A large number of new technology companies have poured into smart driving and even driverless cars under the enthusiasm and support of capital. Xiaoma Zhixing, which has been established for less than two years, has rapidly developed into one of the more prominent enterprises. Different from traditional car companies. The iterative order is implemented step by step. The new technology companies represented by Xiaomazhi's behavior often choose a higher starting point directly, and higher-level technologies cut into this field.
A total of about 230 million U.S. dollars in financing, the first to obtain the highest level (T3) autopilot road test license in Beijing, the recent good news of the small Ma Zhixing has become the strongest domestic and the highest value of unmanned driving Startup company.
Just in July, Xiaoma Zhixing announced the completion of the $102 million A1 round of financing, plus the $112 million financing announced in January. The total round of financing in the A round reached $214 million. It is understood that this round of financing has gathered Xiaoma Zhixing. A group of new and old shareholders: Yan Ming Investment and Fidelity International Co., Ltd.'s own investment institutions, Stora Capital, investment, Songhe Capital, China Merchants Capital, Red Dot Ventures (China) and Delong Capital, etc. 2. Old shareholders Sequoia Capital (China), Morningside Capital, DCM Venture Capital, Hongtai Capital, etc. also invested in this round.
Just in July, Xiaoma Zhixing announced the completion of the $102 million A1 round of financing, plus the $112 million financing announced in January, and the total round of financing in the A round reached $214 million.
What the capital side values, is that Xiao Ma Zhixing is the top team in the industry, the technical solution for autonomous driving and the consideration of the marketization path.
According to reports, Xiaoma Zhixing's founding team has two leading experts in the industry: the original Baidu chief architect, the current Xiaoma Zhixing CEO Peng Jun and the original Baidu's youngest engineer Lou Tiancheng. Among them, Peng Jun used to be at Google. After 7 years of service, he was the earliest pioneer of Baidu Meiyan. The internal technical level was T11, while Loutiancheng was Baidu's youngest T10. Not only these two people, many members of the Xiaoma Zhixing team are from the world. The unmanned vehicle team of top companies (such as Google, Ford, NVIDIA, etc.) has the system development capability of unmanned technology in all directions. The speed of team expansion is not unpleasant. The A1 round of financing is only six in the last round. Month, but in just half a year, the number of teams in Xiaoma Zhixing has more than doubled. In the United States, 80% of employees in Beijing and Guangzhou are technicians.
Technically, Xiao Ma Zhixing chose to develop L4 and above unmanned vehicle technology. According to Zhang Ning, technical director of Xiaoma Zhixing, Xiaoma Zhixing has been developing self-driving systems, including perception and prediction. , planning and control, etc., the entire program is based on sensor fusion plus high-precision positioning and control plus high-precision map algorithm. The advantage is that the technology iteration is very fast, but also make full use of the strengths of its own development system engineering, ' The system capability of the software is very important. From one or two prototypes to one hundred vehicles, the requirements for thousands of unmanned vehicles are different.
As for why he chose to go straight to the L4 level of unmanned driving, Peng Jun has publicly explained his reasons. Although this is the most difficult one in the development of existing autonomous driving technology, it realizes the L4 level of unmanned driving to humans, to the world. The impact is also the biggest. In his eyes, the current L3 level autopilot is a lot of manufacturers' gimmicks, not suitable for commercial scale use. From L2 to L3 to L4 is a gradual process, each level needs The technology, including sensors, is not a small difference, 'L3 can only be used as a transitional phase in the L4 development process.'
Getting a road test license for a new technology company like Xiaoma Zhixing is not only an recognition of its technical capabilities, but also an important part of the company's strategic layout. It is understood that as early as June 2017, Xiao Ma Zhixing obtained the California Road Test License, which realizes 24-hour day and night traffic flow automatic unmanned driving on the open city road. It is the first in the world to achieve unmanned coverage of 10 square kilometers. s company.
However, Zhang Ning admits that the experience of American road test is not enough to cope with the complicated scenes in China. Because of the habits of domestic traffickers, various situations such as the frequency of pedestrians and bicycles appear in an endless stream, which is completely different from the United States, for example, in Chinese traffic. Typical scenes include a variety of lane changes, 'gasser', retrograde, etc., pre-judgment, positioning, and perception of unmanned vehicles have very high requirements. It brings us a very big challenge.
'At present, the identification of vehicles by unmanned vehicles may be 95% or even close to perfection. But for pedestrians, the recognition of bicycles can be 70% better. How to deal with the remaining 30%?” Zhang Ning believes that It is one thing to pursue accuracy in dealing with various emergencies encountered on the road. Pursuing safety, pursuing high reliability, and low recall are another matter. It is necessary to build a brain to cope with various challenges. .
On the eve of the Spring Festival in 2018, Xiao Ma Zhixing landed a small unmanned vehicle team in Nansha, Guangzhou, becoming the first start-up company in China to open the driverless test experience to the public. For several months, Xiaoma Zhixing’s self-driving vehicles have It has received over 1,000 passengers and covers a public section of the city within a range of about 30 square kilometers. The roads include the surrounding areas of the district government, the public square, Wanda Plaza and other complex roads.
Not long ago, Xiao Ma Zhixing won the T3 class driverless road test license (the highest level) issued by the Beijing Municipal Government, and became the first start-up company to obtain the Beijing driverless road test license. The other has already obtained The same level of Beijing test license is the Internet giant Baidu.
It is not easy to get a road test license in Beijing. According to reports, the T3 license issued by Xiao Ma Zhixing is the highest-level road test license currently issued in China. According to the first domestic specification on automatic driving, Beijing The guidance on accelerating the work related to the road test of self-driving vehicles (for trial implementation) is awarded, and the process obtained is comparable to that of ordinary people who obtain motor vehicle driving qualifications. It needs to undergo closed test field training, automatic driving ability assessment and expert review. Procedure. Examine the vehicle's awareness of roads, marking lines and ancillary facilities, and review the ability to properly operate the vehicle under specified driving scenarios, including but not limited to automatic driving, automatic shifting, automatic braking, automatic monitoring of the surrounding environment, automatic lane change, Auto-steering, automatic signal alerting, etc. Third-party authorized agencies also need to examine whether the driver can intervene and take over the driving behavior of the vehicle anytime and anywhere. Obtaining this license means that the company's self-driving car has cognitive and traffic law compliance. , route execution, emergency response and other comprehensive capabilities. It took only ten days, Xiaoma Zhixing completed and passed A series of Beijing T3 licenses for examination and assessment requirements, including 5,000 km of designated closed-cavity fully automated driving tests. During the road test, the self-driving car modified by Xiao Ma Zhixing demonstrated strong environmental adaptability, in different Under the road conditions, the correct judgment has been made, and some unexpected situations can be calmly judged and dealt with.
At present, Xiaoma Zhixing has realized day and night on the central roads of the cities of China and the United States, and mixed traffic between people and vehicles. The automatic driving road test under the whole scene has achieved automatic automatic operation without manual intervention during peak hours. Driving, and making reasonable decisions in the event of encounters with other motor vehicles or bicycles, retrograde, traffic jams, vehicles and pedestrians.
As a start-up technology company, the first challenge for Xiao Ma Zhixing, who has chosen the high-tech technology route from the beginning, is how to find a suitable business model to continue to support further R&D testing of high-tech technology. The company survived and operated normally. In this regard, Xiao Ma Zhixing’s strategy is to 'walk on two legs'. Peng Jun has publicly stated that on the one hand, he is building the world’s most advanced technology in unmanned vehicles to ensure that it can be made truly safe. On the other hand, in order to survive, you must find the landing application in the current state. Xiaoma Zhixing will try to be the commuter in the park or the last mile solution provider in the whole area. For example, unmanned car express, security, mapping of surrounding maps, etc.. In the early commercialization of unmanned vehicles, more should consider the vertical, scened path. It is understood that Xiaoma Zhixing plans to new A round of financing is used to continue to expand the size of the Sino-US team, support and develop partnerships; more importantly, small and medium-sized unmanned vehicles and small-scale commercialization parties will be deployed. According to Peng Jun’s judgment, the most important challenge for the next step of driverless driving is whether it can be mass-produced in small quantities, which has great challenges to the stability and reliability of the system.