On March 25th, the CEO of Polestar, Xia Yiping, stated that whoever fails to implement "smart driving nationwide" by 2024 will fall behind. He also mentioned that Polestar aims to achieve the capability of driving anywhere that Baidu Maps can navigate to by the end of this year.
In the field of autonomous driving, Tesla stands out as quite unique due to its insistence on a pure vision approach, as opposed to the "sensor fusion" approach utilized by many domestic manufacturers with lidar. Similarly, Xpeng Motors shares Tesla's commitment to the pure vision approach. Recently, Xpeng was granted the first L3 autonomous driving test license in China's Optics Valley, enabling it to conduct conditional autonomous driving road tests in designated areas.
It's been reported that the Jiyue 01 has achieved the first domestic pure visual Occupancy Control Chart (OCC) onboard network "boarding," and the high-level intelligent driving point-to-point navigation assistance system PPA based on "pure vision" is about to be "ready for nationwide use."
The most revolutionary aspect of the OCC onboard network "boarding" lies in the enhancement of perception and recognition capabilities. It can finely depict obstacles in 3D, with model accuracy reaching the centimeter level, comparable to or even surpassing lidar.
At the same time, the OCC onboard network can achieve a speed recognition accuracy for moving obstacles within a range of 0.1 meters per second, significantly improving the ability to recognize boundaries of general obstacles such as fences, guardrails, traffic cones, and construction barriers of a certain length, effectively enabling avoidance and braking, thereby allowing for more safety distance for driving.
With the gradual integration of OCC (Over-the-Air) network technology, Ji Yue has now established an advanced intelligent driving system called "BEV+OCC+Transformer", significantly enhancing the pure visual intelligent driving capabilities in recognizing irregular obstacles and generalizing scenes. This enables automotive robots to truly gain perceptual "intuition", further enhancing the driving and riding experience.
Ji Yue's point-to-point navigation assistance PPA has already covered over 400,000 kilometers of roads nationwide. The high-speed PPA can be used on over 90% of highways and elevated roads across the country, while urban PPA, leveraging pure visual capabilities, has pioneered the navigation of major city roads in Shanghai, Beijing, Shenzhen, and Hangzhou.
