Baidu released Apollo plan _ the world's first L3 auto-driving production model

From April 2017, BAT giant Baidu released the "Apollo program", and in July Audi released the new generation A8 with the "world's first L3 autopilot production model", and then in early 2018, the startup company Xiaoma Zhixing And Jingchi Technology started the trial operation of driverless vehicles in Guangzhou. Beijing issued auto-driving road test related documents, clarifying that auto-driving vehicles can be tested on the road in Beijing... Let’s take a look at the related electronic car editors. Content.

Looking back over the past year, in the context of the core technology of hardware and algorithms, and the continuous improvement of relevant laws and regulations, autonomous driving is gradually becoming a hot technology from car companies, parts companies and technology companies. Most of the consumers' travel-related hotspots are getting deeper and deeper into the lives of ordinary people.

However, as Li Jianpeng, director of the Singular Auto Autopilot Architecture, said that while autopiloting brings convenience to the society, it also poses more challenges for everyone. Especially in the past two years, there have been many auto-driving car accidents around the world, which fully demonstrates that this technology is not mature enough yet to meet the conditions for operation.

Baidu released Apollo plan _ the world's first L3 auto-driving production model

A prominent problem is safety. It is well known that when autonomously driving a car is running, it needs to communicate with the cloud to obtain information about the vehicle itself and traffic. At the same time, the self-driving car also frequently uploads some information to the cloud. In this process, if the system itself has loopholes or improper operation, it is easy to provide hackers with the convenience of attack. This has been a real case before, which has given the autopilot a new topic about security, especially cyber security.

In addition, Li Jianpeng believes that autonomous driving faces many challenges in industrialization. For example, algorithms, many automakers are now putting a large computer in the car when they are developing autopilot cars. From a commercial point of view, such a solution is actually difficult to mass-produce - especially considering cost and power consumption. In this case, many companies are now focusing on developing high-performance chips. The same is true for sensors. Although in many companies developing R&D cars, Lidar has a good performance in terms of environmental perception, but due to cost, Lidar is still far away from mass production.

In addition to the above technical challenges, in Li Jianpeng's view, the industry currently faces major challenges in supporting infrastructure construction, high-precision map technology, automatic driving data, laws and regulations, and talents in the process of automatic driving operations. .

“In the case of infrastructure, the synergy between the self-driving car and the road is now only in the demonstration area. Although the car runs very fast, there is no 'road.'” China Automotive Technology Research Center Intelligent Vehicle Research Laboratory and Automotive Software Evaluation Wang Yu, director of the center, analyzed. "There are still legal and regulatory issues. You can see that Beijing's automatic driving has gone very fast and took the lead in the road test. But after establishing the management regulations, there is a problem - what is the road? Government leaders are very cautious about this issue. Who is responsible for the risk after going on the road? It is also a problem that cannot be ignored."

Huang Luoyi, Automated Driving Product Manager of Bosch Chassis Control Systems in China, gave a more in-depth explanation of the technical challenges faced by autonomous driving. In his view, there are five main points: security challenges, including personal safety, functional safety and information security. Especially in the future, autonomous vehicles will be connected to the Internet. Vehicles must have the ability to handle various information security inside and outside the vehicle. Sensing technology challenges, in the case of self-driving cars, if you want to finally mass-produce and go on the road, you need 360°, high reliability, high accuracy, and complementary sensor solutions. Electrical and electronic architecture challenges, such as redundancy of power supplies, redundancy of sensors, redundancy of braking systems, redundancy of steering systems, etc. The challenge of system intelligence, although in many people's point of view, the development of autonomous vehicles is to develop a machine that can replace the human driver, but this machine should not be cold, no feelings, but should be able to think like humans Can understand human thinking. Positioning technology challenges, especially the ability to position autonomous vehicles in real time and accurately, is especially important.

Subsequently, Huang Luoyi specifically analyzed the application challenges of AI technology in autonomous vehicles. He believed that there are four main problems: First, the perception technology of traditional vehicles is based on rules. Once problems arise, it is easy to find problems. Where, but AI technology is like a "black box". When something goes wrong, it only knows how to adjust parameters without knowing the root cause of the problem. Second, the AI ​​algorithm requires very large power consumption, and if it is to be used in a production car, the AI ​​algorithm needs to be compressed and placed on a specific chip, which is currently difficult to implement. Third, in the case of auto-driving cars based on AI technology, how to quickly identify problems and repair them in case of problems, eliminate hidden dangers. Fourth, for autonomous vehicles equipped with a large number of algorithms and software, how to verify and verify the safety and effectiveness of each function, each technology is reliable, reliable and usable.

Huang Yong from the Chery Automobile Smart Car Technology Center is different from the above two guests. He mainly cuts in from technological innovation and market competition, and analyzes the current challenges of autonomous driving. Especially in the market competition, many second- and third-level suppliers are now directly involved in the frontier cooperation of autonomous driving. The industry has not cooperated with automakers and tier one suppliers to produce cars, sell them to dealers, and then dealers. Sold to end users. Rather, with the deepening of the new round of industrial changes, more new models have emerged. Cross-border cooperation and competition are one of them. This is not the case before, and companies that need to participate in it should learn. In addition, Huang Yong believes that there are many places where autopilot needs to work hard in product specialization and talents, especially talents. Currently, automotive electronics from hardware, software to chips, artificial intelligence algorithms, the demand for talents is increasingly The bigger.

However, as the CEO of Geshi Automobile Zhou Xiaoyu said, the prominent features of the Chinese market are that there are more people, more money, a bigger market, and a higher fault tolerance. It is easy to get together and follow the trend, when all the smart people (not necessarily technical experts) After entering, the competition has become extremely cruel, and it is often the do-it-yourself who can adapt to the market and survive the technology. This is China's unique innovation and entrepreneurship. When the industry develops to a certain concentration and depth, and the atmosphere is up, it is very easy to form group innovation, and it is easy to combine regional characteristics and form an application-based innovation model. Therefore, although there are many challenges in the domestic R&D and industrialization of autonomous vehicles, when there are more and more “players”, everyone can brainstorm, and the Chinese auto industry may really be able to use the automatic driving “curve overtaking”.

Speaker

Featuring frame sizes from 10 mm up to 205 mm and SPLs ranging from 70 up to 120 dB, our line of speakers offers you the quality and accessibility needed for your sound requirements. Our standard selection of speakers includes a variety of frame shapes, magnet types, and mounting styles to quickly plug and play into your design. In order to address the trend of miniaturization in mobile devices, we also offer compact, low profile speakers that feature diameters as small as 10 mm and depths as low as 2 mm. If one of our standard offerings does not fit your unique audio needs, then we can custom tailor a speaker that is right for you with a number of available mechanical and electrical modifications.

Speaker System,Indoor Speakers,Outdoor Speakers,Indoor Outdoor Speakers

NINGBO SANCO ELECTRONICS CO., LTD. , https://www.sancobuzzer.com