Introduction
At this year’s Guangzhou Auto Show, in addition to many new models, the most eye-catching is the LiDAR system. The NIO ET7 has one, the ZhiJi L7 and XPeng P5 have two, the WM Motor M7 has three, and the first car of the high-end salon of the Great Wall, the Mecha Dragon, comes with four LiDARs. It is said that the number of LiDARs in the domestic car market has become an important field in the competition.
LiDAR has reached the conditions for mass production in vehicle models
Looking at global car companies, only Tesla is still sticking to a pure visual solution. The reason why Tesla does not want to use LiDAR is not because LiDAR’s performance is not good, but simply because the cost of LiDAR is too high, making it difficult to use in mass-produced vehicle models. In Elon Musk’s mind, the technology that cannot be mass-promoted is not worth Tesla’s research efforts. But with the birth of solid-state LiDAR represented by Luminar, the cost of LiDAR has been significantly reduced.
The price of a single LiDAR can enter the $100 range, becoming the target of major manufacturers. Although the current price of LiDAR is still high, it is not unattainable for vehicles like the Great Wall Salon with a high positioning.
Functionally, the LiDAR’s sensor mainly emits laser pulses and captures backscatter lasers, and calculates the distance between objects by computing the time it takes for the light source to return. Compared with millimeter-wave radar and cameras, LiDAR has the characteristics of high resolution, long distance, wide field of view and no environmental climate conditions. However, compared with cameras, the biggest advantage of LiDAR is that it can effectively recognize non-metallic objects in the distance. This is also the cause of the accidents frequently occurred in previous Tesla and NIO models.
Analysis of Domestic LiDar-equipped Vehicle ModelsAt this year’s Guangzhou Auto Show, the Mechadragon made its debut, attracting attention from all sides with its four solid-state hybrid laser radars from Huawei. Each radar has a viewing angle of 120° * 25°, and is located in the center of the front and rear bumpers, as well as under the left and right side skirts of the vehicle. The Mechadragon’s 360° full-angle coverage allows for stable point cloud formation, which can help the controller make the most accurate judgments for autonomous driving. However, the cost of four laser radars is still a considerable challenge for lower-priced vehicles. It is difficult for ordinary consumers to enjoy such configuration benefits without a significant drop in laser radar costs.
As for the WM Motor M7, it adopts three RoboSense (Suteng Juchuang) second-generation smart solid-state laser radars. Suteng Juchuang is a well-known laser radar company in China, and its technology solutions will be adopted by other companies such as IM and Aiways. The M7’s three solid-state laser radars are deployed on the top front of the vehicle and the left and right front fenders, responsible for forward, left rear, and right rear detection. Each radar can form a viewing angle of 120°, and M7 can form a perception area of 330° to detect obstacles in front of the vehicle more comprehensively. As a strategic model for WM, three laser radars were thought to be enough, but no one expected that the Mechadragon would have an additional rearward laser radar. As the saying goes, it is better to be safe than sorry.The ET7 by NIO is equipped with an Innovusion laser radar. Although this radar and the Huawei 96-line hybrid solid-state laser radar carried by the Mechagodzilla both use the first-generation mechanical scanning technology scheme, the detection distance has been greatly increased. According to NIO’s official statement, the detection distance of this laser radar can reach 500 meters, far outperforming Huawei’s laser radar. Longer detection distance means more reaction time and a significant improvement in safety, but the corresponding cost will also increase significantly. Therefore, even for a financially strong company like NIO, it is unable to install four laser radars at once.
The role of the invisible autonomous driving software platform is more important than the number of laser radars
Actually, it is easy to promote the number of laser radars, but the experience of autonomous driving/driving assistance is difficult to quantify. The bigger purpose of major automakers increasing the deployment of laser radar is for the terminal market promotion. Without a relatively leading autonomous driving software platform, more laser radars are just useless. Of course, the deployment of laser radars is a material cost increase for every automaker, and it is indeed much better than pure visual schemes in terms of recognizing static objects. Therefore, for consumers, it will no longer be possible to judge the level of autonomous driving or the quality of its safety performance solely based on the number of laser radars in the future. After all, no carmaker dares to publicly claim to have achieved L4-level autonomous driving, which means that the driver still needs to pay attention to the road ahead at all times during the driving process and be able to take over the vehicle at any time in case of emergency. Therefore, to some extent, the driving assistance/autonomous driving function now provides an extra layer of protection for driving safety, and from this perspective, the number of laser radars is not that important.
Overall, we still hope that automakers will invest more resources in the development of autonomous driving software platforms, as this is indeed the key to determining the core competitiveness of car companies. Whether it is the previous soul-searching or the recent recruitment of software talents by major automakers, the ultimate goal is to firmly grasp the autonomous driving technology. Compared with such a strategic-level task, the number of laser radars is really not that important.
This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.