In July last year, Guangzhou officially announced the launch of the pilot of autonomous driving mixed driving. Nowadays, in the pilot areas of road testing of intelligent and connected vehicles in Guangzhou, such as Nansha District and Huangpu District, self-driving taxis drive on the road every day, picking up and dropping off passengers in a hurry. With the development of artificial intelligence technology, scenes that could only appear in science fiction movies have entered the daily life of the neighborhood.
At the same time, the situation that artificial intelligence technology is contrary to social ethics has also attracted attention from all walks of life. People are worried: Who is responsible for the accident of a self-driving car? Is it the company to which the vehicle belongs? Or an AI system designer? Or a manufacturer of in-vehicle sensing equipment?
In the face of the new situation, Guangzhou dared to be the first, and decided to establish systems such as operation access, operation supervision, accident liability delineation and other systems and unified monitoring platforms for autonomous vehicles on the premise of maximizing public safety, to ensure road traffic safety, personal and property safety, etc., to promote insurance institutions to set up new insurance for autonomous vehicles, giving priority to protecting public rights and interests; and setting up an autonomous vehicle application demonstration operation guarantee fund to provide assistance to the injured party and effectively control operational risks.
City of Smart Cars:
Guangzhou intelligent network vehicles the whole industry chain ecological layout of the country leading
China is a big country in transportation, in order to achieve the goal of moving towards a strong transportation country, information and intelligent means will play an important role, so the Internet of Vehicles has become an important starting point to support the construction of a strong transportation country. In August 2020, Guangzhou put forward the goal of "trillion strategy to build a city of smart cars", and autonomous driving became a new business card of Guangzhou, and quietly changed the daily life of residents.
In 2020, under the influence of the new crown pneumonia epidemic, Guangzhou's automobile production still reached 2.95 million units, an increase of 1.4% year-on-year, maintaining the position of the first city in the country's automobile production for two consecutive years; achieving an output value of 586 billion yuan, an increase of 3.8% year-on-year, accounting for about 30% of the city's total industrial output value. Surprisingly, in the context of a series of "gray rhinos" in the global automobile industry in 2021, according to the Guangzhou Municipal Bureau of Statistics, from January to November 2021, Guangzhou's automobile production reached 2.5997 million units, of which the output of new energy vehicles increased by 81.9% year-on-year.
Since the reform and opening up, Guangzhou has focused on creating the city business card of "Guangzhou Automobile", creating a pioneering advantage for the ecology of the intelligent network industry, and becoming the ecological agglomeration of the whole industrial chain of "vehicle + network technology + automotive electronics" with the most competitive advantage in China. Gao Yuyue, deputy secretary general of the Guangzhou Municipal Government, first-level inspector and secretary of the party group of the Municipal Bureau of Industry and Information Technology, introduced at the Guangbao Center on July 14 last year that Guangzhou has basically formed a whole industrial chain layout in the fields of vehicles, parts and supporting facilities, automotive electronics, new energy systems, and intelligent network solutions.
"My family lives in Science City in Huangpu District, and in the past two years, I have seen self-driving taxis driving on the road every day. I experienced it with my family the other day, it felt great, safety never pressed the line, after arriving at the destination, the vehicle automatically pulled over and the passengers automatically got off the bus. Ms. Fan, a citizen, said, "Scenes that originally only appeared in science fiction movies have come to life." ”
Autonomous vehicles photographed on Bio Island
Journalist Experience:
Intelligent connected cars are becoming more and more like "old drivers"
Starting from the Guangbao Center, the reporter took the intelligent networked car of Xiaoma Zhixing to Yuejiang Road and experienced the ride test of the main road in the central urban area of Guangzhou for the first time. The reporter's experience found that from the Canton Tower to the Pazhou Bridge, intelligent networked cars can automatically lead passengers to enjoy the scenery on both sides of the river. The speed is about 30 km to 40 km / h, and there is no speeding during the whole process. If the leftmost lane is a U-turn lane, the vehicle will automatically switch lanes, but the line will not be compacted. The vehicle automatically recognizes the traffic lights and starts the vehicle or brakes to stop.
In the event of a sudden acceleration of the vehicle on the right and the possibility of occupying the forward lane, the smart connected car will slightly deviate to the left, but will not switch lanes, and automatically decelerate to maintain a safe distance. The staff told reporters that this is the result of artificial intelligence learning, and intelligent networked cars are becoming more and more like "old drivers". In the past, when encountering the situation of speeding and congestion of the right vehicle, the intelligent networked car may brake sharply on its own, but after the big data learning of artificial intelligence, it will be determined that the vehicle will not be jammed, so it is slightly offset to the left side of the same lane and maintain a safe distance.
Several self-driving cars were displayed in front of the Guangbao Center
How to deal with autonomous driving accidents?
Guangzhou: Establish a new type of insurance to give priority to the protection of public rights and interests
The most discussed area of AI ethics is autonomous driving. On the evening of March 18, 2018, in Tempe, Arizona, a pedestrian named Elaine Herzberg was hit by a self-driving car being tested and died of his injuries. There was a driver in the car involved, but the vehicle was controlled by a fully self-driving system (artificial intelligence). The incident became the world's first self-driving car collision and death case, prompting the global public to think about the ethical, moral and legal issues of artificial intelligence, and the repeated question is: Who is responsible for the incident, the testing department of the self-driving car company, or the designer of the artificial intelligence system, or even the manufacturer of airborne sensing equipment?
Guangzhou has given its own direction of exploration. According to the "Guiding Opinions of Guangzhou Municipality on the Work Related to the Road Testing of Intelligent And Connected Vehicles", it clarifies the organization and management, application conditions, daily supervision, test types, accident liability determination and other matters of the road test of intelligent and connected vehicles in Guangzhou, and requires enterprises to purchase the necessary commercial insurance for each test vehicle.
According to reports, Guangzhou has promoted the pilot of mixed traffic on the premise of maximizing public safety. Construct a technical safety standard system, network security system, and operation safety guarantee system for autonomous vehicles as a whole, establish systems for operation access, operation supervision, accident liability delineation, and a unified monitoring platform for autonomous vehicles to ensure road traffic safety, personal and property safety, and other legitimate rights and interests. Promote insurance institutions to set up new types of insurance for autonomous vehicles, giving priority to protecting public rights and interests; and set up a demonstration operation guarantee fund for autonomous vehicle applications to provide assistance to injured parties and effectively control operational risks.
Mr. He, an artificial intelligence engineer at Aerospace Cloud Network, told reporters that intelligent networked cars with artificial intelligence are getting smarter and smarter, which lies in smart learning, and its process is actually the collection and analysis processing process of big data.
Keeping the impact of the accident to a minimum is also Guangzhou's goal. According to reports, after Guangzhou carries out the mixed traffic pilot, 12 municipal departments such as industry and information technology, transportation, public security, justice, housing construction, urban management, emergency response, etc. will clarify the division of responsibilities, define the work authority, and form a work synergy from top to bottom. The application demonstration operation area should establish "three mechanisms" of rapid reporting of accidents, determination of accident responsibility, and advance compensation. If an accident occurs, the operator should report within 15 minutes, and the third-party agency should submit a written report within 3 hours; the traffic police department of the jurisdiction shall handle the road traffic accident of the intelligent networked vehicle (automatic driving) application demonstration operation in accordance with the law; the insurance company, the party responsible for the accident and the application demonstration operation guarantee fund shall pay the corresponding compensation for the accident, establish a set of rapid emergency response mechanism, and strive to control the impact of the accident to the minimum extent and maximize the control of large-scale demonstration application risks.
The backstage of self-driving cars is also inspected by personnel
Industry Trends:
It is expected that the production of intelligent and connected vehicles will exceed 4 million by 2025
In a 2018 report, McKinsey predicted that based on estimates of the cost curve of the underlying technology of autonomous driving, 2025-2027 will be the inflection point of autonomous driving, because then it will be the economic parity point between autonomous driving and human driving. This means that the total cost per kilometer of autonomous driving will be roughly the same as the cost of a driver driving a conventional car.
The future commercial model of autonomous driving is roughly like this: the vehicle company is responsible for mass production, the solution provider provides technical services, the service operator builds a big data platform, is responsible for the management and operation of the fleet, and provides services for end users. Among them, service operators will implement asset-light operations, which will be financially supported by financial institutions.
The reporter learned that it is expected that by 2025, guangzhou's vehicle production capacity will reach 5 million, of which the output of intelligent networked vehicles will exceed 80%. According to the "Opinions of the General Office of the Guangzhou Municipal People's Government on Promoting the Accelerated Development of the Automobile Industry", by 2025, the assembly rate of new autonomous vehicles below the L3 (inclusive) level of intelligent networked vehicles in Guangzhou will exceed 80%, and the L4 level of autonomous vehicles will be industrialized. Last year, JD Logistics and Xiaoma Zhixing jointly launched the first L4 level autonomous driving heavy truck trunk logistics demonstration operation line in China. This is also the first city in Guangzhou to land an L4 level autonomous driving heavy truck trunk logistics demonstration operation line after achieving eight national firsts in the field of automatic driving.
According to the plan, Guangzhou will carry out mixed traffic pilots in the pilot areas of road testing of intelligent networked vehicles in Guangzhou, such as Nansha District and Huangpu District, and the specific operating scope (including highway types) in the area will be delimited by the pilot district government. In the early stage of the pilot, it mainly supports the development of passenger transport activities such as cruise taxis, online ride-hailing vehicles, buses and other passenger activities and general road freight transport, and in the later stage, according to the technical development, the municipal pilot office will review and successively open application scenarios such as urban intelligent sanitation, express delivery, and postal services.
Smart sanitation vehicles
Guangzhou exploration in the era of smart cars
1. "Full coverage, tight coupling" to create an intelligent networked automobile industry ecology
In 2020, under the guidance of the Ministry of Industry and Information Technology and the Provincial Department of Industry and Information Technology, Guangzhou, together with Shenzhen and Huizhou, will establish a regional coordination mechanism for the three cities of Guangzhou, Shenzhen and Huizhou, form a leading group for the construction of intelligent and connected vehicle clusters in the three cities, and the three cities will work closely together to jointly promote the development of the intelligent networked vehicle industry cluster, and actively create an intelligent networked automobile industry ecology from vehicle manufacturers to autonomous driving head enterprises, from testing to demonstration operation, from hardware to software, from smart cars to smart cities, "full coverage, tight coupling". Strive to become the most competitive advantage of the intelligent networked automobile industry agglomeration in China.
2020 has thus become the first year of the development of Guangzhou's "smart car", and vehicle companies have launched intelligent networked vehicles with core competitiveness. Today, Guangzhou's intelligent network technology, products, and enterprise achievements are constantly emerging, and at the same time, the ecosystem of coordinated development of intelligent networked vehicles and smart cities has gradually improved.
In April 2020, GAC Group and Huawei independently developed the world's first 5G+V2X intelligent communication system, integrated Huawei's latest generation of 5G chips, and created the first mass-produced 5G smart car "Aian V" in China, and launched the ADiGO 3.0 automatic driving system based on high-precision maps.
Dongfeng Nissan has gradually introduced technology and 10 Nissan Zhixing technology models from the fields of intelligent driving, intelligent power, and intelligent interconnection, and completed the network layout of all main models in 2020. As of July 2021, the number of users connected to the vehicle connection system has exceeded 2 million, and Xiaopeng Automobile successfully launched the P7 intelligent networked car with NGP automatic navigation assistance driving system, with sales exceeding 15,000 units in that year.
Beidou satellites have become the "clairvoyance" guiding intelligent networked vehicles. According to reports, the number of satellite navigation and positioning enterprises in Guangzhou accounts for about 20% of the total number of similar enterprises in the country, of which Taidou Microelectronics is the first manufacturer in China to integrate the "three-in-one" solution of radio frequency, baseband and flash memory; Guangdong Xingyu Technology is committed to building a "cloud-network-terminal" trinity of accurate spatio-temporal platform and precise spatio-temporal ecosystem, and its Beidou high-precision positioning network has achieved 90% of the country's population area coverage.
In addition, Guangzhou Gaoxin Group successfully launched the world's first batch of 5G+V2X vehicle terminals and vehicle scale modules for automotive front and rear installation.
2. Enter the main roads of the central urban area and take the lead in carrying out road tests
2020 is the first year of Guangzhou's smart transportation construction. Huangpu District to build the country's first 5G automatic driving application demonstration island, as the pilot area of intelligent networked vehicle testing, Huangpu District has issued the "Guangzhou Huangpu District Guangzhou Development Zone Intelligent NetworkEd Vehicle Road Test Implementation Rules", initially formed a new type of intelligent network industry ecology of various market entities mutual integration and symbiosis, division of labor and cooperation, benefit sharing, has been opened in the biological island 6 primary roads, 11 secondary roads, 5 tertiary roads a total of 21.894 kilometers.
Nansha District has the most open test sections in the city. At present, Nansha District has accumulated 90.7 kilometers of open test roads, accounting for about 58% of the total mileage of open test roads in the city, ranking first in the city. The basic platform for intelligent driving IoT perception has been built at the tip of Lingshan Island in nansha pearl bay area, which can fully perceive pedestrians, vehicles, road conditions, weather information, etc. on the road. The next step will focus on the opening of the test road for intelligent and connected vehicles and the construction of a closed test base for Qingsheng automatic driving.
Guangzhou thus became one of the first cities in China to carry out road testing of self-driving cars. According to reports, after the first batch of 24 intelligent networked vehicle road test notices were issued in Guangzhou in 2019, 59 test notices were issued successively. Last year, for the first time, an 11-kilometer test road was opened on the main road in the central urban area (Haizhu District), supporting companies to carry out road testing work in a broader and more complex road environment. At present, a total of 88 first-, second- and third-level test roads have been released and opened, with a mileage of about 156 kilometers.
Since 2020, Guangzhou has continuously innovated and made breakthroughs with the goal of automatic driving application demonstration, forming eight national firsts. The first to approve the 5G remote driving test; the first to recognize the road test license of intelligent and connected vehicles in other regions; the first to issue passenger test licenses; the first to carry out Robotaxi technology verification in batches; the first to carry out road tests on the main roads in the central urban area; the first autonomous driving R&D enterprise to obtain a license for the operation of the online ride-hailing platform; the first autonomous driving R&D enterprise in the jurisdiction to create a standardized process for the mass production of L4 autonomous driving; and the first to issue a comprehensive specification for the city-level vehicle networking pilot zone.
According to the data, in 2020, there will be 5 autonomous driving enterprises in Guangzhou, 71 autonomous vehicles will carry out automatic driving tests in Sui, with a cumulative test mileage of nearly 700,000 kilometers, more than 60,000 passengers, zero active safety accidents, and basically have the conditions for demonstration operation.
3. Blueprint for the future Complete large-scale urban transportation tests in five stages
The reporter learned that in the next step, Guangzhou will support enterprises to carry out various types of open road testing and demonstration applications on a larger scale, and strive to achieve the demonstration deployment and operation of intelligent networked vehicles in Guangzhou. By 2025, Guangzhou will complete multi-dimensional, comprehensive and large-scale urban transportation experiments in five stages with different mixed traffic environment proportions, different degrees of vehicle-road collaboration, and a variety of new travel services. On the basis of 180 days of no active safety accidents in each stage of autonomous vehicle demonstration application, and on the basis of assessment and acceptance by the municipal special working group, we will further increase the amount of autonomous vehicles, carry out a new stage of automatic driving mixed tests, and steadily promote the mixed driving pilot.
Following the pilot project of mixed traffic in Nansha District, Huangpu District and other pilot areas of guangzhou intelligent networked vehicle road test, guangzhou approved the construction of intelligent networked vehicle pilot area in Huadu District after joint deliberation by relevant member units of the Municipal Intelligent Networked Vehicle Road Test Office.
At the same time, Guangzhou will build a pilot project of Pazhou Car City Network. Pazhou Checheng Network Pilot Project is a demonstration project for the coordinated development of Chezhou, which will provide advanced means for urban governance at the two levels in urban areas in terms of transportation, community and safety through the upgrading and integration of urban infrastructure, the construction of the "Checheng Network" platform, the construction and operation of the Checheng integration demonstration application and the formulation of relevant technical and engineering standards, and create a "new city construction Pazhou model". It is reported that the Pazhou "Checheng Network" project involves about 30.11 kilometers of roads, covering an area of about 10 square kilometers, and building "one network, one platform, N applications". "One network" refers to the construction of Pazhou global Checheng perception network, focusing on Pazhou Convention and Exhibition Center Area and Internet Innovation Agglomeration Area, integrating infrastructure such as 5G, smart light poles, traffic management, and CIM basic data in Guangzhou to form Pazhou Checheng Perception Network; "one platform" refers to the construction of "Checheng Network" platform, the construction of a "Checheng Network" platform with unified standards, logical collaboration, open source and openness, and support for multi-class applications and city-level data processing; "N application" refers to exploring and building a variety of Auto city integration applications. Based on the empowerment of Pazhou Checheng Perception Network and the "Checheng Network" platform to multiple urban management and service fields, combined with the needs of multiple fields, the integration application of Checheng city is built. At present, we are promoting the construction of infrastructure such as smart light poles, and all applications are expected to be online by the end of December 2022.
Wu Dongsheng, senior vice president of Gaoxin Group and general manager of the intelligent networking business unit, introduced that at present, foreign countries are mostly engaged in the research and development of a single intelligent vehicle, that is, through the vehicle perception of the surrounding roads; and with the emergence of smart light poles in China, through 5G to achieve vehicle-road collaboration, not only vehicles can perceive the road, the road can also sense the vehicle, both through 5G networking, so that intelligent networked vehicles are safer.
Text/Guangzhou Daily. New Flower City reporter Liu Xing
Photo: Guangzhou Daily. New Flower City reporter Chen Youzi
Guangzhou Daily. New Flower City Editor Cai Lingyue