Domestic lidar manufacturers are at a key point on the eve of profitability.

As the leading companies in this field, Sagittarius Juchuang and Hosai Technology both mentioned profit prediction in their quarterly reports. The former set the time for achieving balance of payments at 2025, or said that the fastest time would be the fourth quarter of this year.

This year, the growth rate of domestic lidar shipments can be described as rapid. According to Sagittarius's financial report, the company shipped 120000 units in the first quarter, up 457 percent from a year earlier. The number of Wo Sai is 59000, which, although less than the former, is also a significant increase of 69.7 over the same period last year.

Both companies' growth is supported by ADAS (Advanced Driver Assistance Systems) lidar products.

Looking at the entire industry, the growth rate of lidar shipments is also considerable. In 2023, the global passenger vehicle lidar shipments of nearly 600000 units, by 2024 is expected to reach 1 million units. Wang Chuanfu, chairman and President of BYD, has repeatedly said publicly that "the first half of the automobile industry is electric and the second half is intelligent."

Although the growth rate is fast enough, the penetration rate of lidar is not high, only 3% in 2023. Due to the high price, most of its main models are above 200000 yuan, which is not the mainstream model sold in the market. In addition, lidar manufacturers also need to deal with the route of pure vision technology, as well as price cost, enterprise strength, market competition and other challenges.

In the wave of intelligence, the fate of domestic lidar manufacturers is also ups and downs.

Saw between cost and price

Using lidar as the main sensor is one of the technical routes in the field of intelligent driving, and the other is to use the camera as the main sensor and abandon the purely visual route of lidar.

From 2018 to 2022, lidar is in the early stage of industry development, mainly serving L3 and L4 autonomous driving technologies (the technology is divided into six levels according to the degree of automation), such as Robotaxi (driverless taxi). However, due to the delay in the commercialization of L3 and L4, in recent years, lidar manufacturers have begun to focus on providing perception for ADAS.

In steering, the biggest problem facing lidar is the see-saw between price and cost. The cost of early Robotaxi is more than one million yuan, and the price of lidar is the most expensive or even hundreds of thousands of yuan. It is difficult for passenger car companies to afford such expensive hardware costs.

Take a 200000 yuan new energy vehicle as an example, usually the cost of smart driving scheme will be controlled within 4%, that is, about 8000 yuan. Just three years ago, the price of lidar was generally 5000 yuan a piece, which was more than half of the cost of smart driving. As a result, few car companies were willing to carry lidar on cars at this price. This concept has gradually spread to more non-mainstream models.

Musk has added a "fire" to the pure vision automatic driving scheme ". In 2021, under his leadership, Tesla became the first company to apply pure vision autonomous driving solutions to mass production models, completely abandoning the assistance of lidar, thus greatly reducing the hardware cost of vehicles. After seeing Tesla's success, a number of domestic smart driving companies and car companies have raised their flags to put into the pure visual tide.

The follow-up of car companies is closely related to the rational purchase of cars by consumers in the past two years, and they have to "sacrifice" the price banner to compete for the market. In order to alleviate the pressure on profit margins due to the intensification of price wars and rising marketing expenditures, cost reduction has become a priority for car companies, and the ranking of lidar has naturally continued to move backward.

This pressure is also transmitted directly to the lidar manufacturer. "We are now being forced by OEMs customers. They hope that the lower the cost of sensors and domain control chips, the better. I fully understand." Hesai co-founder and CEO Li Yifan said to the interface news and other media.

But he also pointed out that in order to control these apparent costs, higher "comprehensive costs" may be required ".

Tesla, which chooses the visual route, not only has an advantage in algorithm, but its comprehensive strength is also based on a series of capabilities such as on-board chip self-research, data loopback and automated labeling, and self-built supercomputing center training model. The company plans to spend about $10 billion this year on AI training and reasoning.

The end of the wave has previously expressed a similar view on the interface news, that although the abandonment of lidar can reduce hardware costs, but pure visual perception program for software research and development requirements are higher, the upfront investment is greater.

"Not every car company is Tesla." The relevant person in charge of YI Technology said that due to the huge development cost of choosing a pure vision route, since the second half of last year, few domestic smart driving companies and car companies have chosen to insist on pure vision.

As far as lidar technology itself is concerned, it is irreplaceable-it is not only a functional part, but also a safety part.

Lidar belongs to active light, compared with the camera passive perception of the visual system can not be affected by ambient light. A 2022 study by the Massachusetts Institute of Technology (MIT) team showed that the fusion method equipped with lidar improved the accuracy of night-time perception by 3 times, and could also reduce the number of false triggers of the automatic emergency braking system (AEB).

"It's hard for anyone to sacrifice security to save some money." Li Yifan said.

Sailing against the current, descending

Although the "pool" is still shallow, the rapid development of the domestic automobile industry also drives the suppliers of the intelligent driving industry chain to enter and exit frequently.

"The auto industry has changed a lot, and the pace of consumer electronics is gradually approaching." The relevant person in charge of a path technology told the interface news.

The average selling price of consumer electronics products is low, the iteration is fast, and suppliers need to do research and development, reduce costs, and push new products. This rhythm is also increasingly mapped within the automotive industry. The life cycle of a model in the traditional automotive industry is 7 years, with a 3-year mid-term facelift. But now Chinese car companies, especially new energy car companies, the life cycle of the basic generation of models is 3 to 5 years, and the models are changed once a year.

The person in charge of the above-mentioned first-path technology mentioned that it has only been a few years since the first generation of lidar was put on the bus. The products of the past era have not kept up with the current new architecture products in terms of cost and performance.

This is undoubtedly a huge challenge for suppliers. In the past, as long as their products entered the supply chain of a model, they were "safe" within 3 years or even 7 years. Now they may be eliminated in the annual facelift. Of course, there are opportunities, and the window for more projects is open.

According to data from the Gestapo Automotive Research Institute, throughout 2023, Hesai Technology, Sagittarius Juchon, Tupatong and Huawei Technologies lidar load (excluding imports and options) were in the forefront, accounting for a total of 80% of the market share. Wesai Technology ranked first with a 37.3 per cent share, with tudatong, Sagittarius and Huawei Technologies in the last few places.

This year's ranking has changed rapidly, from which you can see the intensity of competition in the industry. From January to March, Sagitar's lidar installed capacity turned to first, with a 51.6 per cent market share; Hosai Technology ranked third, with a 15.2 per cent share; and Huawei rose to second place, with a 18.6 per cent share.

In the face of the test of survival, manufacturers are fighting for price and technology, the production line is not stopped, and the inner volume is not stopped. At present, it has entered the "thousand yuan era".

During the Beijing auto show, a number of domestic laser radar manufacturers launched their own new products. Among them, the new medium and long-range lidar MX released by Sagitar Juchuang will achieve mass production of its first project at a cost of less than $200. Previously, the price of Sagittarius Juchuang's products was around US $500, and this MX has dropped by more than 60%.

The company also plans to lower the cost price further in the future. According to Qiu Chunchao, CEO of Sagittarius, the price will be locked to about 1000 yuan as soon as possible.

One Path Technology quickly released the ZVISION EZ5 after lowering the cost of the lidar product ZVISION EZ6 to below 2000 yuan at the beginning of the year. The relevant person in charge said that the price of the new product will also be within 200 US dollars, and there is a clear path to reduce the cost to less than 1000 yuan.

Compared with the previous prediction of the price of single lidar in 2023-2027 by Guolian Securities, which is 4500 yuan, 3800 yuan, 3300 yuan, 3000 yuan and 2800 yuan respectively, the current price reduction rate is far faster than the industry expectation.

Improved architecture and scale are two common paths for lidar manufacturers to reduce prices.

Different vendors have different choices of architecture. Take Sagittarius as an example, the company uses a two-dimensional scanning technology route to avoid NVH (noise, vibration and vibration roughness) risks caused by the motor structure and deterioration after long-term use by chipping each plate of the lidar as much as possible, while breaking through the limitations of the system.

On the processing side, the company has integrated the analog front end of sending and receiving, including the digital link end and the algorithm end, to a system-level dedicated chip (Sagittarius Self-Research SoC:M-Core).

After the structure becomes simpler, the production difficulty and capacity pressure decrease, and the maturity of the upstream process is increasing.

Qiu Chunchao believes that the structure of mass production products is not the end point. In the future, the lidar products produced by Sagittarius can be changed from five transceiver modules to one, further reducing the number of internal components and the complexity of the system, and driving the cost to continue to go down.

For the development and layout of products, Sagitar Juchuang has chosen a platform-based iterative model, so that each generation of products will not change too much, production lines and supply chain resources can be reused, product quality stability is higher, and some costs can be saved. Qiu Chunchao hopes that through a large-scale approach, the barriers will be gradually raised, and a generation gap will be gradually formed to obtain a higher gross profit margin.

"If the price reduction is to provide consumers with a product with incomplete performance, it is actually meaningless." A lidar industry insider told the interface news analysis that in the future, the pure vision scheme will gradually achieve the same performance as the laser radar in intelligent driving below L3. The performance of the laser radar should be maintained or even continuously upgraded while the price is reduced in order to be competitive.

Not only laser radar manufacturers unilaterally take the initiative to reduce the cost, car companies also play an important role in it. The relevant person in charge of Yijing Technology told Interface News that on some general components, some head car companies will help lidar manufacturers talk about a lower supply price as large customers. Car companies do not want suppliers to be in a state of negative gross profit supply all the time, the upstream and downstream of the industrial chain are actually working together to reduce costs.

Bet on a future

As the cost of lidar continues to decrease, it becomes possible to carry lidar on 150000-200000 models. This also lays the foundation for manufacturers to enter a broader market space. According to calculations, the penetration rate of lidar in the new energy market above 150000 will exceed 16% in 2024. According to Li Yifan's judgment, after 2024, the penetration rate of lidar will be like smart phones and new energy vehicles, ushering in leapfrog growth.

In the first half of April this year, retail statistics of China's passenger car market showed that the market penetration rate of new energy vehicles exceeded 50% for the first time, and the market share exceeded that of fuel vehicles for the first time. At the Beijing Auto Show held at the end of April, new energy vehicles also occupied the C position.

"This may be the first time that we can hardly see the launch of a new fuel car at the auto show." Gong Min, head of UBS's China Automotive Industry Research Team, mentioned in an interview with Interface News and other media that the current pattern of the automotive and new energy vehicle industry will also affect the upstream industry's shift to electrified components and intelligent electronic hardware and corresponding software. Both the most cutting-edge technology and the most intense competition are in China.

Several major lidar companies rely on a small number of head car companies and models for sales. Such as the ideal of cooperation between Wesai, Xiaomi, Sagitar Juchuang cooperation of Xiaopeng, Geely, etc., will affect the entire market share of shipments due to the emergence of explosive models. Data show that in 2023, the delivery of standard lidar in China's passenger car market (excluding import and export) will account for more than 70% of the TOP 10 models, while the average delivery of other models will be less than 6000.

Li Yifan believes that although it has not become a profitable business to carry lidar on the Robotaxi, lidar has been recognized by car companies and consumers on ADAS. Such a car has a higher gross margin and a closed commercial loop.

But cars can't all carry the ambitions of lidar manufacturers. Qiu Chunchao said that only in the field of vehicles, Sagittarius has been able to make up losses, and the company still has a large investment in research and development in AI and robotics.

"The robot market can eventually achieve 10 times, or even 100 times the amount of cars." He said, "including humanoid robots, logistics delivery robots, service robots, sweeping robots, etc., as long as the robot needs to move, it will basically be applied to lidar to improve its intelligent level."

At present, Sagittarius has customers of more than 2000 robot companies. Different from the automobile industry, the customers of robot companies are characterized by multiple and scattered customers. The demand of each customer is relatively small, and it is still in the application exploration stage. Therefore, vehicle lidar is still the fundamental aspect of revenue. However, thanks to the introduction of new concepts such as humanoid robots, the industry can see a wider application space for lidar.

Although lidar manufacturers are generally at a loss, they are still increasing their investment in research and development. In the first quarter, Sagitar Juchuang invested about 0.15 billion yuan in research and development, up 97.4 percent from the same period last year, while Hesai Technology spent 0.194 billion yuan on research and development, including investment in robots.

The two companies are also trying to expand into overseas markets. This quarter, Hosai and two of the world's top ten international automakers by revenue reached a fixed-point cooperation in lidar mass production, including a global mass production model of an international automobile brand. Official information shows that it has cooperated with the world's largest vehicle factory in 2021-2022, which is speculated by the media to be Toyota of Japan.

Among the 10 international automakers with the highest income in the Fortune 500 list, 8 automakers or their joint venture partners have reached a fixed-point cooperation relationship with Sagitar Juchuang, and 6 automakers or their joint venture partners have reached a fixed-point cooperation relationship with Hesai Technology.

The relevant person in charge of a path technology also has confidence in the future of laser radar. He believes that if we consider from the perspective of cost, we will choose between several sensors, but when the cost is reduced to a certain level, it is almost the same as that of high-end cameras, or from the perspective of cost and system gain, lidar has reached a relatively high cost performance ratio, then enterprises will certainly hope to incorporate lidar into the system.

Big waves, fast-changing, lidar manufacturers are waiting to win the future.