Xinna Sensing: A "new" show of high-precision positioning standing on the shoulders of giants

The global craze for autonomous driving continues. Not long ago, VinFast, which was founded only five years ago and is known as the "Tesla of Vietnam", announced the launch of its listing plan and officially submitted an IPO prospectus to the U.S. Securities and Exchange Commission (SEC), planning to be listed on Nasdaq.

According to official news, VinFast has received approximately 65,000 orders worldwide, and its two medium and large pure electric SUV models have begun to be delivered globally. The new vehicles incorporate multiple autonomous driving perception sensors, and the vehicles will reach L3 assisted driving capabilities. . Among these many sensors, Gasgoo found out that the indispensable high-precision positioning system comes from Xinna Sensing, a company headquartered in Wuxi, China. The INS combination positioning system INS401 provided to VinFast has been selected for four models and mass-produced for shipment.

In addition to VinFast, Gasgoo further discovered that Xinna Sensing has also won multiple model appointments from leading passenger car companies in China and achieved mass production.

So, as a company that keeps a low profile in the industry and focuses on mass production, how can Xinna win the favor of domestic and foreign OEMs? Recently, Dr. Zhao Yang, CEO of Xinna Sensing, was interviewed by Gasgoo and shared Xinna’s “past and present life”.

Going back 10 years ago, many people may not have heard of autonomous driving, but at that time, MEMSIC Semiconductor, founded by the global leader in the microelectromechanical systems (MEMS) industry, had already developed autonomous agricultural machinery using its solutions. Into the American fields. At that time, Dr. Zhao Yang, as the founder, deeply realized that “all these things will be indispensable core components in future cars.”

Centimeter-level high-precision positioning enables autonomous driving of agricultural machinery; Source: Xinna Sensing

Today, 10 years later, as cars accelerate towards the "autonomous driving" era, this very forward-looking prediction is becoming a reality. Zhao Yang also took this opportunity to start his second entrepreneurial venture in his life, establishing Xinna Sensing.

Start a second business

In the company introduction of Xinna Sensing, there is this paragraph: "Xinna was spun off from MEMSIC Semiconductor Company in 2017. MEMSIC was founded in 1999 by Dr. Zhao Yang, CEO of Xinna, and has been committed to Designs and produces advanced MEMS sensors, and was the first company in the world to be listed in the United States specializing in MEMS sensors."

So, why choose to spin off operations? What energy does MEMSIC give to Xinna, and what are the business differences between the two?

"MEMSIC is a sensor chip company. Everything is made independently, including equipment, process manufacturing, software algorithms, etc. Xinna was spun off from MEMSIC. In principle, the two are based on core technologies. It’s similar, but the products are different,” said Dr. Zhao Yang. After the spin-off, MEMSIC's products focus more on consumer electronics, while Xinna focuses on automobiles and industry.

Dr. Zhao Yang said that the reason for establishing Xinna was to concentrate resources and energy on the automobile and industry with higher product requirements. "Consumer businesses such as mobile phones have large volume and low unit prices, and can be fast in and out. However, the unit prices of automobiles and industry are high and the system stability requirements are also high. We must concentrate on that."

In addition to the inheritance of core technologies, MEMSIC has given Xinna a lot of practical experience in the development process of an enterprise from 0 to 1. As the founder of two companies, Dr. Zhao Yang said frankly that from his personal perspective, if he does the same thing again, the mistakes made in the past can be effectively avoided, successful experiences can be learned, and the whole process will be smoother.

Xinna vehicle-grade high-precision positioning products and solutions; Source: Xinna Sensing

Standing on the arms of MEMSIC's more than 20 years of technology accumulation, Xienna Sensing has developed a rich product line with persistent innovation since it became independent in 2017, including high-frequency and high-bandwidth current sensors based on magnetoresistive films, High-performance code open source inertial measurement unit (IMU), high-precision real-time kinematic (RTK) navigation system and centimeter-level precision positioning services. Its high-precision positioning software and hardware one-stack solution has complete self-research capabilities from modules to systems, from hardware to algorithms, and has achieved extensive mass production in passenger cars, commercial vehicles and precision agricultural autonomous driving. Application and verification.

Dr. Zhao Yang pointed out that Xinna is growing very fast now and will continue to invest in developing more new products in the future.

vital inertial navigation system

In the development process of autonomous driving, vehicle-level high-precision positioning is the key support for L2+ to L5-level autonomous driving, and the high-precision inertial integrated positioning system (INS) is the core of achieving high-precision positioning.

Judging from the current actual situation, with the development of domestic autonomous driving technology so far, most mass-produced vehicles have only achieved L2 and L2+ assisted driving, and there are still only a few that achieve L3 functions.

In this regard, Dr. Zhao Yang pointed out that it is not difficult to realize autonomous driving in a good environment. What is difficult is how to do it in extreme environments. To achieve autonomous driving in all climates, all bands, and in any scene, the current The technology is far from there yet.

As far as the inertial navigation system required for autonomous driving is concerned, the current L2 level autonomous driving does not require high accuracy for the inertial navigation system. Once it reaches high-level or even fully autonomous driving, very high-precision sensors are necessary. This is crucial.

Dr. Zhao Yang emphasized: "The inertial navigation system is a national strategic technology. Countries are competing for its accuracy. It is the hardware, not the software, that determines the performance of the high-precision positioning system." Just like a computer, If the hardware is not in place, no matter what software is used to optimize it, it will not reach the ideal speed. Therefore, the hardware determines the system. The better the hardware, the simpler and more reliable the design of all aspects of the software will become.

Therefore, for the high-precision inertial integrated positioning system (INS), the strategy adopted by Xinna is to continuously improve the performance of current products while increasing investment in self-developed next-generation IMU chips.

It is understood that the IMU is the only sensor in the car that is not affected by the external environment and can work normally under various conditions. Especially when the GNSS signal is weak and sensors such as lidar and cameras are affected by bad weather, the IMU will become the last protection for vehicle positioning.

Xinna's next-generation self-developed chip; Source: Xinna Sensing

It is reported that in the past few years, the bias instability of IMU has been improved by nearly an order of magnitude, which can basically meet the current high-precision positioning requirements of L2+ autonomous driving. The performance of Xinna's next-generation self-developed IMU chip will be greatly improved and can even match Positioning requirements for L5 level autonomous driving. "Currently, there are few companies in the world that can make IMU MEMS chips, and there are currently no chips that can achieve such high performance. We not only independently develop sensors, mixed signal processing ASICs, and core algorithms, but also independently produce semiconductor sensor chips. Meixin has strong enough experience to support it,” Dr. Zhao Yang said confidently.

In the context of the technological competition between China and the United States, it is rare for Xinna to be a domestic company capable of designing and manufacturing core chips for inertial navigation systems. Faced with some restrictive policies of the United States against China, Dr. Zhao Yang pointed out that if Xinna wants to take root in the Chinese market, it must adopt localized operations: Xinna is headquartered in Wuxi, China, and has a R&D and production base, fully equipped with a complete core The full-process operation capabilities of R&D, engineering design, manufacturing, technical support, and after-sales service have been completed and are in normal operation. At the same time, Xinna has global R&D centers in Boston, Chicago and Silicon Valley. According to reports, Xinna has full-stack self-research capabilities, covering different fields such as chips, modules, IMU algorithms, RTK algorithms, and fusion algorithms.

It is worth mentioning that based on full-stack self-research, Xinna’s cumulative shipments of inertial navigation products have exceeded 1 million units.

Technological innovation is the only way out

At present, although there is still a long way to go before L4 or even L5 is achieved, there is no doubt that the global trend of autonomous driving is developing. Among them, car-grade high-precision positioning, which is a key support for autonomous driving, is also attracting more and more new players, including traditional high-precision positioning companies and cross-border players from auto parts suppliers. There are also entrepreneurs who started from scratch, all trying to get a share of this emerging market.

Regarding new players on the car-grade high-precision positioning track, Dr. Zhao Yang said bluntly, "To ensure profitability and survival, the only way out is technological innovation." Because car-grade products will have more stringent performance and specification requirements, This will lead to an increase in product costs.