing it to test autonomous vehicles with passengers on designated roads in Beijing. This makes Baidu one of the first companies to conduct such tests in China. So far, Baidu's autonomous vehicles have been tested in 23 cities in China, with a total mileage of 3 million kilometers.By the end of December 2019, Beijing had allowed 151 roads (with a total length of over 500 kilometers) for autonomous vehicle testing, although these roads are not open to the public during the testing process. The Chinese government has designated five levels of autonomous driving test permits, from T1 to T5. The Beijing Municipal Transportation Commission allocated all of its first batch of T4 autonomous driving test permits to Baidu in early July this year. At that time, the company pointed out that T4 is the highest level permit in China, which is an open road test permit that allows it to deploy prototype vehicles on urban roads, tunnels, school areas, and other places.



Continued Partnerships


Baidu recently established a strategic partnership with Geely to equip Geely vehicles with DuerOS for Apollo, an AI-based IoV solution that features voice assistants, augmented reality, and motion detection capabilities. Baidu has also reached an agreement with FAW to deploy autonomous vehicles in Changsha, Hunan Province in the coming months. Changsha will become a pilot city for Baidu's ongoing autonomous taxi project. The future 5G network and the intelligent road upgrade of DuerOS for Apollo (both part of the Hunan Smart City Plan) will lay the foundation for Baidu to become China's largest autonomous taxi company.



Apollo Go and Apollo 5.0


Apollo Go is the latest version of Baidu's open-source autonomous driving platform. It can dynamically calibrate vehicles at a rate of 100 vehicles per minute through cloud computing in just 30 minutes. It also comes with Baidu's valet parking framework, which is an economically efficient way to transform traditional parking facilities into intelligent infrastructure.


Apollo has grown to over 400,000 lines of code (more than twice the 165,000 lines of code announced by the company in January 2018) and is now being tested, contributed to, or deployed by well-known brands such as Intel, NVIDIA, and NXP. According to Baidu, the number of developers obtaining Apollo code from the project's GitHub repository has increased by 20% since mid-2018, reaching 12,000.


Baidu and its competitors are vying for a lucrative market. According to a McKinsey report, the value of China's autonomous driving cars and mobility services will exceed $500 billion by 2030, with approximately 8 million autonomous vehicles driving on public roads.

RECENT AI NEWS

RECENT AI TOOLS