The first year of the outbreak of autonomous driving? Level 4 autonomous driving technology has alre

发布时间:2021-08-27 13:53:46

阅读量 : 69

Recently, the news of Huawei's entry into smart cars has attracted widespread attention. The general trend of automotive smart networking has come, and the future market space is broad. As autonomous driving models will be mass-produced on the market, 2021 is expected to usher in the first year of the outbreak of autonomous driving. The potential market size of the autonomous driving industry has exceeded 100 billion yuan in the near term, and it has reached trillion yuan in the long run. Huawei stated that the smart car architecture “adopts a design approach that starts from the end, and builds a full-stack solution for L4~L2+ level autonomous driving based on the L4 autonomous driving architecture.”

In fact, L4 autonomous driving technology has long been mature applications in the aircraft field.


The intelligent drone developed and produced by Suzhou EAvision Robot Technology Co., Ltd. is the industry's first fully autonomous drone. The application of auto-driving technology to drones has promoted the development of intelligent drone technology. At present, the E-A2021 UAV, which is mainly used in the field of plant protection, has realized all-terrain, all-weather (day, night), all-area, and all-crop operations.



What is the difference between autonomous driving and autonomous driving?

Autonomy is different from autonomy. Autonomy is machine operation without human intervention, and the whole process is unmanned, while autonomy is machine-assisted human operation. From the beginning of the research and development of UAVs, in 2016, EAvision established a fully autonomous technical route, directly crossing the manual and semi-automatic stages. This greatly reduces the threshold for the use of drones and helps users to operate more efficiently. After being freed from heavy agricultural work, farmers can devote more energy to fruit processing and sales, and use their manpower in higher value-added links. For example, in 2020, an old man over 60 years old in Anhui used EAvision UAV to operate more than 10,000 mu times a year.


How does EAvision realize fully autonomous operations?

The perception system of EAvision Autonomous UAV adopts multi-sensor fusion technology, which is the current trend of auto-driving in automobiles, and uses binocular vision sensors, lidar, millimeter-wave radar and ultrasonic mines to realize three-dimensional environment perception. The powerful computing platform adopts the CPU and GPU architecture, which has the largest computing power in the UAV industry today, solves the bottleneck problem of high-density 3D point cloud computing, realizes real-time high-speed computing, and obstacle detection time is less than 100 milliseconds. It is precisely because of these superior technologies that EAvision UAV does not require surveying and mapping during operations, and can flexibly avoid obstacles in complex environments such as hills and mountains.


How to measure the technology of autonomous driving?

Availability, is one of the indicators for measuring autonomous driving technology. Due to the complexity of the scene and the maturity of the technology, the application of many autonomous driving technologies needs to limit the scene, that is, the availability is not high. The extremely high-availability autonomous driving technology has achieved high availability. The operating terrain covers plains, gentle slopes, hills and mountains. It is suitable for 90% of China's terrain and can still fly autonomously in a complex mountain environment with a slope of 70°.


How does EAvision realize the path planning of complex plots?

To deal with complex and irregular plots, EAvision uses the most advanced artificial intelligence algorithm to quickly and efficiently plan the optimal path according to the characteristics of the plot, reduce path duplication and air flight, thereby improving operation efficiency.


For users, will technology upgrades increase hardware costs?

Similar to today's autopilot technology, the intelligent algorithm of EAvision UAV can learn independently and continuously upgrade. With the increase of operating data, the performance becomes better and better without frequent replacement of new intelligent hardware.