Tesla Rashomon: How far is it from true autonomous driving?

Most American and Chinese automakers have chosen the Internet approach. Google's autonomous driving technology has been adopted by many automakers, and Apple is actively cooperating with Mercedes-Benz, Ferrari, Infiniti, Volvo and other automakers. Mainstream automakers are introducing the iOS operating system into new models. 

In China, Baidu has invested heavily in the research and development of driverless technology since 2015, and has obtained 2,900 patents. It has formed three open platforms: autonomous driving, vehicle-road collaboration, and intelligent vehicle networking. In the 2019 Autonomous Driving Takeover Report released by the California Department of Motor Vehicles, Baidu ranked first in the number of disengagements, surpassing the previous year's champion Google Waymo. The other two Chinese companies, AutoX and PonyAI, ranked fourth and fifth respectively. On June 27, 2020, Didi and CCTV jointly broadcast the test ride of autonomous driving online car-hailing, and announced the opening of autonomous driving taxi services to the public in Shanghai. 

Market analysts believe that by 2030, all taxis in industrialized countries will be automated. By 2040, shared self-driving cars will dominate. By around 2050, most human-driven cars will disappear. 

People see the autonomous driving revolution as a repeat of the automobile revolution in the 20th century. In the 1920s, cars began to flood into cities, and then all motorization was completed. A hundred years later, history is repeating itself. Just as the automobile industry changed history, autonomous driving will also change the present. 

Autonomous driving will overturn the original automotive industry chain, cross-border competitors will enter the market, the top of the value chain will shift from traditional OEMs to new technology elites, and travel, vehicle sales models and value distribution patterns will be redefined. Cars will become large-scale mobile intelligent terminals, and the core components of cars will shift from transmission systems that reflect power and control to intelligent software systems (algorithms) and processor chips that reflect the level of autonomous driving technology, realizing software-defined cars. 

In the last century, the automobile industry made the United States a "nation on wheels"; today, which countries will autonomous driving create opportunities for?   

4. New infrastructure may help China “change lanes and overtake” 

The 2021 Two Sessions have just concluded. Compared with previous years, this year's Two Sessions saw more proposals on "autonomous driving", "intelligent networking", "automotive chips", etc. The proposals of Baidu's Robin Li, 360's Zhou Hongyi, NetEase's Ding Lei, Great Wall Motors' Wang Fengying, Hozon Auto's Fang Yunzhou, Chery Automobile's Yin Tongyue, GAC's Zeng Qinghong, SAIC's Chen Hong, JAC's Zhou Fukang and others are all related to autonomous driving. How to commercialize autonomous driving in advance and realize intelligent transportation and smart cities under the background of "new infrastructure" is one of the focuses of this session of the Two Sessions. 

There are three mainstream technical routes for the development of autonomous driving: 1. The "Google-style" single-vehicle intelligent route represented by lidar and high-precision maps 2. The "Tesla-style" single-vehicle intelligent route represented by visual perception and shadow mode 3. The "new infrastructure" type of smart transportation route that takes the lead in networking and vehicle-road collaboration Traditional automobile powers such as Germany and Japan have steadily advanced automation upgrades with their accumulated automobile manufacturing capabilities. The United States has carried out cutting-edge exploration with its strong basic scientific research strength and world-leading artificial intelligence technology. In contrast, China's advantages in promoting autonomous driving are reflected on another level. 

With the vigorous promotion of the country's new infrastructure, Chinese technology companies are focusing on the third technical route, which is more inclined to "intelligent cars with vehicle-road collaboration". That is, through the combination of sensors with drivers, other vehicles, road infrastructure and cloud communications, massive information interaction between vehicles and roads is achieved, forming a high degree of collaboration between "people-vehicle-road-cloud". 

In 2019, the China Highway Society's Autonomous Driving Working Committee and Autonomous Driving Standardization Working Committee released the "Intelligent Connected Road System Classification Definition and Interpretation Report", which clearly defined China's road intelligence classification standards and provided clear guidance for the construction of smart roads. In April 2020, the National Development and Reform Commission clarified the scope of new infrastructure, including 5G, the Internet of Things, satellite Internet, artificial intelligence, data centers, and intelligent transportation infrastructure. Autonomous driving is both the intersection of these fields and a specific application scenario. 

The battle over mobile communication technology standards behind autonomous driving is also a battle over the dominance of interest distribution between countries. The United States is a global leader in artificial intelligence, with strong basic scientific research capabilities, and the technology accumulation advantages of technology giants are obvious. It is difficult for China to catch up with the track of single-vehicle intelligence alone, but it is expected to achieve "lane-changing overtaking" in the field of autonomous driving through vehicle-road collaboration. (See the previous article "Tesla's Chinese "Harvester") 

China's communications companies, represented by Huawei, are world leaders in 5G technology, and have a large number of 4G and 5G base stations with wide coverage. The Chinese government has vigorously promoted the construction of new infrastructure such as 5G networks, the Internet of Things, satellite Internet, data centers, and intelligent transportation infrastructure. In comparison, the construction of new infrastructure in the United States is generally market-driven and lacks government leadership, and the progress is significantly slower than that in China; in addition, the public pays more attention to personal privacy, which may also lead to a slow progress in promoting networking. 

The gradual development of the new infrastructure blueprint has created a rare strategic opportunity for my country's autonomous driving. This is a social change in which traditional OEMs, new car-making forces, Internet giants, ICT companies, and parts suppliers are deeply involved. Driven by the new infrastructure, vehicle-road collaboration is expected to enter the fast lane, making up for China's shortcomings in the development of single-vehicle intelligence, and is expected to take the lead in realizing the large-scale commercialization of L4-L5 high-level autonomous driving technology.  

References:

[1] John Markov, A Brief History of Artificial Intelligence, 2017

[2] Kai-Fu Lee, Artificial Intelligence, 2017

[3] Qiu Zeqi, Analysis of Social Action Subjects in Autonomous Driving [J/OL]. People's Forum Academic Frontier, 2021-03-15

[4] Deloitte report, Autonomous driving under new infrastructure: the battle between single-vehicle intelligence and vehicle-road collaboration, 2021

[5] Ma Naiduo, Wei Yawen, Zhuang Mengmeng, Li Longhui, Research on the development trend of autonomous driving policies at home and abroad, 2020

[6] Li Qiuwei and Shen Tong, Analysis of the Current Status and Trends of Autonomous Driving Vehicle Development in Foreign Countries, 2020