With foreign companies dominating and domestic brands rising, who will dominate the 100 billion market for automotive sensors?

Liu Xi, a perception fusion expert at Furuitech, told Chuxingyike (ID: carcaijing) that multi-sensor fusion technology can achieve redundancy, that is, when a single sensor fails, it can be cross-verified through other sensor data to improve fault tolerance. In addition, when the cost of lidar is high, multi-sensor fusion can improve the economic benefits of the solution.

Not only will the in-vehicle sensors be integrated, but the road-side sensor data will also be incorporated into the digital stream to supplement the vehicle-side autonomous driving.

The 5G-V2X vehicle-road collaboration route of "from vehicle to road, from terminal to cloud" has been mentioned by more and more autonomous driving companies. Waymo, which has always emphasized the intelligence of a single vehicle, has also opened up a lane to study vehicle-road collaboration technology solutions.

Research shows that vehicle-road collaboration can solve the problem of insufficient confidence in vehicle sensor perception alone, or improve vehicle safety by 2-3 times. By adding sensors such as cameras and radars to the roadside, it is like adding a pair of "God's eyes" to the car's perception equipment, reducing the dimension to solve complex working conditions and scenarios.

Against the backdrop of vigorously promoting new infrastructure, some test roads in the Shanghai and Xiongan autonomous driving demonstration zones have been equipped with high-definition 8K cameras, lidar, millimeter-wave radar, etc.

"These technologies help the car understand the traffic conditions on the road, such as identifying traffic accidents and which lane is occupied. This information is transmitted back to the cloud and then to the car. The driverless car can foresee and plan routes to avoid them in advance," said Liu Xi.

Wang Xiaobin pointed out that vehicle-road collaboration is equivalent to a "beyond-visual-range sensor" for a car, providing information beyond the detection range of existing vehicle-mounted sensors, especially in complex scenarios.

"From the mainstream point of view, to achieve high-level autonomous driving, information from the road side is essential. But now, the cost of vehicle-road collaboration is a limiting factor. It is difficult for a single car company to bear it, and it requires the government or industry to take the lead in promoting it," said Wang Xiaobin.