The construction of a high-level autonomous driving demonstration zone with cloud-connected control has begun, ushering in the era of driverless driving

Beijing Economic and Technological Development Zone will build a networked cloud-controlled high-level autonomous driving demonstration zone by the end of this year, which means that the development of intelligent connected vehicles has entered a new stage.

At the 2020 Zhongguancun Forum theme event "Beijing High-level Autonomous Driving Demonstration Zone Launch Conference", Kong Lei, member of the Beijing Economic and Technological Development Zone Working Committee and deputy director of the Management Committee, released the construction plan for Beijing's high-level autonomous driving demonstration zone.

After comprehensive research, Beijing has chosen the economic and technological development zone with the most suitable resources for the development of the intelligent connected industry.

Image source: auto.hexun

The Economic and Technological Development Zone has a special administrative system, has full authorization from the municipal government for trials first, and has strong government resource integration capabilities; the Economic and Technological Development Zone has the country's leading innovation resource advantages, with industry institutions such as the National New Energy Vehicle Technology Innovation Center authorized by the Ministry of Science and Technology, the China Automotive Intelligent Connected Vehicle Innovation Center authorized by the Ministry of Industry and Information Technology, and the Beijing Intelligent Vehicle Industry Innovation Center authorized by Beijing, and has outstanding innovation resource advantages.

The economic development zone also has superior location conditions. It is not in the city center, but has a complete urban form, a concentration of scientific and industrial talents, and a high degree of social acceptance and tolerance for new things.

The demonstration zone will be built around the Beijing Economic and Technological Development Zone (60 square kilometers). By 2022, the construction of the five major systems of "smart roads, smart cars, real-time cloud, reliable network and accurate map" will be completed, the key technical and management links of networked cloud-controlled autonomous driving will be connected, a city-level engineering test platform will be formed, and a series of application scenarios will be commercialized and a batch of intermediate products will be promoted and applied.

Yang Diange, professor at the School of Automotive Engineering and Transportation at Tsinghua University, introduced at the press conference that this networked cloud-controlled high-level autonomous driving demonstration zone in Beijing Economic and Technological Development Zone can not only target single intelligent vehicles, but also low-level autonomous driving vehicles. These vehicles can achieve high-level autonomous driving once they enter the demonstration zone.

As for autonomous driving, it is the product of the deep integration of electronics, communications, artificial intelligence, big data and the automotive industry. It is the commanding heights that countries are currently vying for in industrial development.

Domestic and foreign autonomous driving companies mainly adopt the single-vehicle intelligent technology route, but due to insufficient single-vehicle perception and high production costs, the single-vehicle intelligent technology route cannot achieve large-scale operation of L4 and above autonomous driving vehicles in the short term.

In recent years, China has proposed a technology route for connected autonomous driving. At the same time, the European Union has released a technology roadmap for connected autonomous driving. The U.S. state of Michigan has begun building its first connected autonomous driving corridor. Well-known vehicle manufacturers such as Audi have begun to deploy V2X vehicle-road collaboration-related functions in their products.

In this regard, after in-depth research, Beijing has decided to build a high-level autonomous driving demonstration zone to accelerate the large-scale operation of L4 and above high-level autonomous driving. The autonomous driving demonstration zone will be promoted step by step according to the steps of stage 1.0 (test environment construction), stage 2.0 (small-scale deployment), stage 3.0 (large-scale deployment and scenario expansion), and stage 4.0 (promotion and scenario optimization). After forming a mature model, it will be gradually replicated and promoted to other areas in Beijing.

Currently, Beijing has started the 1.0 phase of construction, deploying intelligent infrastructure for 10 kilometers of urban roads, 10 kilometers of expressways and 1 AVP parking lot.

According to the technical standards set by the Society of Automotive Engineers, autonomous driving is divided into multiple levels, from L1 to L5. According to the construction plan of the Internet-connected cloud-controlled high-level autonomous driving demonstration zone, the demonstration zone will support high-level autonomous driving above L4, while being backward compatible with the test operation and vehicle networking application scenarios of low-level autonomous driving vehicles.

It is reported that the construction of the demonstration zone in the future will adopt the approach of "small steps and fast progress, iterative improvement", with an iteration cycle of 3 to 6 months, constantly revising and improving the subsequent construction methods and content, and advancing step by step according to the steps of test environment construction, small-scale deployment, large-scale deployment and scenario expansion, promotion and scenario optimization. After a mature model is formed, it will be gradually replicated and promoted to other areas of Beijing.

Kong Lei introduced that the Internet-connected cloud control is a higher stage of the development of intelligent Internet-connected vehicles, that is, vehicles can not only perceive information, but also use cloud data to make decisions, judgments and controls. In the future, the Internet-connected cloud control automatic driving system will gradually be integrated into the system of smart city construction.

Autonomous driving is the product of the deep integration of electronics, communications, artificial intelligence, big data and the automotive industry. It is the commanding heights that all countries are currently vying to seize in industrial development.

At present, domestic and foreign autonomous driving companies mainly adopt the single-vehicle intelligent technology route. Due to various factors, the single-vehicle intelligent technology route cannot achieve large-scale operation of L4 and above autonomous driving vehicles in the short term. First, there are deficiencies in single-vehicle perception. It is impossible to achieve "beyond visual range" and solve the "visual blind spot". Second, the production cost remains high and is not enough to support the commercial scale production of autonomous driving vehicles. Third, it is difficult to solve the long-tail challenge. Extreme scenarios and other problems require a lot of tests and real scenarios to solve, which is difficult to achieve in reality. Fourth, group intelligence planning cannot be coordinated. Single-vehicle intelligence cannot achieve the optimal group traffic or regional traffic.

In recent years, my country has proposed a technology route for autonomous driving based on networking, which has gradually been widely recognized by the international community. At the same time, the European Union has released a technology roadmap for connected autonomous driving, and Michigan, the United States, has begun to build the first connected autonomous driving corridor. Well-known vehicle companies such as Audi have begun to deploy V2X vehicle-road collaboration related functions in their products, and the industry consensus is constantly building. But overall, there are still many difficulties in the implementation of autonomous driving based on networking: First, in terms of technology, key links such as high-reliability, low-latency communication technology, and algorithms for integrated perception and decision-making have not been fully resolved. Second, in terms of management, the decentralized and isolated management system does not match the needs of integrated road digital facilities. Third, in terms of business model, the traditional government infrastructure model is difficult to sustain. Fourth, in terms of industry promotion, no company can take on everything, and it is far from enough to support according to the traditional industry support rules.

After in-depth research, Beijing has decided to build a high-level autonomous driving demonstration zone to accelerate the large-scale operation of L4 and above high-level autonomous driving. On the basis of supporting the continued iteration and improvement of single-vehicle intelligence, Beijing will focus on building the world's first networked cloud-controlled high-level autonomous driving demonstration zone. Through the deep integration of vehicle and network, vehicle-road collaboration can be achieved, the role of network terminals can be played, vehicle-side costs can be reduced, and automotive product innovation and industrial transformation can be promoted.