Software-defined cars have become a trend? How should traditional car companies change?

The smart driving track, which seemed extremely sexy at first glance, has experienced ups and downs and entered its true highlight moment. After a wave of investment, the industry has returned to rationality, and only those companies with real technical strength remain. The story of smart driving is getting more and more exciting: the leading companies in the track are proving the value of technology with mass production performance; players in the automotive field are also entering the field of autonomous driving with a more suitable attitude for themselves, welcoming the era of software-defined cars.

From July 12 to July 14, the 2019 Fourth Global Artificial Intelligence and Robotics Summit (CCF-GAIR 2019) was officially held in Shenzhen. On the first day of the conference, after a full day of "Artificial Intelligence Frontier Special Session" and "Artificial Intelligence 40th Anniversary Special Session", the 2019 AI Best Growth List Award Ceremony was grandly held at the CCF-GAIR 2019 dinner.

Compared with previous years, the 2019 AI Best Growth List is more authoritative and representative. After three months of registration and nomination, and 15 days of concentrated expert review, the 436 AI startups nominated and applied for the list were finally selected for the 13 most popular AI + fields, including the Best Barrier Growth Award, Best Product Growth Award, Best Commercial Growth Award, and Best Future Growth Award. Finally, 52 AI companies with the most representative product capabilities, technical capabilities, commercial value, and future growth potential were selected. Among them, in the highly-watched AI + automotive field, four major companies, Moshi Intelligence, Newtech, Moran Cognition, and GAC, won awards.

Best Commercial Growth Award - Magic Vision Intelligence

Magic Vision Intelligence is a supplier that realizes mass production of autonomous driving based on deep algorithms.

For autonomous driving systems, environmental perception and vehicle positioning capabilities are key. Magic Vision Intelligence has done a lot of work in environmental perception and vehicle positioning, and has ranked first in important international competitions in the field of deep learning algorithms many times.

Based on deep learning-based forward-looking and multi-view surround vision perception, and advanced VSLAM technology, Magic Vision Intelligence supports L0~L2 level advanced driver assistance systems (ADAS), and can provide LDW, FCW, PCW, SLI and other functions; supports autonomous/automatic parking solutions based on 3D surround vision and deep learning, and can identify the environment around the vehicle in real time, and output information such as vehicles, pedestrians, obstacles, parking spaces, lane lines, etc.; and based on deep learning and visual SLAM L3, L4 high-level autonomous driving solutions, the world's top visual perception, fusion, positioning, path planning and control algorithms are implemented on embedded low-power automotive-grade chips, which can realize autonomous driving under different scenario conditions.

Moshi Intelligence is focused on achieving pre-installed mass production of passenger vehicles in China and has reached cooperation with many OEMs.

Best Product Growth Award - Newtech

As a technology company focusing on the field of autonomous driving, Nullmax, founded in 2016, is committed to using advanced computer vision technology, deep learning technology and artificial intelligence technology to provide the world with safe, efficient and economical autonomous driving solutions.

After more than two years of development, Newtech recently launched "MAX 1.0", which is very similar to Tesla's Autopilot and can realize three scene functions: highway driving (HWP), traffic jam following (TJP), and autonomous parking (AVP). The hardware configuration of the MAX solution consists of 12 cameras, 5 millimeter-wave radars, 12 ultrasonic radars, 2 ECUs, DMS, and GPS+IMU, and the software system consists of the MaxOS operating platform and core algorithms independently developed by Newtech. The solution is low-cost and can meet the perception accuracy required for highways and urban expressways.

In addition to the MAX solution currently released, which focuses on L3 autonomous driving based on visual perception, Newtech is also developing L4 autonomous driving solutions. Its goal is to achieve mass production of L3 and L4 autonomous driving solutions around 2020.

Best Future Growth Award - Sudden Cognition

Voice interaction has become a standard feature of smart cars. Moran Cognition tends to integrate voice user interface and graphical user interface to achieve a more natural interaction method that is more in line with human senses.

So far, Moran Cognition is promoting the fusion of long-term and short-term memory in devices through "streaming" interaction. Once the fusion is formed, there will be two major benefits: one is that it can run across devices, and the other is that the application will be relatively free in the process of device operation. More importantly, such a product form is a manifestation of greater respect for users.

In addition, Moran Cognition is also building knowledge graphs based on massive big data to provide users with more intelligent decision-making services in life encyclopedias, travel assistance, etc. It has a full stack of core technologies including semantic understanding, dialogue system, learning system, service docking, speech recognition, language generation, noise reduction, voiceprint recognition, command words, etc., all of which are independently developed by the Moran Cognition team.

As a supplier of in-vehicle system products, Moran Cognition not only provides core technologies such as voice and semantics to car manufacturers, but also connects to more than 50 life travel services, which coincides with the idea of ​​car manufacturers to build a whole vehicle ecosystem. The result of the project cooperation between Mercedes-Benz Passenger Cars China R&D Center and Moran Cognition - providing technical support for voice purchase of movie tickets for the former, will also be equipped in the MBUX intelligent human-computer interaction system of many future Mercedes-Benz models, and is expected to be officially launched in 2020.

Best Barrier Growth Award - GAC

In response to the development trend of electric, intelligent, lightweight and emotional automobiles, GAC Group established the GAC Research Institute in 2006. Currently, there are more than 4,000 R&D personnel in this field, and about 200 employees are engaged in the research and development of autonomous driving.
According to Leifeng.com, in 2013, GAC developed the first driverless concept car WitStar. Currently, the car has developed to the second generation. In June 2018, the "Qiyun" concept intelligent connection system jointly developed by GAC Trumpchi and Tencent was officially released, and the GAC Tencent Car Connection AI system that was born immediately was also regarded as the industry's cutting-edge in-vehicle AI system.

At present, advanced driving assistance systems, including ACC, FCW, AEB, LKA, SVM, etc., have been mass-produced in GAC Trumpchi GS8, GM8, GS5 and GAC New Energy Aion S. Among them, GAC Aion S is equipped with the world's leading Adigo autonomous driving, which is the first mass-produced L2 autonomous driving product in China. Adigo integrates ICA integrated cruise, APA automatic parking and other auxiliary systems; it is equipped with AI system version 2.0, the latest Tencent car connection system built-in multimedia, with intelligent voice interaction function, etc., and the whole system can be iteratively upgraded by OTA.

A GAC New Energy Aion LX equipped with an L3 level automatic assisted driving system was also officially released at the 2019 Shanghai Auto Show, which can realize functions such as automatic driving and valet parking on highways. For L4, GAC has already cooperated with pony.ai to carry out demonstration operations of L4 driverless taxis. Just last month, GAC also obtained the first batch of intelligent networked car road test licenses in Guangzhou.

The wave of software-defined cars brings opportunities and challenges to technology companies and traditional car companies. Facing intelligence, digitization, and automation, even industry giants still need to make many profound changes in consciousness, strategy, organization, and mechanism to reshape the new ecology of the automotive industry and new ways of playing with autonomous driving. And the massive entry into the era of automobile iPhones is not a simple copy, but a new editing of IT genes and the inherent genes of automobiles to evolve into a new species. On this road, the leading players have completed a gorgeous transformation.