LiDAR 2022: Walking in the Air Vents, Bubbles, and Hopes

This is the starting point, not the end. To win this race, in addition to technology, you also need to choose the right car company, but also need to tell users what is lidar, and the car equipped with lidar is good

Text | Guo Yu

Edited | Shi Zhiliang

In the past three months, the world's lidar companies have gathered the most in Las Vegas.

Whether it is Velodyne, Luminar, Valeo, Ibeo and other foreign companies, or Sagitar Juchuang, Hesai Technology, Innovusion (Tuttong) and other Chinese companies, they all came to the U.S. casino in early January, with their own new products, to participate in one of the world's most high-profile electronic technology exhibitions - 2022 American Electronics Consumer Show (CES).

When the self-driving function gradually became the standard in cars, the hardware lidar behind automatic driving also stood in front of the spotlight from behind the scenes. The lidar companies participating in CES range from elders like Velodyne, who were founded in the last century, to rookies like Hesai Technology, who were born a few years ago.

"The first year of lidar on the car" - the industry describes the lidar market in 2022. With the maturity of high-performance vehicle-grade lidar in terms of performance, cost and reliability, and the ability to meet the demand for mass production of automobiles, the market awareness and application scale of lidar will begin to show explosive growth in 2021. According to the incomplete statistics of the "Caijing" reporter, there will be 17 models equipped with lidar in 2022.

Models and suppliers that have recently been equipped with lidar

Data source: CICC research report, Internet synthesis

Consultancy Yole predicts that the vehicle-mounted lidar market will reach $5.7 billion by 2026. Gu Jianmin, CTO of Valeo China, predicted to Caijing that the market size will reach 50 billion US dollars by 2030, and China will play an important role in it.

A lidar battle is underway. Bao Junwei, co-founder and CEO of Innovusion, told Caijing that at present, the world's head lidar companies are four with fixed-point and mass production capabilities: Innovusion, Luminar, Hesai Technology, and Sagitar Juchuang, "Other manufacturers who have not yet obtained real fixed-point projects or have no mass production planning have basically fallen behind in this round of competition." ”

There is still room for rewriting. Is the lidar after getting on the bus really as good as advertised? Is the lidar company's chosen OEMs competing to sell enough of the relevant models? Especially when the lack of cores continues. Is the current lidar technology route the ultimate choice for car companies? Zhang Junyi, a managing partner at Oliver Partners, believes that those lidar companies that have not yet boarded the car still have a window of two or three years, "If they are not able to produce at that time, they will basically give up." ”

Along with it are bubbles and confusion. A number of industry insiders told the "Finance" reporter that the lidar market is becoming a pile market. Zhang Junyi believes that the performance, cost and accuracy of lidar from different manufacturers are not the same, and they cannot be purely based on quantity.

The wind outlet is coming

"Happy, the first lidar project invested by the fund and individuals is ready for mass production." At noon on January 26, Zhang Junyi posted this message in the circle of friends and attached a video related to Innovusion. It's been more than three years since he invested in Innovusion.

Founded in November 2016, Innovusion has R&D centers in Silicon Valley, USA and Suzhou and Shanghai in China. It is understood that Innovusion has reached a cooperation with NIO, and lidar will be installed on NIO ET7 and ET5, two models listed in the first quarter and third quarter of 2022 respectively.

Zhang Junyi introduced that investing in parts and components such as lidar is more challenging, not only to consider technology, but also to consider the possibility of large-scale mass production in the future, the controllability of costs and so on. After much deliberation, he and several of his partners themselves increased their investment in the project, proving a high degree of confidence in the field.

"Even if a company has 10 shortcomings, but 1 point convinces you, and you don't have too good choices, you may also invest in it." 」 A technology track investor told Caijing that investment in the primary market is often a matter of probability. If you encounter a particularly hot track, it is even more important to throw in first than to examine the company's reliability first.

The lidar that is getting on the car is one such track. As long as the track continues to be hot, it means that someone has always taken over the plate and withdrawn to make a profit. "As for whether the company you invested in has finally become the head, most investors are not paying much attention now." If you see that they have only installed tens of thousands of cars in total, you will have a valuation of billions or tens of billions. These companies look more expensive than in the secondary market. ”

More than three years later, Zhang Junyi's identity has changed from a co-founder of NIO Capital to a managing partner of Oliver Wilshire Consulting, and now there is little hope of casting an Innovusion. "It is difficult for famous lidar companies to invest in now, and it is not of much value to invest in some relatively early companies now, and the most secure investment is to invest in some projects endorsed by industrial enterprises." Zhang Junyi told the "Finance" reporter.

It is undeniable that although the technological invention of lidar originated overseas, these Chinese companies are the best at engineering and mass production.

Between 2004 and 2007, the Defense Advanced Research Projects Agency (DARPA) launched three DARPA Grand Challenges, and Velodyne's multi-harness mechanical lidar was born. For a time, Velodyne was almost synonymous with lidar and the only option for all-driverless technology companies, accounting for most of the market share of in-vehicle lidar.

Today, the original inventor Velodyne is no longer dominant, and the volume of lidar has undergone a change from large to small box due to changes in technical routes. Buyers have also changed, and buyers of lidar are no longer only self-driving companies, but also car companies. According to the incomplete statistics of the "Caijing" reporter, there will be 17 models equipped with lidar in 2022.

"We have indeed caught this wave of dividends from China's new energy and autonomous driving, and the demand in the domestic market will obviously come faster and more fiercely than overseas." Qiu Chunchao, co-founder and CEO of Sagitar Juchuang, admitted to the "Finance" reporter that although the demand in the field of robotics is also growing by an order of magnitude every year, it is not an order of magnitude at all with the demand in the automotive field.

According to the incomplete statistics of the "Caijing" reporter, since last year, the total financing of the global lidar market has exceeded 8 billion yuan, and Chinese companies have obtained more than 90% of it. At the end of February, domestic lidar company Sagitar Juchuang raised more than 2.4 billion yuan in financing to prepare for an IPO in the near future.

Skimming off the foam

The rapid development of china's intelligent electric vehicles has shrunk the changes in the in-vehicle lidar market to 9.6 million square kilometers. To some extent, the current state of the car lidar in the Chinese market represents the entire industry.

"Will publicize" is the most heard description of the lidar market by Caijing reporters in interviews.

On the industrial side today, "one car with X laser radar" is becoming the latest way for car companies to promote car intelligence. And "xx mass production lidar" that "won the fixed point of x car companies" and added various qualifiers is also becoming the latest way for lidar companies to publicize their technology and mass production capabilities.

A number of interviewees told the "Finance" reporter that the current piling atmosphere of car companies is serious, resulting in a strong cost orientation in the market, but the number of lidar can not represent good performance. "It's like installing multiple lidars with a ranging distance of only 50 meters, and it can't clearly identify objects 250 meters away." Bao Junwei gave an example.

An industry insider admitted to the "Caijing" reporter that some lidar companies will blur related concepts when promoting, such as "mass production". Lidar companies' customers are mainly divided into autonomous driving companies and car companies, unlike the large-scale orders of car companies, autonomous driving companies because of the small demand, even if they sign a cooperation order, they should not be called "mass production". However, when it comes to external publicity, the two are often added together.

When more and more lidar companies installed cars, the slogan of "the first year of lidar mass production" also followed.

"This is more of a propaganda tool for some lidar companies." Hesai Technology believes that although lidar manufacturers will start to deliver mass production vehicles this year, the real industry mass production outbreak should be in 2025, "Our framework orders are generally signed in a few years cycle." ”

In terms of the number of lidar deliveries, Valeo is currently the only company in the world to achieve large-scale mass production. Gu Jianmin told Caijing that as of now, Valeo has delivered more than 150,000 lidars. "More than any other player in the industry."

Laura Wrisley, Velodyne's vice president of global sales, took a similar view. Laura Wrisley told Caijing that there are still many years away from the full application of self-driving cars, and the largest and fastest-growing lidar market in the next few years will be driven by industry and robotics, intelligent infrastructure, and smart cities.

An industry consensus is that the effect of vehicle-mounted lidar has not yet been verified by the market, at least until the first batch of lidar-loaded cars enter the market two or three years before the preliminary conclusions can be drawn. "At this stage, we can't say which lidar company dominates the market."

It will take time for such a bustle to spread to ordinary consumers.

What the user can directly understand is the detection distance and resolution of the lidar, that is, how far and how clear the lidar can see. But in addition to the performance of lidar itself, how hardware-integrated autonomous driving system experience is also important.

The lower problem is to first explain to the user what lidar is. "How do you tell if the lidar installed in this car is good?" When the Caijing reporter threw this question to a prospective car owner, he was immediately asked back: "What is lidar?" ”

How to land

There is a time window, and there are variables that rewrite the lidar pattern.

Which variables? The first is the lidar company itself. At present, most of the lidar companies are start-ups, and how the "small workshop" adapts to large-scale mass production orders is a big test. In addition, if the technology is not passed, it will also face the risk of being replaced halfway through.

The second is car companies. How many lidar companies can sell and how much money they make depends more on the sales of the model itself. "This is unreasonable, not that honor students will be able to succeed." Car companies also shoulder the role of educating users, "if the same car uses two different manufacturers of lidar, ordinary users can not perceive the difference between the two, can only be judged through the experience of automatic driving, which is also closely related to the manufacturer's algorithm." Zhang Junyi believes that only when the automatic driving experience is improved, the reputation of lidar will slowly rise, and the two are bound.

The third is the supply capacity of core components such as chips, which tests lidar companies and car companies at the same time. All along, whether it is the core components of lidar, or car chips, most of them rely on imports. At present, the transmission of epidemic factors has caused a shortage of chips, and who can supply at this time is a step faster. However, this also requires the same frequency resonance of lidar companies and car companies. "Many car companies have orders that have no sales and no deliveries because of lack of cores, so even if the lidar is installed on the car, it will be very bad."

From a larger perspective, the technical route currently used by vehicle-mounted lidar is not necessarily the final choice of car companies, but the best choice at present.

At present, the on-board lidar on the market adopts the ranging principle of the time-of-flight method (ToF) method, and is mainly based on mechanical or semi-solid-state (MEMS micro-galvanometer) technology route. Lidar is mainly composed of four modules: laser transmitter, receiver, processor and laser control module. The time-of-flight method is simply that the transmitter emits a laser, controls the module to control the direction of injection, the laser will shine on the surface of the target object, and instantaneously produce a reflection, the reflected beam is received by the receiver, the processor calculates the time of the round trip of the beam of light, you can get the precise distance of this point, that is, distance = speed x time (the speed of light is known). When an infinite number of rays of light are emitted, an infinite number of points can be used to outline the object.

Specifically, the mechanical and semi-solid technology routes are suitable for two different application scenarios, and there is no strict distinction between technical advantages and disadvantages. Mechanical lidar is mainly used in unmanned vehicles, because high performance is the priority, large volume, high cost and even time-consuming and laborious manual debugging is not the most concerned thing for unmanned vehicles. Semi-solid-state lidar is mainly used in the current mass production of passenger cars, because mass production to consider the cost, to meet the standardization process available for mass production, and even the aesthetics of the car.

New technology routes are still being developed. At present, there are two main technical routes recognized as promising in the industry, one is OPA and Flash in solid-state lidar, and the other is the chipping of stacked components of mechanical lidar. But either way, there is no ability to get on the bus in the short term.

Valeo sees solid-state technology as the next trend and uses it in Valeo's near-field lidar (NFL) products. "The combination of these technologies will depend on the technical requirements of our customers and the corresponding applications." Gu Jianmin said. Hesai Technology told the "Finance" reporter that the company attaches great importance to the research and development of new technology lines, according to the prospectus of Hesai Science and Technology Innovation Board, the company's research and development investment in 2020 (as of the end of September) accounted for 64% of the operating income. "We are constantly exploring higher performance and lower cost solutions through chip-based methods, including the fact that pure solid-state lidar products are already in the mass production plan."

More application scenarios are being explored. Closed and semi-closed scenarios such as factories, warehouses, and ports, and intelligent infrastructure such as the Internet of Vehicles, rail transit, and highways, are requiring one or more lidars. Bao Junwei believes that the future of lidar is likely to be as ubiquitous as the current camera and become a necessity in life.

It will take some time to realize this future. When chatting with an interviewee, the Caijing reporter asked the other party whether he was considering buying a car equipped with lidar, "To be honest, I am still struggling with whether to buy a fuel car or buy an electric car, and I have not considered the level of lidar." ”

This article was first published in the "Automotive and Mobility" section of Caijing magazine on March 14