Unmanned driving on the road to Shenzhen, Shenzhen has pulled the future into reality

On August 1st, the "Regulations on the Management of the Shenzhen Special Economic Zone Intelligent Connection Automobile" was officially implemented, and Shenzhen became the first city for autonomous driving legislation. In fact, as early as the "Regulations" landed, many Shenzhen companies rubbed their hands, making some streets in Shenzhen a most concentrated experimental area for intelligent driving. The unmanned online ride Robotaxi has sprung up after the rain. How to guarantee the safety of unmanned driving? The "unicorn" company, which has been optimistic about the prospect, also admits that unmanned driving urgently needs to "return blood" in business. So how far is the unmanned driverless driver of science fiction from entering the ordinary people's home?

Shenzhen Yuan Rong Qixing, Xiaoma Zhixing, Baidu's autonomous driving and vehicle road collaborative application test bases, etc.

How to make unmanned vehicles have a "prediction" like "old driver" and can see, understand, and make correct decisions is the primary prerequisite for ensuring high -end autonomous driving safety. The team led by Yuan Rong Qixing CEO Zhou Guang was the first fusion algorithm in the industry. Zhou Guang said that the current L4 -level autonomous vehicle is equipped with sensors such as lidar, camera, and millimeter wave radar. These sensors have been perceived by a key perception fusion algorithm to form a perception of 360 degrees of surrounding environment. The vehicle is abstracting the weight of the surrounding environment, and all people and vehicles can be accurately identified. In addition to the integration perception, planning control is also a key link. Continuous game -based methods can make vehicles also highly flexible in complex urban road conditions. Out overtaking is not a problem. These are algorithms established to ensure safety.

Regarding the dispute between the route of autonomous driving technology, Zhou Guang said that there is a major disadvantage of Tesla's pure visual technology, and it may be inaccurate to some less objects. Today's perception model is a fusion perceptual model, and there is no so -called distinction between the Lord and the secondary. All the data comes in is just information. The algorithm is to use this information to learn deep learning.

To improve the safety of autonomous vehicles, in addition to software algorithms, obtaining a lot of data is also critical. At the same time, Mo Luyi, vice president of Xiaoma Zhixing, a road test in China and the United States, told us that autonomous driving is a very complicated technology. China's road conditions are much more complicated than the United States. Help the entire technology can be improved rapidly. Use the same set of software and hardware systems to test under different road environments across China to help the system improve the generalization capacity, so that when it will expand new areas in the future Finish. She also specifically mentioned that many underground tunnels in Shenzhen have several kilometers of underground waterways, and these scenes will put forward different requirements for the entire autonomous driving system. There are also weather conditions, such as the scenes of rainy days in Guangdong, and typhoons. These scenes actually make a lot of software and hardware improvements.

In the autonomous driving technology, the chip has a "central hub" position. It controls the operation of autonomous vehicles through the autonomous driving platform. How to solve the problem of "stuck neck" by foreign giants? In this regard, Mo Luyi, vice president of Pony Zhixing, said that facing the problem of stuck neck, we need to rely on independent research and development or other alternatives. At the same time, she believes that the entire chip industry in China is also booming, and in the future, Chinese chips have the ability to support complex autonomous driving algorithms.

Urban roads are more "wisdom" to accelerate unmanned driving popularization

The "Regulations of the Regulations" of Shenzhen Legislative also encouraged conditional intelligent connected automobile -related enterprises to build roads and traffic scene simulation simulation platforms to simulate testing and technical verification of autonomous driving systems. It is necessary to make the vehicle intelligent and unmanned, but also make the urban roads more intelligent and efficient, and make unmanned driving more secure in cities. In this regard, Sun Jing, general manager of Baidu Intelligent Transportation Guangdong Province, pointed out that in the end, all vehicle coordination provides drivers with better services. While it can improve the traffic efficiency of the entire road, and to ensure safety. The biggest advantage of car road collaboration is that the recognition perception of the road is more accurate, and it will support multiple scenarios such as transportation, municipalities, and road incidents. The scene -based algorithm is more accurate, making its services more targeted. In response to high -level autonomous driving vehicle collaborations, high -precision recognition and information transmission of the intersection effectively improved the assistance of autonomous driving, helping the vehicle can better get information on the road, auxiliary vehicles make decisions, and serve autonomous driving. When the road is getting smarter, the cost of autonomous vehicles can also be reduced. In response to data security issues that are concerned about, Sun Jing emphasized that the data obtained should be effectively provided to the society, but it must be based on data security. In the process of industrial development in the future, more data will be opened, but it must be empower to various ecological manufacturers under the premise of ensuring the security of national data.

As a pilot city of "Smart City Infrastructure and Smart Connected Automobile Cooperation", Shenzhen has attracted autonomous driving star companies such as Baidu Apollo, Xiaoma Zhixing, and Yuan Rongqixing in recent years. Autonomous vehicle testing and providing free to passenger experience, preparing for the upcoming official on the road. As a technology company focusing on R & D and application of L4 -level autonomous driving technology, in the past three years, Yuan Rong Qixing has accumulated over 7 million kilometers in Shenzhen, Wuhan, and Hangzhou CBD region. At the same time, Xiaoma Zhixing, who deployed China and the United States, also established a research and development center and related road tests in Qianhai, Shenzhen at the end of last year. Regarding the layout in Shenzhen, Mo Luyi, vice president of Xiaoma Zhixing, said that the company has taken the entire Shenzhen's opening and legislative level in the entire Shenzhen. Essence At the same time, the company also hopes to bring the driverless technology to Shenzhen and get the technology improvement in the scene of Shenzhen. What should I do if there is an accident in driverless driving? Shenzhen legislation attempt to clarify rights and responsibilities

The "Regulations on the Management of the Intelligent Connected Automobile in Shenzhen Special Economic Zone" shall be implemented from road testing, access registration, to use management, traffic violations, and accident handling to carry out full -chain legislation for autonomous vehicle management. Legal obstacles have also opened more rights for the commercialization of autonomous driving.

Who should be responsible for an accident? This is a common concern at all sectors of society. The legislation also provides for the division of responsibilities after the traffic accident that the intelligent connected cars that are widely concerned about by all parties. As long as the vehicle is drove, the driver is responsible for violations or accidents. In the state of unmanned driving, there is no driver in the car, and the responsibility falls on the owner and manager of the vehicle, and can further request compensation from producers and sellers.

With the implementation of the regulations, many problems such as autonomous driving cars cannot enter the market, can not be on the license plate, and cannot operate charges will be solved. This also brings a new round of development opportunities for autonomous driving companies. Nie Yuren, the general manager of Baidu Intelligent Transportation Solution, predicts that by the end of this year, more than 300 autonomous driving sites that can be connected in Shenzhen will be further improved by citizens' travel convenience. He emphasized that operating the large -scale operation of running through unmanned commercialization in China is an important goal of Baidu's layout in Shenzhen in the future, and it is also the most important direction of Baidu in the field of unmanned commercialization. Zhou Guang, CEO of Shenzhen Yuanrong Qixing Technology Co., Ltd., was introduced to the details of the implementation of the regulations. He believes that how to charge and which authentication systems need to be charged, which is critical for the next operation of autonomous driving companies. In the future, what kind of verification of the main driving of autonomous driving taxis needs to be obtained. The first step is to introduce first.

How did the unmanned R & D investment invested in the huge "unicorn"?

Zhou Guang, CEO of Shenzhen Yuanrong Qixing Technology Co., Ltd., said that after the implementation of Shenzhen's regulations, autonomous driving has really achieved laws. Enterprises can promote deeper unmanned tests in the future, improve research and development efficiency, and carry out commercial charges. Can make the company "return blood". Mo Luyi, vice president of Xiaoma Zhixing, believes that Shenzhen answers the access of autonomous cars in a legal provision to help the entire industry gradually move from the original test technology research and development to commercialization. Essence

Different from other tracks, autonomous driving involves various emerging technologies such as AI, big data, the Internet of Things, and 5G. From technical research and development, preparation to commercial monetization, autonomous driving companies are experiencing increasingly severe tests. In this regard, Zhou Guang, CEO of Yuan Rong Qixing, pointed out that true commercialization must provide better services with cheaper prices, which is the social value of autonomous driving. In April of this year, Yuan Rong Qixing's first batch of L4 -level autonomous driving pre -driving plans to Shenzhen landed in Shenzhen. According to Yuan Rong Qixing, they have developed the current L4 -level autonomous driving solution with the lowest cost and about 20,000 yuan, and can soon be mass -produced. Zhou Guang is expected to start profitable in 2025. In order to achieve commercial landing as soon as possible, autonomous driving companies also set their sights on the smart logistics industry. Yuan Rong Qixing reached a cooperation agreement with Debon Express in June. Its autonomous driving light card will provide normalized freight transfer services for Debon Express. This is also the first domestic L4 autonomous driving light card commercialization project. When it comes to difficulties, Zhou Guang also admits that the biggest difficulty of commercial landing is to find strategic partners to go forward together, so the company's current strategic direction is to cooperate in depth with car companies. How to grit the teeth and highlight the siege, and it is a problem that the self -driving enterprise needs to face. Mo Luyi, vice president of Xiaoma Zhixing, pointed out that there is a prerequisite whether it is a large -scale mass production or commercial landing, and it is unmanned. To achieve completely, the system handles all the scenes. In this case, it is necessary to do mass production and reduce costs. There is indeed some difficulty and challenges in it, and the entire upstream and downstream cooperation needs

- END -