Article | Theory of Intelligent Relativity, Author | Chen Xuanbin

I don't know when it started, but autonomous driving seems to have suddenly awakened and frequently appeared at international events, becoming an important means for major organizers, technology manufacturers, and even countries to showcase their technological innovation capabilities.
Before and after the Tokyo Olympics, Toyota's autonomous driving successfully entered the international arena through this grand event. Toyota not only provided a large number of autonomous vehicles for the Tokyo Olympics, but also launched the commercial operation of L4 level autonomous vehicles at the Tokyo Imperial Stadium.
Time has arrived for the Beijing Winter Olympics, and autonomous driving has ushered in a comprehensive and multi-dimensional application in the venue of China. The most eye-catching thing is undoubtedly Baidu Apollo's car robot, which successfully completed an 800 meter torch relay as a torchbearer. Following closely behind, autonomous driving products from manufacturers such as Meituan, Neolithic, Zhixing, Ubisoft, No.9 Company, and BAIC Group have successively landed in Beijing Shougang Park, undertaking a series of service work such as transportation, food delivery, security, and cleaning.
Similarly, whether it is the Boao Forum already held this year or the postponed Hangzhou Asian Games, both have coincidentally made the display and service of autonomous driving technology a key focus of promotion. Domestic well-known technology manufacturers such as Wenyuan Zhixing and Geely Automobile have also seized such opportunities to enter the international stage, cooperated with event organizers, and accelerated the implementation of their own autonomous driving technology solutions.
So, behind this phenomenon, for autonomous driving, is this a pre training of "fixed point force" or a "pure muscle show"? What will the frequent appearances at international events bring to the development of autonomous driving?

01 Autonomous driving, continue "enclosure movement"

Before being fully commercialized, autonomous driving often requires extensive road testing to meet commercial requirements. Whether it is simulation training in enclosed areas or fixed-point training on open roads, autonomous driving is basically still being tested in a clearly defined area.
For autonomous driving, this development model is like a gradual "enclosure movement", starting in specific scene areas, then gradually advancing towards the periphery, and then expanding to the entire urban scene.
Looking back at the development process of the intelligent connected vehicle industry in Changsha, this context is clearer.
In 2016, Changsha began constructing a national intelligent connected vehicle (Changsha) testing area, which was completed and put into use in 2018. It mainly provides closed testing scenarios for major autonomous driving manufacturers and various types of autonomous driving vehicles, such as urban simulation testing area, high-speed simulation testing area, rural simulation testing area, off-road simulation testing area, and so on.
In 2019, Changsha launched the construction of two "100" projects, namely the "100 kilometer smart expressway" and the "100 square kilometer urban open road". As a result, autonomous driving entered the testing phase of urban and regional open roads in Changsha. Subsequently, the total mileage of the test road gradually increased to 175 kilometers, covering both urban and intercity areas.
During this process, Baidu's autonomous driving travel service platform, Radish Run, conducted trial operations within the open road range of Changsha, gradually covering the 135 kilometers of roads in Changsha from the initial 50 kilometers, gradually expanding the regional scope step by step. Now, Carrot Run has started providing autonomous driving services in cities such as Beijing, Shanghai, Guangzhou, Shenzhen, and Wuhan.
Similarly, the autonomous driving testing areas in Guangzhou are gradually opening up, covering six urban areas including Baiyun, Haizhu, Panyu, Huangpu, Huadu, and Nansha. A total of 202 first, second, and third level testing roads (sections) have been opened for well-known autonomous driving manufacturers such as Wenyuan Zhixing and Xiaoma Zhixing to use for passenger and load testing training.
The reason why such a situation exists is that, on the one hand, autonomous driving itself has a certain degree of danger. The division and limitations of specific testing areas are easy to manage, ensuring the personal safety of urban residents as much as possible, and meeting the practical testing environment required for autonomous driving.
On the other hand, the "vehicle road collaboration" model favored by most domestic autonomous driving manufacturers currently has a crucial technical prerequisite - the intelligent transformation of roads and the laying of communication infrastructure, thereby forming a synergy between vehicles and the traffic network system, in order to further achieve the effect of autonomous driving.
Therefore, for current autonomous driving, whether from the perspective of administrative management or technological development, the best mode is basically to "circle" in a specific area for testing, training, and iterative upgrading.

02 is another "new infrastructure"

From a technical perspective, this "enclosure movement" is more necessary for the development of the "vehicle road coordination" mode of autonomous driving.
In the future development path of autonomous driving, road end transformation and support will only become increasingly important. As autonomous driving continues to develop, the "perceptual limitations" exposed by bicycle intelligence are becoming increasingly apparent, which further drives technology manufacturers and local governments to seek optimization ideas for the "vehicle road collaboration" model.
Compared to bicycle intelligence, in the vehicle road collaboration mode, through the transformation and upgrading of the road end, it is often possible to achieve beyond line of sight and more accurate perception and recognition, thereby ensuring the safety level of vehicle driving.
At present, in key scenarios such as urban intersections or highways, the installation of LiDAR and other road end sensing devices can achieve full dimensional restoration and real-time monitoring of road segment scenes, effectively improving the traffic management efficiency of specific road segments.
So, once these road end perception and recognition information are fed back to the vehicle end, autonomous driving can provide good safety feedback and ensure normal driving. This is where the application advantages of vehicle road collaboration lie.
However, such a path also means more investment, especially in the intelligent transformation of road ends and the layout of communication infrastructure, which is often indispensable.
During the Winter Olympics, in order to fully ensure the efficient and safe services of autonomous driving, China Unicom deployed a total of 17 5G base stations, 1 Beidou ground-based enhancement station, 30 5G roadside terminals, 20 C-V2X roadside units RSUs, 55 roadside edge processing units, 40 camera heads, 15 microwave radars, and 3 lightning vision all-in-one machines in Shougang Park, covering the main roads of the entire Shougang Park, Only then can we bring a stunning demonstration of autonomous driving technology to the world.
Therefore, in the "vehicle road collaboration" mode, the development of autonomous driving cannot avoid a necessary infrastructure construction.
And such a path will inevitably limit the driving range of autonomous driving to a certain extent to only stay on the road sections that have already been opened and undergone intelligent transformation. At the same time, due to the wider connection of infrastructure construction, the overall development of autonomous driving will also slow down, and can only gradually expand with the upgrading of the road end.

03 Road construction is not just a technical issue

To develop autonomous driving, road construction is an essential technical project, but it is not just a technical issue.
Of course, the main issue is still a technical one.
The first step in road construction is to establish a technical channel for data transmission. In terms of communication technology, both China and the United States have basically defined the technical route of C-V2X as a direct communication technology for vehicle road coordination. Although highly valued, it still cannot change the predicament of C-V2X in its early stages of development, such as incomplete construction, unclear standards, and high costs.
Nowadays, China is vigorously promoting the development of the C-V2X standard and recently released three new C-V2X standards, namely the "C-V2X Scale Test Data Interface Specification", "Technical Requirements and Testing Methods for Vehicle Road Collaborative Roadside Sensing System", and "Technical Requirements and Testing Methods for Direct Communication Positioning and Synchronization of Vehicle Networking Wireless Communication Technology without GNSS Signal Based on LTE".
At the same time, the Ministry of Industry and Information Technology of China has clearly stated in the "14th Five Year Plan" for the development of the information and communication industry that by 2025, key highways and urban roads will achieve C-V2X scale coverage. On the market side, technology related manufacturers such as Huawei, ZTE, Baidu, and Gaoxing are also continuously following up to promote the research and application of C-V2X communication modules.
The comprehensive implementation of such technological paths is crucial for the development of China's autonomous driving industry.

Furthermore, this is also an engineering issue.
As a type of infrastructure project, road construction is widely related and the construction difficulty is far more complex than that of bicycle intelligence. A series of problems, such as how to repair existing urban roads, who will repair them, who will operate them after repair, and how to clarify the rights and responsibilities of operation, are common difficulties that plague the development of autonomous driving.
At present, the "iron triangle" model of "government+state-owned enterprise platform+ecological enterprise" being formed in Changsha is attempting to answer this question. That is to establish a state-owned enterprise platform between local governments and autonomous driving ecological enterprises (such as Xiangjiang Intelligent in Changsha, Hunan) to be fully responsible for infrastructure construction, demonstration scenario operation, etc.
However, this is only a model for the Changsha region. Previously, some cities had outsourced similar infrastructure projects to autonomous driving manufacturers, and Baidu Apollo had won bids for new infrastructure projects in cities such as Guangzhou, Chongqing, and Hefei.
Both different modes are currently in the validation stage. What kind of market entity should this project ultimately be entrusted to operate, and perhaps the results of current practice will ultimately be needed.

Conclusion

Overall, with the maturity of autonomous driving technology and the opening of policies, autonomous driving applications have the ability to extend to more cities and scenarios.
International events such as the Olympics and the Boao Forum have extremely high requirements for participating in autonomous driving enterprises. Being able to participate in them is not only a manifestation of technical strength, but also attracts global attention and traffic. For some autonomous driving startups, this can precisely attract the attention of the capital market, and for mature companies, it is also a rare opportunity to use technology to serve society.
In addition, from the perspective of the "vehicle road collaboration" mode, the specific areas divided by international events also have natural basic advantages for autonomous driving demonstrations, and are a rare testing scenario - autonomous driving manufacturers can conduct comprehensive verification of autonomous driving technology in a specific, complex, and standardized real environment time and time again.
Currently, there is still a considerable journey ahead for the full commercialization of autonomous driving. The "enclosure movement" of autonomous driving will continue, and more importantly, during this period, we need to collaborate to achieve policy breakthroughs.
So, before that, showcasing in the most eye-catching demonstration area (which is the venue of major international events) is also an attempt to explore future commercialization!
This article is reprinted from Titanium Media, article: Intelligent Relativity, author: Chen Xuanbin.
First image source @ Visual China
Original link：When did the "enclosure movement" of autonomous driving start- Titanium Media Official Website (tmtpost.com)