Breaking through the technological blockade, how can domestic millimeter-wave radar stage the "rise of a great power"?

Source: Cheyun

Author: Huang Rongzhao (Investment Director of Intelligent Vehicles at Fortune Capital)

If we compare the development of millimeter-wave radar in China in recent years to a puzzle, people who are interested in it take the small piece in their hands, observe the piece in others' hands during communication, and then try to put it together to depict its full picture. As a person in the industry research and investment industry, I have been fortunate to collect many pieces in recent years. On this summer night, I will put together my picture for everyone to see.

It is said that in the 1960s, automotive radars were already being developed in the United States, but they were just 10GHz transmitters and receivers, and they were presumably built with discrete components, so the availability was not very good. At least by the 1990s, the big names in Japan, the United States, and Germany had already begun systematic research, and gradually began to use them in commercial vehicles. After 2010, the shipments of A (Autoliv), B (Bosch), C (Continental), and D (Delphi) basically reached tens of millions. Of course, the term ABCD is for easy memorization. In fact, Hella, Denso, Fujitsu, TRW, and Valeo all have strong strengths.

▲ Market share of millimeter-wave radar of each company, source: CITIC Securities

Back to China, the earliest millimeter wave radar developers in China were mostly from the military and universities with military backgrounds. Most of the universities were Southeast University, Xidian University, Chengdu University of Electronic Science and Technology, and the Chinese Academy of Sciences. Most of them were security radars, ship radars, etc.

It can be said that in 2014, Mobileye, which had been silent for 10 years in research and development, went public and entered the public eye, with a market value of over 10 billion US dollars, which can be said to have opened the enlightenment movement and the first year of science popularization of China's ADAS. In 2015, Baidu's driverless car was released, shouting the slogan of "three years of commercial use, five years of mass production", which completely ignited people's beautiful imagination of driverless driving, and various entrepreneurial teams surfaced. Mixed with various discussions on L2, L3, and L4, people and the capital market's enthusiasm quickly jumped from ADAS to driverless driving, completing the shock of driverless driving from the automotive circle to the Internet circle. It took only a few months to complete the science popularization of the investment circle and the media circle. In this way, cameras, millimeter-wave radars, and lidars became the focus of everyone's discussion.

To a certain extent, Micro Smart is a prototype product of many domestic radar manufacturers. This German company provides an overall radar solution, and several traditional automotive Tier 1 companies have purchased solutions from it and paid high licensing fees. Later, it was discovered that there are many "wonderful" parts of millimeter-wave radar, and vehicle-mounted radar has more and more functions. This licensed production business model limits the flexibility of development and is increasingly unable to meet new market demands. At the same time, cost reduction under this business model is also a big challenge.

A few years ago, an article said that 77G chips were banned, which seriously misled a lot of people about 77G. Now the giants have already started to compete fiercely. Infineon, Ti, ST, and NXP have their own characteristics and their voices are very loud. Infineon said, I am the originator of the industry, and the Infineon-Bosch alliance has shipped more than 20 million pieces. Ti said, I have CMOS technology, high integration, high cost performance, and I am willing to support you little brothers.

From the author's observation, Infineon is limited by the Infineon-Bosch alliance and its domestic promotion is limited. NXP's chips are also less used due to their high prices. Several major startups have adopted Ti's solution. From the perspective of product strength, the integrated solution reduces the difficulty of development and greatly reduces costs. On the other hand, Ti's active support and lobbying have also made it a large number of supporters. The cost of the chip is basically in the range of US$10-20. The higher cost performance may help Ti become a new winner.

Special mentions here are Xiamen Yixing and Calterah. As domestically produced chips, and in addition to being in the automotive industry, it is always difficult to get started, but product strength is the ultimate measure, and at present, the momentum is good. After returning to China from the automotive chip giant Infineon in 2011, Dr. Yang Shoujun of Xiamen Yixing began the research and development of millimeter-wave radar front-end RF chips (MMIC). With outstanding R&D capabilities, products were soon launched. In 2015, when domestic radar vendors and solution providers had not yet started, it was very difficult to simply promote the chip market. Xiamen Yixing made various attempts, continuously sent samples, and patiently cultivated customers. At present, I feel that the company's transformation goal can be compared to Smart Micro, turning the most difficult part of the radar into a solution, better developing support, lowering the technical threshold, and promoting faster product implementation.

The Gatland team, after returning from the United States, has achieved certain results in the 77GHz RF CMOS process, but is currently facing the problem of improving integration and the lack of support from solution providers and radar developers. The author believes that the idea of ​​downstream and solution integration is also the best choice for RF chip companies to survive and break through in the current status of the millimeter wave radar industry.

There is no doubt that the entire millimeter-wave radar industry will be in the incubation period in the past three years and at least in the next two years. Companies will be in a situation where they will burn money on R&D and patiently accept car factory projects. The current situation is that there are no less than 30 radar departments of startups and mature companies. Those who cannot get financing will face a serious survival crisis. During the millimeter-wave radar financing boom in 2015-2016, some of the money raised has basically been spent. Therefore, the author believes that there will be a concentrated financing wave for ADAS startups in the near future.

Currently, companies engaged in millimeter-wave radar can be basically divided into two categories: one is the newly formed product departments of existing automotive supply chain companies, and the other is the entrepreneurial team.

Let's first talk about the newly formed product departments of automotive supply chain companies: I think China's automotive electronics industry can also be divided into the southern and northern factions. The northern faction can be called the traditional automotive faction, which is mainly centered in the Yangtze River Delta and focuses on safety, communication, and control. The southern faction can be called the electronics faction, or simply the car computer faction, which is mainly located in the Pearl River Delta and has grown up in the past decade by focusing on navigation and audio-visual entertainment systems. In recent years, they have also been listed in large numbers.

Traditional Tier 1 companies are facing the pressure of reducing costs every year, and they are racking their brains to launch new products and occupy a better position. Millimeter wave has become the easiest choice for them to understand. Judging from the announcement, Desay SV, Huayu Automotive, and Tongzhi Electronics are all actively deploying. Of course, there are many companies that I am not convenient to disclose, and they are almost all actively making technical reserves. The first idea of ​​these companies is to buy solutions first, trying to quickly build products and gain the upper hand. But from the results, the progress of the products is always less than expected, and it is estimated that the mass production of the products is still facing difficulties. After all, the veterans of these companies are familiar with automotive electronics and EMC, but the talents of microwaves are all in the military and communications industries. Traditional automotive electronics companies have never played with them, and they always have to step into a few pits once they get started. On the other hand, it is difficult to reduce costs. If you do forward research and development, it is still difficult to reserve talents. From RF, antennas to signal processing, to control strategies, etc., there is a lot of investment and accumulation.

Besides, most of the first batch of Chinese millimeter wave radar startups were established in 2014-2016. The team often has a background in scientific research, military industry, or suppliers who changed jobs and started their own businesses. For example, the overseas returnees: Sensortech, Hangzhou Zhibo. The research institutes: Xingyidao, Nanjing Falcon Eye, Suzhou Anzhi, Suzhou Haomibo. The industrial transformation: Shenzhen Anzhijie, Shenzhen Chengtai, Hunan Naray. There are also Yilaida, Muniu, Raybotek, Shenzhen Zhuoying.

There are about 20 millimeter-wave radar startups in the country, and no less than 10 of them have received financing. The characteristics of these companies are also very obvious. Generally, those who do not have a doctorate are embarrassed to open a store. The R&D foundation is relatively strong, the execution is good, and the entrepreneurial idea is often to take the initiative to develop and bite the hard bones. The positioning is also mainly concentrated on solution providers or directly aiming at the new Tier 1. The shortcomings of these companies are also obvious. The first is the lack of money. If you want to do a good job in millimeter-wave radar, it is quick to make a demo, but you have to consider system construction, market promotion, vehicle adaptation, system design, calibration and stability, process difficulty and production stability, which is an investment of more than 50 million. If you want to be a Tier 1, it is indeed like this. There is no small step and fast running in the Internet industry. Only when the system is perfect and excellent enough can you knock on the door of the OEM.

▲List of major domestic millimeter wave radar companies, compiled based on public information

At present, according to the author's observation, the companies in the entire industry that are in mass production and shipment are mainly Shenzhen Anzhijie and Wuhu Sensortech. Anzhijie is mainly testing the waters in the aftermarket, and has obtained thousands of sets of BSD and LCA radar shipments. According to the company, it has also obtained 2 domestic OEMs and 4 projects. Sensortech has benefited from the strong support of Hikvision. According to Hikvision's annual report related-party transaction announcement, it has obtained more than 30 million in revenue in the security field.

▲Source: Hikvision 2017 Annual Report

The millimeter wave market is always puzzling. Through grassroots research, the author has an analytical idea: the current mass production applications of millimeter wave radars are nothing more than forward ACC, AEB functions, and rear BSD, LCA functions. According to the data of MEMS Consulting Company, 41 million units have been shipped worldwide. If an average car uses 2-3 units, that is more than 10 million cars, and the global penetration rate is about 20%. MEMS also said that domestic shipments are 1 million units, which is about 500,000 cars, and the penetration rate is about 3-5%.

In 2016, MEMS Consulting estimated that the number of pre-installed millimeter-wave radars in Chinese cars reached 1.05 million, of which 24GHz radars accounted for 63.8% and 77GHz radars accounted for 36.2%. 24GHz radars are now mainly used for blind spot detection (BSD), and the market demand comes from Chinese brand cars; 77GHz radars are mainly used for adaptive cruise control systems (ACC), and some companies also use them for forward collision warning (FCW) and automatic emergency braking (AEB). According to the China New Car Assessment Program (C-NCAP), the automatic emergency braking system (AEBS) will be included in the scoring system in 2018, so the demand for 77GHz radars will increase. In addition, looking forward to the future, the smaller size and longer detection distance of 77GHz radars will squeeze the market space of 24GHz radars.

Here we focus on analyzing the domestic millimeter-wave radar market. With the five-star rating of C-NCAP and the implementation of the Ministry of Transport's 1094 document on commercial vehicles, the domestic market will enter a stage of rapid growth. In recent years, it is even expected to grow exponentially. The current penetration rate has reached 5%. According to experience, when a product breaks through the 5% penetration rate, it will often enter a period of rapid growth. It is estimated that around 2020, if the market penetration rate can exceed 15%, the market shipments will exceed 3.75 million. For front view + rear view, the average value of a single vehicle is 800-1000 yuan, and the market space is at the level of 3-5 billion. If the assembly rate is further increased, the market space will be even larger.

Internationally, the AEB configuration rate of new cars in the US market in 2017 was 19%. According to the NHTSA survey, only 19% of vehicles listed in the US in 2017 were equipped with AEB automatic emergency braking systems. In terms of quantity, the Japanese automaker Toyota has the most. Due to the large overall sales volume and 56% of vehicles equipped with AEB, 1.4 million of the 2.5 million Toyota products produced in 2017 were equipped with AEB, which is much higher than other brands; General Motors ranked second, with 551,777 of the 2.8 million vehicles produced in 2017 equipped with AEB, a ratio of about 20%; Honda ranked third, with about 30% of its vehicles equipped with AEB, and 492,330 of its 1.6 million vehicles equipped with AEB.

In terms of proportion, among the car manufacturers equipped with AEB systems, the top few with the highest ratios are not surprisingly all luxury brands. Among the cars produced in 2017, the electric car manufacturer Tesla ranked first, some of which were equipped with AEB, while others were enabled by the cloud download system after leaving the factory. A total of 99.8% of Teslas have AEB functions, including the 2017 Model S and SUV product Model X. In addition, among the cars produced by German car manufacturer Mercedes-Benz in 2017, as many as 96% of the cars were equipped with AEB, followed closely by Audi with 73%, Volvo with 68%, and BMW with 58%.

▲NHTSA announced the proportion of AEB configuration of each brand in 2017

Among the new models in China, according to the author's statistics through the configuration table of Autohome, there are also some findings. From the installation of forward and rear radars, these joint venture car companies will rely more on the safety function of forward braking. Relatively speaking, American and Japanese cars are ahead. Among them, brands such as Mercedes-Benz, Volvo, and Lynk & Co. have already equipped all series with millimeter-wave radars as standard. Fortunately, the configuration of millimeter-wave radars has been extended to the mid-range of B-class cars, and the latest generation of Camry is one of them.

▲Vehicle models equipped with millimeter-wave radar by brand

Among the new models in China, according to Infineon statistics, the following models have been equipped with BSD functions.

▲List of models equipped with BSD, source: Infineon

In this section, the style of writing changes greatly because a large amount of data is cited. In general, the author also observed that domestic OEMs generally work on millimeter wave radar projects. It basically shows that first-tier OEMs first set benchmarks and learn from experience, and then seek replacement. Second- and third-tier OEMs are even more radical and directly specify domestic suppliers.

Obviously, from the perspective of the competition landscape, the current focus of domestic radar manufacturers is not on domestic peers, but on the encirclement of international Tier 1. International Tier 1 has mastered mature radar technology and also controls key links such as chassis, brakes, and control. It seems that the competition barriers are already unbreakable, and the market generally lacks confidence in domestic radar manufacturers.

The localization of reversing radar and car machine has great reference significance. Before and after 2010, the pre-installation rate of reversing radar was less than 5%, and the market was almost dominated by Bosch and Valeo. However, after Audi made a breakthrough in sensor technology, integrators such as Haun started quickly. Local companies often have the advantages of strong service capabilities, fast response speed, and high cost-effective products. Start with small customers, accumulate experience, expand scale, improve quality, and gradually seize important customers to grow. Now, the pre-installation rate of reversing radar has increased to a mature stage of 70%. Domestic first-line independent brands have basically cut into domestic suppliers. Haun also cut a certain share of Volkswagen this year.

Summarizing the experience of localization of reversing radar and car machine, we believe that the localization of millimeter-wave radar will definitely come, and cost-effectiveness and emphasis on service may become the magic weapon for domestic millimeter-wave radar to break through. Surrounding the city with the countryside is currently the most reasonable idea. At present, the international Tier 1 has been alerted and has directly hit the price to a micro-profit model, and is not as concerned about development costs as before. However, the main focus of the international Tier 1 is on first-tier car manufacturers, and small and medium-sized car companies with shipments of about 200,000 units are in a vacuum. By surrounding the city with the countryside, the domestic tier 1 will first train the team, expand the scale and complete the quality improvement. At that time, the domestic Tier 1 will adapt to the characteristics of the large number of models of domestic OEMs, and the advantages of cost-effectiveness and emphasis on service will be fully utilized, and it will be expected to cut into the supply chain of first-tier car companies. After all, the competition is so fierce, and cost reduction is always the theme that OEMs hold tightly.

The author believes that once the quality of domestic Tier 1 is recognized and they gain a firm foothold in the OEM market, the flexibility of localized R&D will become a major advantage. The progress of domestic Tier 1 in recent years is still worthy of recognition. Although there is still room for improvement in original technology, the ability to develop applications is the strength of domestic Tier 1.

The past few years of millimeter wave development have been a time of fierce competition, a time of hard work, and a time of ups and downs. In the future, there will be years of staking out territory, years of growth and loss, and years of domestic companies striving to rise and giants suppressing them.

Looking back at the past few years of millimeter-wave radar, we always look forward to the years when reversing radar and car central control appeared. As a Chinese and an investor, we also hope that domestic companies can seize the opportunity in the wave of automotive intelligence and gradually grow into large companies.

"Xinshiye's media"