Chip giants are accelerating the expansion of automotive radar, and lidar replacements are about to emerge

NXP Semiconductors, a leading manufacturer of RFCMOS 77GHz automotive radar sensors, recently announced that it has signed an investment and strategic cooperation agreement with Nanjing Hawkeye Electronic Technology Co., Ltd. to expand its share of the Chinese automotive radar market.

NXP and Falcon Eye Technologies Sign Investment and Strategic Cooperation Agreement

Ecological development model reshapes the automotive industry value chain

The financial terms of the agreement were not disclosed. However, both parties promised to change the traditional business model and seek an ecological development model based on technical services, providing customers with reference design solutions based on NXP radar products and system-level technical support and services. Specifically, it includes: providing domestic Tier 1 users with the design and simulation verification of vehicle-mounted millimeter-wave radar systems, as well as the design of antennas and RF electronics; in the case of limited time, providing Tier 1 users with a full set of design, verification, and evaluation services, and even opening up back-end processing software.

"This model is actually very mature in Europe, the United States and Japan, and in the entire automotive electronics industry, but it is not mature in China." Shi Xuesong, CEO of Falcon Eye Technology, hopes that this move will bring together NXP, Falcon Eye Technology, Tier 1 and car manufacturers to help Chinese local companies cope with complex and ever-changing technological challenges and accelerate the vigorous development of China's automotive radar and autonomous driving markets.

Founded in April 2015, Falcon Eye Technology, relying on the National Key Laboratory of Millimeter Waves of Southeast University, focuses on the research, development and application of 76-81GHz automotive millimeter wave radar technology in automotive active safety driving information systems and automotive assisted autonomous driving systems.

Southeast University Millimeter Wave State Key Laboratory is the only national key laboratory in the field of millimeter waves in China, representing the highest scientific research level in this field in China. Four years ago, Falcon Eye Technology cooperated with it to establish the "Southeast University Falcon Eye Automotive Electronics Joint Research Center", which has accumulated a large number of outstanding engineering talents and rich product experience in millimeter wave multi-system radar imaging, millimeter wave new antenna arrays and millimeter wave new circuit structures.

Millimeter wave is a very difficult technology. But through cooperation, Falcon Eye Technology was able to apply two of the latest technological achievements of the National Key Laboratory of Millimeter Waves into the solution: one is the medium and long-range integrated radar antenna technology, which can double the angular resolution of radar sensors; the other is the substrate integrated waveguide (SIW) antenna technology, which can greatly expand the radar field of view (FOV) and reduce radiation power consumption.

Examples of advanced design concepts at the Southeast University Millimeter Wave State Key Laboratory

"Radar talents do not grow on trees. They are a very scarce resource, and this is true globally. So if you want to achieve global innovation, you can't just do it from the perspective of Europe or the United States. You must have the participation of the best Chinese talents, otherwise the ecosystem will be too narrow and too slow!" emphasized Lars Reger, senior vice president and chief technology officer of NXP.

Getting out of the maze

In an interview with the media, Lars Reger said that the traditional hierarchical differentiation model of the automotive supply chain can no longer meet today's automotive industry's demand for innovation. The emergence of a large number of new car-making forces and the rapid development of autonomous driving technology require a more flexible and efficient approach. This is an opportunity for cooperation between the two parties.

"The automotive industry has changed a lot recently, with a large number of startups pouring in from both China and the United States." Lars Reger believes that these companies have strong financial support, but are relatively lacking in technology and experience, and NXP is eager to provide them with reference designs and customized development. However, hundreds of Tier 1s and car manufacturers are like a "maze" for NXP, so it is very necessary to have local companies with rich experience in a certain field act as a bridge to help it quickly complete the design. The cooperation between NXP and Falcon Eye Technology is a representative example.

According to Liu Fang, General Manager of NXP Greater China Automotive Electronics Business, for the Chinese auto market, traditionally, new technologies from OEMs come from global mainstream Tier 1s, such as Bosch. However, in recent years, the requirements of OEMs, especially for electronic components, have been growing, and these requirements are precisely what some large global Tier 1s cannot flexibly customize for the Chinese market, which to a certain extent has created obstacles for the development of Chinese OEMs. Secondly, due to the protection of technology and intellectual property rights, it has also affected the rapid entry of Chinese OEMs into the international market to a certain extent.

She pointed out that China's basic research level in the millimeter wave field is actually not inferior to the international advanced level, but many enterprises and universities lack industrialization capabilities. With the help of such an industrial platform, China's local Tier 1 is likely to narrow the gap with the international Tier 1.

According to Shi Xuesong, Falcon Eye Technology has begun cooperating with OEMs and international and local Tier 1 customers based on NXP's reference design platform. Some of these products are expected to be mass-produced this year, while radar products for corners and blind spots will reach mass production next year.

Will imaging radar replace lidar?

According to IHS Markit, as the fastest growing segment in the field of autonomous driving, up to 50% of new cars will use radar technology in 2020; by 2023, China is expected to become the world's largest automotive radar market, with a 40% growth rate that is more than twice the global market, and local car manufacturers are very willing to adopt it. Among them, the driving effect of the China New Car Assessment Program (C-NCAP) is more obvious. The program requires greater popularization and innovation of radar in automotive safety-related applications, such as blind spot detection, automatic emergency braking, front and rear cross traffic detection, and precise environmental mapping.

Automotive radar is the fastest growing segment in autonomous driving

Lars Reger believes that the average price of semiconductor devices in current vehicles is around $400, but in the next 10-15 years, this figure will rise to $1,200. Of this, $400 will be used entirely for autonomous driving. If this $400 is further broken down, about a quarter will be used for vehicle artificial intelligence technology, and the remaining three-quarters will be used for sensor technology including radar. This is definitely a huge market that cannot be missed.

The figure below shows how the in-vehicle sensor system is upgraded during the evolution of autonomous driving from level 1 to level 5. It is not difficult to see that in the level 1-2 autonomous driving stage, 1-3 radar systems and 1 auxiliary camera are usually required; in level 3 autonomous driving, at least 4-6 radars and more than 4 auxiliary cameras are required; and in the level 4-5 stage, 6-10 radar systems, 6-8 auxiliary cameras and 1-3 lidars are required. 

At present, the three main technologies for the development of autonomous driving are LiDAR, camera and radar. Limited by resolution and pixels, both camera and radar still have more or less problems, but Lars Reger said that if the "imaging radar" with higher resolution can be mass-produced as soon as possible, the expensive LiDAR can be removed from the entire module, and the imaging radar with >350GFlops is one of the key investment focuses of NXP.

Imaging radar is expected to become a key technology to drive L3-L5 autonomous driving

Indeed, although radar has the ability to measure speed, distance, angle, and work all day, the lack of sufficient resolution has always been its biggest challenge. New imaging radars have been trying to solve this problem by controlling output or input in various ways, hoping to eventually present an image of the external environment with a wide field of view and a long range. Industry insiders believe that imaging radars with 4D (distance, elevation, azimuth, speed) or 5D (the first four, plus direct target classification from radar data) functions are very likely to replace lidar in positioning applications.

In a fall 2018 radar report, Yole Développement wrote that innovative startups such as Metawave and Uhnder are using disruptive technologies to meet the requirements of high-resolution sensors, such as "ultra-thin steerable beams and AI engines for deep learning or unprecedented high channel counts for high-resolution imaging radars."

However, "future technological development is not as simple as choosing one of several directions." Liu Fang believes that, on the one hand, whether from the perspective of autonomous driving or assisted driving, the multi-sensing functions of vision and radar will coexist for a long time. The development of imaging radar is still in its early stages, and it is too early to talk about replacement. Everything is still unknown. On the other hand, because the development level of China's camera technology is far higher than that of foreign countries, more and more Chinese ADAS manufacturers will use millimeter-wave radar combined with camera technology to achieve autonomous driving Level 1, Level 2 and even some Level 3 functions.