Camera VS LiDAR: Autopilot Industry Chain Controversy
With the arrival of the earnings season, 35 of the 82 self-driving concept stocks have so far disclosed their performance forecasts for the first half of the year. Among them, Changan Automobile, Great Wall Motor, and BYD have "booked" the top three positions in net profit.
Although BAIC Blue Valley has in-depth cooperation with major manufacturers such as Huawei and Baidu, it is still difficult to escape the fate of losses.
In addition, Shenglu Communication and Gaohong Co., Ltd. broke through the siege and are expected to achieve a net profit increase of up to 10,529.92% and 3,787.76% respectively. The former can provide millimeter wave products for unmanned driving, while the latter provides C-V2X car networking technology solutions , which can enhance the interaction between vehicles and traffic lights.
It is not difficult to see from the financial report that under the wave of unmanned driving, both upstream and downstream enterprises in the industry chain have benefited.
Among them, the camera technology route chosen by Tesla and the lidar technology route chosen by a large number of domestic car companies have brought a group of companies alive.
Brokerages predict that the lidar market space in my country's passenger vehicle sector will reach 98 billion yuan by 2030.
However, Qiu Bin, an associate researcher at the School of Vehicles and Transportation of Tsinghua University, told Shell Finance reporters that if the cost of autonomous driving for a single model is too high, it will inevitably affect the price of the model, and thus affect the market share.
The main problem facing automakers is how to balance technology and cost.
The two companies with the largest increase in net profit have "unique skills"
Wind data shows that as of August 23, there were 82 unmanned concept stocks, of which 35 have announced their performance forecasts for the first half of this year.
Among them, 20 companies are expected to be profitable, and 15 are not.
It is estimated that the three companies with the highest net profit caps in the first half of this year are Changan Automobile, Great Wall Motor and BYD. The net profit is expected to reach 6.2 billion yuan, 5.9 billion yuan, and 3.6 billion yuan respectively, and the minimum net profit is expected to reach 5 billion yuan. yuan, 5.3 billion yuan, 2.8 billion yuan.
Changan Automobile said that the profit growth in the first half of this year was mainly due to the continuous improvement of the company's brand, the continuous optimization of product structure, and the continuous improvement of the profitability of its own brands.
BYD's unmanned driving has also made considerable progress this year.
The third-generation pure electric driverless delivery vehicle jointly developed and designed by BYD and the American technology company Nuro has been officially released.
It is understood that in this cooperation, BYD will be responsible for vehicle development, vehicle testing and manufacturing, and provide core assembly components such as blade batteries, motors, electronic controls and human-computer interaction.
In addition to the above-mentioned three large OEMs, it is expected that the net profit of the fourth to tenth with the highest net profit cap will remain between 800 million and 100 million yuan.
Among them, Top Group, Thundersoft, and Gaohong Co., Ltd. are expected to achieve a maximum net profit of 737 million yuan, 415 million yuan, and 290 million yuan, and a minimum of 667 million yuan, 388 million yuan, and 250 million yuan, respectively.
In addition to this year's "King of Profits", there are still some companies whose net profit has soared. A total of 9 companies expect their net profit to rise by more than 100%.
Among them, Shenglu Communication estimates that the highest net profit increase can reach 10529.92%, and Gaohong shares predicts that the net profit can increase by 3787.76%.
Shenglu Communication's millimeter wave products are quite well-known.
The company's 2021 annual report shows that the company's civil communication business is mainly engaged in the research and development, production and sales of products and services such as mobile communication antennas, radio frequency devices and active integrated communication equipment, and vertical industry overall solutions. Key technologies for mmWave product design and manufacturing.
The company's products cover automotive antennas, etc., forming a relatively complete communication antenna product line. Customers mainly include domestic and foreign telecom operators, communication equipment integrators, and automobile manufacturers.
In July this year, Shenglu Communication responded to investors' questions about "what applications of the company's products in driverless driving", saying that the company's UWB positioning technology for autonomous driving has been mature and applied, and the company is currently cooperating with SAIC Group.
Another company with soaring profits, Gao Hong, mainly provides C-V2X solutions for driverless companies, and has cooperated with many customers to carry out testing and verification in various places.
According to reports, the C-V2X Internet of Vehicles technology can mainly strengthen the interaction between vehicles and traffic lights, so that autonomous vehicles have the ability to know the status and driving intention of other surrounding vehicles in real time, and the ability to obtain changes in surrounding road information in real time, so as to improve the phase of signal lights. Real-time and reliable information acquisition, and effectively improve the ability to cope with complex working conditions such as passing through non-signal-controlled intersections and passing under severe obstructions ahead.
Some are happy and some are sad.
BAIC Blue Valley, Celis, Oufeiguang and other three companies are expected to lose up to 2.2 billion yuan, 1.76 billion yuan, and 950 million yuan respectively in the first half of this year, becoming the company with the largest pre-loss so far.
In addition, Jianghuai Automobile, *ST Sunsea, King Long Automobile, etc. are also expected to suffer losses.
Among them, BAIC Blue Valley has released its semi-annual report on August 5. At that time, it said that the company and Huawei continued to promote product development, brand marketing and other related cooperation, and the new HI version of Jihu Alpha S was delivered during the reporting period.
On the other hand, the company and Baidu continued to promote the cooperation in the field of Apollo intelligent driving, and the self-driving αT vehicle based on the Polar Fox model obtained the license for the unmanned commercialization pilot in Beijing.
The company and world-renowned battery suppliers continue to deepen the cooperation in model projects to develop and improve the supply guarantee mechanism.
BAIC Blue Valley's Intelligent Vehicle Motion Control System (IVMCS) has completed software process improvement and capability assessment in the automotive industry, and the L4 autonomous driving project has obtained the T3 license for the Beijing autonomous driving road test.
However, why did BAIC Blue Valley, which has already cooperated with the two giants Huawei and Baidu, suffer losses?
BAIC Blue Valley said that in the first half of 2022, the shortage of chips and the continuous rise in the prices of raw materials such as batteries, coupled with the repeated outbreak of the new crown pneumonia, had a greater impact on the company's performance; in addition, although the company's sales of new energy vehicles in the first half of 2022 were higher than the previous year There has been an increase in the same period, but it is still in the climbing stage, and the scale effect has not yet appeared.
At the same time, in order to improve the competitiveness of products, the cost of the company's brand and channel building and R&D investment continued to increase.
In addition, Wanan Technology is hovering between profit and loss. The company expects to achieve a maximum net profit of 5 million yuan and a minimum net loss of 3 million yuan.
Camera and lidar, who is the mainstream?
The development of autonomous driving is in full swing, so which technical route will become the mainstream?
Qiu Bin introduced that in the traditional driving mode, people rely on human eyes and brains to collect and process information, and then adjust the state of the vehicle according to the information. Autonomous driving actually relies on machines to collect information to make decisions.
The most difficult part at present is whether the machine can accurately collect and recognize various scenes.
In this scenario, cameras, lidars, millimeter-wave radars, and ultrasonic radars can all be the choices of car companies.
Among them, Tesla is a typical car company that chooses cameras.
Tesla CEO Elon Musk has publicly stated that any company that relies on lidar is destined to have no future.
Some of Tesla's Model 3 and Model Y models were previously equipped with millimeter-wave radar, but not long ago, they also modified cameras with pure vision solutions to assist driving.
Obviously, Tesla chose the purely visual route.
On the other hand, many manufacturers are also very optimistic about lidar, and 2022 is even called the first year of lidar.
According to the reporter's incomplete statistics, there are more than 20 models that have been announced to be equipped with lidar in China.
The Zhiji car under SAIC Group can be equipped with a LiDAR fusion intelligent driving hardware system, including a complete set of hardware such as high-precision LiDAR and Orin X chip with high computing power.
In mid-June, NIO's new SUV ES7 was equipped with the 1550nm LiDAR "Falcon" provided by InnoVusion.
On June 21, Lili's new SUV Lili L9 was also unveiled with lidar.
On July 16th, the new HI version of the Alpha S was officially launched in batches at the Blue Valley Magna Intelligent Manufacturing Base in Zhenjiang, Jiangsu.
Under the support of the entire computing platform with 3 lidars, 34 sensors and 400T computing power, the new HI version of Jihu Alpha S can realize "high-speed, urban, parking" full-scene high-level intelligent driving assistance; based on HarmonyOS development The smart cockpit supports the "one-button transfer" function, which can realize the docking of the mobile phone and the car machine.
In addition to the above models, Feifan, Weimar, Nezha and other brands have launched models equipped with lidar.
Liu Yu, deputy general manager of BAIC Group and chairman of BAIC New Energy, said publicly, "Around the two key features of intelligent cockpit and high-level assisted driving, cars will become completely different from the past, and will provide a lot for the development of the entire industry. Chance."
At the same time, lidar manufacturers have also ushered in an upswing.
According to IDC data, the penetration rate of new L2 autonomous passenger vehicles in China reached 23.2% in 2022Q1, a significant increase from 7.5% in the first quarter of 2021.
According to the Ministry of Industry and Information Technology's "Intelligent Connected Vehicle Technology Roadmap (Version 2.0)", the domestic L2 and L3 new vehicles will reach 50% in 2025, and exceed 70% by 2030, and L4 will account for 20%.
According to the Huaxi Securities Research Report, in 2022, China's passenger car lidar installations will exceed 80,000.
According to the statistics of Zosi Automobile Research Institute, the installed number of lidars in H1 domestic passenger cars will reach 24,700 in 2022.
In the second half of 2022, more than 10 new lidar models are planned to be delivered in China, including Xiaopeng G9 and Weimar M7, which will greatly increase the number of lidars installed. It is expected that the total installed capacity for the year is expected to exceed 80,000.
According to estimates, it is estimated that the market space of lidar in the passenger car field in my country will reach 26.1 billion yuan in 2025 and 98 billion yuan by 2030.
It is expected that the automotive lidar market will benefit from the penetration of high-level autonomous driving and maintain rapid growth.
According to our calculations, the compound growth rate of the lidar market in my country's passenger vehicle field can reach 200%+ in the next three years, and the compound growth rate from 2025 to 2030 will reach more than 30%.
In the fast-growing market, the sensor products launched by Hesai, DJI, Sagitar and other Chinese manufacturers have obvious price advantages, breaking the leading position of developed countries such as the United States in this field.
From the perspective of the entire industry chain, in terms of upstream components, Changguang Huaxin, Focuslight Technology, YOFC Fiber Optics, Sunny Optical, Yongxin Optics, Fujing Technology, Tengjing Technology, Lante Optics, Crystal Optoelectronics, Union Chuang Electronics, Tianfu Communication and many other listed companies have been involved.
From the perspective of midstream OEMs, Hesai Technology, Sagitar Juchuang, Wanji Technology, Huawei, DJI Lanwo, Aobi Zhongguang and other companies are very prominent.
Among them, Sagitar Juchuang supplies to GAC Aian, SAIC Zhiji, Xiaopeng, Weimar and other manufacturers.
Qiu Chunxin, founder and CEO of Sagitar Juchuang, has publicly stated that after years of technological changes, the human dream of autonomous driving has finally ushered in the dawn of mass production in the era of artificial intelligence.
Cost is the roadblock to commercialization
On the road to popularization of autonomous driving, cost is still a roadblock.
"It is expected that by 2025, driverless taxis may be commercialized on a small scale. If it reaches the stage of commercialization of driverless cars in the whole industry, it may have to wait until 2035," said Zhang Xiang, a specially-appointed expert from China Bolian Think Tank. Mass production has begun, and the computing power of millimeter-wave radar and chips is also increasing. However, the development of driverless software still lags behind the development of hardware, which will become a technical obstacle.
Qiu Bin said that even if the technical bottleneck is not considered for the time being, my country has become a big car country, and the production of cars is relatively high. If the cost of autonomous driving for a single model is too high, it will inevitably affect the price of the model, and then affect the market share.
The main problem facing automakers is how to balance technology and cost.
"Now car companies can put a large number of perception methods such as lidar and multi-eye cameras in the autonomous driving system, which can better demonstrate the company's algorithms and capabilities, but if mass production is considered, it is necessary to calculate whether this method can be used. Profitability and long-term competitiveness.”
In Qiu Bin's view, Tesla's current choice of cameras instead of radar is mainly due to cost considerations. If the cost of lidar, millimeter-wave radar, and ultrasonic radar all drop significantly, then Tesla may also use these technologies. Integrate into the car.
"The concept of 'redundancy' is often mentioned in the automotive industry. In fact, this is the case. There are cameras and multiple radar technologies in a car. One technology may make mistakes, but the probability of misjudging multiple technologies at the same time is great. Reduce, if you can control the cost and combine more, I believe this can solve the perception problem of the car to a certain extent," Qiu Bin said.
Qiu Bin speculates that if the cost of lidar can drop to between $100 and $400 in the future, the number of vehicles equipped with lidar will increase significantly.
Qiu Bin said that, on the one hand, car companies need to improve their performance technically and train the system more reasonably and more heavily. On the other hand, with the continuous popularization of autonomous driving, the urban transportation system can also be improved.
"Urban traffic occasionally has abnormal conditions. Isolation piers are placed during road construction, roadside facilities are damaged, some small-scale natural disasters such as landslides, and traffic accidents cause road conditions to be too complicated in a short period of time. It is equivalent to a special scene in urban traffic, the vehicle system may have never learned it, and there may be loopholes in identification.” He said that if the urban traffic system can report abnormal situations in time, it will be synchronized to nearby vehicles, road and vehicle ends. Combined, the probability of identification errors may be greatly reduced, and the safety probability will be greatly improved.
"The intelligentization of road traffic facilities is the future development direction. If sensors are installed at intersections in the future, the cloud computing will be sent to the cloud, and then synchronized to nearby cars, which will greatly improve safety, but how to ensure investment and safety is still a problem." Industry analyst Zhong Shi said that the current high-level autonomous driving technology is very expensive to apply to cars, that is, the cost of bicycle intelligence is too high to prevent popularization. If the level of the intelligent transportation system is increased, it can effectively transfer and reduce the cost of smart cars. , but this requires huge investment in government infrastructure.
Looking at it now, who will be responsible for these huge investments is still an unsolved problem.
In Zhong Shi's view, the use of monthly and annual fees to pay in installments is a possible model for the future commercialization of unmanned driving, but even so, too high fees will still make consumers hesitate.
How to realize the commercialization of autonomous driving and how to realize it may continue to trouble car companies for some time to come.
Beijing News Shell Finance reporter Lin Zi