Founded only two and a half years ago, Millima Wisdom Is Showing Amazing "Millima Speed".
Following the launch of the high-speed pilot assistance NOH in November last year, The 2022 HAOMO AI DAY held on April 19 announced that the NoH of the City Pilot Assisted Driving will also be mass-produced in the near future.
In about half a year, the assisted driving system extends from the high-speed domain to the urban domain, which can be said to be the fastest landing speed, and it may also become the first large-scale mass production enterprise in China.
In terms of technology development and mass production landing, why is the growth of Zhixing so fast?
"China's first mass production of autonomous driving"
Relying on the great wall of large customers, Zhixing won the title of "the first place in China's mass production of autonomous driving" in 2021.
Since the launch of the first model equipped with the HWA high-speed intelligent driving system in May 2021, the Weipai Mocha has been mass-produced in five models, including the Tank 300, latte, Macchiato, and Haval Divine Beast, and in November 2021, the high-speed pilot intelligent driving system NOH was successfully launched, which was installed on the Weipai Mocha.
As of April this year, the total mileage of users equipped with the model of the Zhixing Intelligent Driving System has exceeded 7 million kilometers. At the same time, in this month, the tank 500 equipped with the second-generation intelligent driving system of Millide also successfully achieved mass production delivery, and the Tank 500 became the sixth model equipped with the intelligent driving system of Millipede Zhixing.
Zhang Kai, chairman of Miller Zhixing, believes that the difficulty faced by "getting on the car" is much greater than that of the new power brand, because the new forces are basically a single model, and the millima zhixing needs to face different cross-platform models, including fuel vehicles and pure electric models, as well as small cars, small cars and large cars and other different types.
In the field of terminal logistics automatic delivery vehicles, the end of the zhixing also ushered in a new trend, the second generation of terminal logistics automatic delivery vehicle Xiao Mo Camel 2.0 was officially released, the bicycle price of 128,800 yuan, becoming China's first 100,000 yuan terminal logistics automatic delivery vehicle, is expected to be put on the market in May 2022.
Little Devil Camel 2.0 has been fully upgraded in terms of configuration, equipped with vehicle-level perception suite, ICU3.0 large computing platform; supports the rapid power exchange function, which can achieve a real mileage of 100km in a single power change; in terms of intelligent driving, it covers the full road conditions of low speed in urban open roads, and optimizes the performance of complex traffic scenarios such as mixed traffic and congestion.
In addition, Zhixing has also comprehensively upgraded the production base of terminal logistics automatic delivery vehicles, and the upgraded production workshop covers an area of 10,000 square meters, which can achieve the production capacity target of 10,000 unmanned delivery vehicles per year. The new production line meets the customized needs of customers, and can realize small batch trial production and large-scale mass production.
The city NOH debuted, and at the end of the day, it will be open to cooperation
"2022 will be the most critical year for the development of the autonomous driving industry, and the competition in the autonomous driving industry will officially enter the second half. The competition in the second half is mainly concentrated in the field of pilot intelligent driving in the urban open scene. ”
Zhang Kai judged in his speech that with the continuous improvement of the penetration rate of high-level intelligent driving of passenger cars, the NOH intelligent pilot assistance driving system has gradually entered the city, and the intelligent driving of urban scenes is becoming a new place of contention.
At the AI DAY, The "Miller City NOH" equipped with HPilot3.0 was officially released, and the vehicle equipped with the Millipede City NOH system will also be officially mass-produced in the near future, when the Millipede Zhixing may become the first company to truly land urban auxiliary driving products.
At present, the NOH system of Miller City has been polished in depth in Beijing, Baoding and other cities, and in the actual experience of Cheyun bacteria a few days ago, The NoH of Millime City has achieved zero takeover of 22 traffic lights in the whole process. In the future, the number of cities where this function will be landed will exceed 100, covering all first- and second-tier cities in China.
In terms of hardware configuration, the system is equipped with an ultra-high computing chip with AI capability of 360T, cache 144M, and CPU computing power of 200K+ DIMPS; it is also equipped with 2 lidars, 12 Cameras, and 5 millimeter-wave radars.
In terms of function realization, according to the driving route provided by the navigation, the noh of the city can realize the main functions of automatic lane change overtaking, traffic light recognition and control, complex intersection traffic, unprotected left and right turning in the urban environment, and can also cope with complex urban traffic scenarios such as vehicle close entry, vehicle obstruction, intersection, roundabout, tunnel, overpass.
It is worth noting that Zhixing has also seen players in the field of passenger car autonomous driving, more for the "co-creation cooperation" rather than the "supply" model. Zhang Kai believes that with a more open and flexible business model, technology companies and car companies will jointly develop software and hardware for autonomous driving in a co-creation and cooperation manner, which will become an inevitable trend in the development of the industry.
To this end, in the field of passenger car autonomous driving, Miller Zhixing has designed a "6P open cooperation model", and partners can choose to adopt the full-stack technology solution of Milliper, can choose to cooperate at the level of data intelligence cloud services, or can choose the cooperation of software or hardware, or the functional module level; and can even choose to customize with Milliper at the source code level.
In short, the "6P open cooperation model" means that the possibility of cooperation can be explored openly at the 6 product levels between the full-stack solution and the source code.
Hou Jun, COO of Zhixing, said in an interview that some domestic and international OEMs were surprised by the feedback of this open cooperation model, "How can it be so open? ”。 He revealed that Mo Mo is talking about cooperation with OEMs outside the Great Wall, and at the same time he also said: "Mo Mo will not serve all the world's OEMs, because the service cannot come." ”
Core source power, MANA evolution upgrade
As we all know, data is the biggest driver of artificial intelligence. The cornerstone and core source power of the evolution of all products is MANA, an intelligent system for autonomous driving data released in December 2021.
MANA is a complete set of data processing tools based on massive amounts of data, including algorithm models, test verification systems, simulation tools, and computing hardware. After more than 3 months, MANA has ushered in a number of upgrades in terms of reducing costs and improving iteration speed.
"The perception technology of autonomous driving has undergone tremendous changes in the recent year, including the geometric growth of chip computing power, the emergence of transformer transmodal models and the rapid increase in Camera pixels." Gu Weihao, CEO of Miller Wisdom Walk, said that the change of basic technology drives the change of solutions, and Miller began to think about new solutions from the first principle, the first of which is to solve the most challenging "traffic light recognition" problem.
Therefore, in terms of perception intelligence, Miller has created a "double-stream" perception model, which decomposes the traffic light detection and the road binding problem into two channels, so that the daily passenger car test of Millimae realizes the traffic light recognition under the light map. In addition, the self-developed BEV Transfomer is used for lane line recognition, which realizes the automation of labeling through the two stages of "target coarse positioning" and "fine estimation of attributes", and realizes multi-sensor fusion lane line recognition on urban roads, so that there are no lane lines in China that cannot be recognized.
In terms of "cognitive intelligence", MANA replaces the traditional handwritten rules and parameters with machine learning models, which solves the problem of bloated code and easy to collapse and fail in the face of complex scenarios, making scene decisions more generalizable and applicable, and greatly improving interpretability and generalization capabilities.
In terms of "cost and speed", Mo Mo and Alibaba have carried out in-depth cooperation in big model data processing technology, realizing the first encounter with the M6 model in the field of autonomous driving.
At present, MANA has completed nearly 200,000 hours of learning time, and the virtual driving age is equivalent to 20,000 years of human drivers. In the AI autonomous driving industry, there is a saying: "Whoever can mine the value of data efficiently and cheaply can become the king of competition." "And MANA is the core of the effective acquisition of high-quality data and the success of the competition in the second half of autonomous driving."
Che Yun Summary:
Not long ago, Zhixing has just announced the completion of the A+ round of financing of hundreds of millions of yuan, so that it has sufficient "ammunition" in research and development investment and talent system construction. Next, the goal is to win the three major battles of 'data intelligence technology war', 'intelligent driving city scene war', and 'scale war of terminal logistics automatic delivery vehicles'.
Zhixing is providing a new development path and model for the autonomous driving industry with hands-on practice.