Autopilot, the founder of lidar, will face a big reshuffle?

Why is there such a difference when the enterprise that invented the car regulation laser radar goes bankrupt and the domestic investment is still in the air?

Tesla announced the cancellation of all radars, relying solely on vision to achieve automatic driving. Almost at the same time, the German technology company Ibeo, which invented the car size laser radar, formally filed for bankruptcy. The company that first engaged in the development of car size laser radar fell down, but it was a different scene in China.

Can the R&D company of lidar go bankrupt? This should not happen in 2022. This year is an important time node for the centralized mass production and loading of lidar. This year, the number of mass production vehicles with laser radars has increased to 16, and the number of vehicles equipped with one laser radar is a minority group, including 9 vehicles equipped with more than one laser radar. The Great Wall Salon machine Jialong directly carries four Huawei laser radars.

The founder is bankrupt, and the laser radar is no longer popular?

However, the key point is that the main mass production models of lidar are all in China, and there are few mass production models of lidar overseas at present. In fact, the bankrupt Ibeo is really an old man. Since 2010, it has won the order of Audi's autopilot lidar hardware, but it was not officially delivered until seven years later, that is, the new generation A8 model of Audi with L3 level autopilot function. You may say that Audi does not use Fareo's lidar? Yes, that lidar was developed by Ibeo and Valeo in the same year.

Now that the mass production is over, the most difficult R&D stage has passed. In the past seven years, Ibeo has not received any orders, only Audi is a partner, and what worries Ibeo is that even if Audi A8 realized L3 in that year, there was almost no sales of the version with lidar, because there was no use scenario allowed to go on the road, because until May this year, Audi's hometown, Germany, has officially opened the license for the L3 automatic driving model on the road. It has been five or six years since then. Now Audi has also changed its thinking. New models equipped with laser radar no longer regard laser radar as the hardware for automatic driving, but as a part of the vehicle's active safety perception. In China, Audi's automatic driving has applied the vehicle road coordination system.

Another partner of Ibeo is Great Wall Motors of China. In 2019, Ibeo and Great Wall Motors announced that they will provide solid state lidar for Wei brand SUVs. Just a few days ago, the DHT-PHEV lidar version of Mocha officially met with us. We also analyzed its automatic driving logic of "emphasizing perception, ignoring maps". However, the lidar is not from Ibeo, but from domestic suppliers Sumiton Juchuang, Now Ibeo is bankrupt.

In fact, we can see that even though we have been arguing for so many years, there are still a few car companies that use pure vision solutions, especially those as radical as Tesla. Although Toyota said in April this year that it would use a pure vision scheme, it still saw a laser radar on the Lexus test vehicle recently. Of course, it may be testing whether the laser radar is feasible, because Tesla has done this before.

In fact, laser radar is still necessary for domestic car enterprises, because the data redundancy brought by laser radar is still helpful for the complicated road conditions in China. We do not deny that some independent brands or new power models use laser radar as a gimmick to sell cars, but we really need to face up to the role of laser radar in automatic driving.

At present, most high-order auxiliary driving or autonomous driving models still use visual perception data as the main reference for their solutions, but not every enterprise has the accumulation of visual data like Tesla's, as well as the first mover advantage. Not everyone can learn Tesla's visual algorithm solution (occupying the network), or can afford to learn it. Therefore, other radar sensors, including millimeter wave radar It is necessary to enhance the sensitivity and redundancy of laser radar, or even 4D imaging millimeter wave radar.

Especially for China's latecomer autonomous driving enterprises and new car building forces, the point cloud data of the laser radar can greatly simplify the development difficulty of automatic driving. In the dark and poor vision, the radar point cloud has the advantage of crushing the imaging effect of the visual camera. Although the cost will be higher and consumers need to bear this part of the cost, the time cost of research and development will be greatly reduced, This is similar to our electrified overtaking on curves.

But this does not mean that we will always rely on lidar. The visual algorithms of independent brands and autonomous driving companies are also rapidly iterative optimization, and lidar can win more time for us.

summary

While foreign lidar enterprises are bankrupt, China's lidar market is full of vitality. Hesai Technology, Sagitar Juchuang, Lashen Intelligence, Yijing Technology, Beixing, Mulai Laser, Luowei Technology, Reich Zhiguang... But this scene should also be temporary. Now more and more enterprises are pouring into this chess game, which is likely to muddy the water, and finally only a few enterprises can win, Like Weixiaoli, the market is ruthless. When you are no longer recognized and gradually marginalized, you can either wait for other large enterprises to merge or go on the path of Ibeo.

Author: Zou Yuyuan

Source: OFweek Laser Network