In an era of respecting traffic, the traditional concept of selling cars no longer seems to be feasible in the “spoiling” of technology companies such as Huawei, Xiaomi and Baidu.
In particular, Lei Jun held a Xiaomi car press conference on the evening of March 28, which directly gave a lesson on “Internet thinking” to the entire automobile industry.
As a result, it can be seen that the leaders of car companies, who used to be behind the scenes all the year round, have recently come out one after another in person, either turning on live broadcasts or appearing on major social platforms such as Weibo and Douyin to “scan” the brand’s face with an active posture to earn hot topics.
On April 15, Wei Jianjun, chairman of Weibo, launched an hour-long NOA test of Great Wall Motor without high-precision map in Baoding, Hebei Province, and passed the super-difficult and complex road conditions in Baoding.
Photo source: live screenshot of Great Wall Motor.
According to reports, Baoding represents the complex and changeable road conditions of China’s second-and third-tier cities, and the traffic environment is full of challenges, from complex “Liudaokou” to narrow streets, etc.
have put forward a severe test to every action of the intelligent driving system, which can fully test the true level of the intelligent driving system.
At the same time, Great Wall Motor is also actively promoting smart driving tests in first-tier cities, covering a wider range of test scenarios.
In this Smart Drive live broadcast, Great Wall Motor’s full scene NOA can accurately identify standard traffic lights and all kinds of non-standard traffic lights, ensuring safe and reliable traffic at various intersections.
The farthest identification distance of the system reaches 150 meters, whether it is a standard intersection or a special intersection, the turning success rate is more than 95%.
Source: live screenshot of Great Wall Motor.
Secondly, in addition to accurately identifying static obstacles, Great Wall Motor NOA can also accurately identify dynamic obstacles, such as pedestrians and non-motor vehicles.
The system can flexibly adjust the driving route and improve the driving safety according to the identification of obstacles and real-time traffic conditions.
In addition, in the whole scene lane change and lane management, it can provide comprehensive lane change support, whether it is the confluence and outflow of the highway or the congested sections of urban traffic, it can achieve efficient and safe lane change operation, and the success rate of lane change is 98%.
And can effectively reduce the risk of congestion on congested roads.
Even in the section where there is no lane, the system can keep the vehicle running steadily by virtue of its strong lane simulation ability.
However, Wei Jianjun has also repeatedly stressed that the premise of smart driving must be based on security, and we cannot talk about smart driving without security.
“car safety is no small matter, safety is always the first priority.
We should have a more serious attitude, be more fully prepared, dare to challenge complex scenes, not treat users as guinea pigs, and make users feel more at ease.
” Photo source: live screenshot of Great Wall Automobile, Great Wall Auto Intelligence driving force, the goal to enter the first echelon, for a long time, the relevant technical layout of Great Wall cars in the field of intelligent driving appears to be quite “low-key”.
In the live broadcast process, Wei Jianjun said bluntly that Great Wall Automobile attaches great importance to intelligence and started the research on self-driving as early as a decade ago.
People think that we are a traditional car company, but in fact we are an industry leader in smart driving, cockpits, large model applications, and so on.
In addition to applications in cars, in the field of vertical AI, such as automotive modeling, we used AI technology to assist styling design three years ago.
” He went on to say.
According to public data, in July 2020, on the occasion of the 30th anniversary of the founding of Great Wall Motor, it formally put forward the goal of “transforming into a global intelligent technology company.
” in 2023, Great Wall Automobile entered the first year of intelligent transformation.
Subsequently, Great Wall Motor has an intelligent cutting-edge organization AI Lab under the technology center, which will be responsible for building the full-link AI technology system of Great Wall Automobile and building AI capability, which will also build a technical communication bridge for the intelligent brand “Coffee Intelligence” of Great Wall Automobile from the R & D end to the whole vehicle end.
At present, Great Wall coffee intelligence has evolved and iterated to the 2.0 era. In this full-scene NOA live test, Great Wall Motor’s latest generation of intelligent driving system Coffee Pilot Ultra (referred to as CP Ultra) was also officially unveiled.
Source: live screenshot of Great Wall Motor.
It is reported that the system is based on Orin-X ‘s high computing power domain control platform, with 1 lidar, 3 millimeter wave radar, 11 HD cameras and 12 ultrasonic radars, and has all-weather sensing capability.
Based on the characteristics of strong perception, safety and fast iteration, it can realize the panoramic connection from driving to parking, from highway to urban and rural areas, and can “drive anywhere and park anywhere”.
Provide users with a full-scene, point-to-point intelligent travel experience.
At the same time, CP Ultra has achieved 100% high-precision maps, and this is the first time that the full-scene NOA capability without high-precision maps has been launched on the products of Great Wall Motor.
In addition, CP Ultra has the functions of cross-layer memory parking, automatic parking and tracking reversing, and supports 200 + parking spaces.
The success rate of parking identification is 98%, the success rate of parking is 98%, and the maximum driving distance is 3 km.
According to the plan of Great Wall, CP Ultra will first be installed on the flagship model of high-end brand Wei Brand, and there should be different versions of Coffee Pilot, such as pro and base, for different brands and models.
Wei Jianjun also set up flag in the live broadcast to enter the first echelon position in the field of intelligent driving.
It is worth noting that Wu Huixiao, vice president of intelligent driving at Great Wall, also put forward his own views on the controversy over whether the intelligent driving in the industry should follow the route of pure visual technology or perceptual fusion based on lidar.
She believes that pure vision does not mean that it is more advanced, and that Great Wall will stick to the lidar fusion scheme for high-end models in order to be safe.
The more intelligent driving goes backward, the more it needs the support of computing power, high-quality data, and so on.
Great Wall believes that high-quality data will directly determine the ability of smart driving in the future.
As a matter of fact, the “brain” of the NOA of Great Wall Motor comes from the SEE integrated intelligent driving model of Great Wall Motors, which is the first integrated intelligent driving model in China, according to officials.
The model not only realizes the link integration of perception and decision integration, but also realizes the transformation from rule-driven to data-driven and from machine to brain.
At the same time, based on the unlimited growth of SEE, as long as the continuous feeding of data, it can continue to evolve, laying a technical foundation for L2 to L4 evolution.
Of course, data training inevitably requires a huge amount of arithmetic support.
To this end, Great Wall Motor spent a lot of money to build a line.
Jiuzhou, the leading supercomputing center in the industry.
Its ultimate scale reaches 1.
64EFLOPS, achieving high-performance storage of 5T/s, and a high-performance network with a communication bandwidth of 3.
7TB/s, providing a solid computing power base for model training.
Photo source: Great Wall Motors, Wei Jianjun personally went to the live broadcast this time to show off his smart driving standards to the outside world during the full-scene NOA challenge on difficult road conditions.
In the second half of the intelligence, as more and more top leaders of car companies move from behind the scenes to the stage and be active in the public’s vision, it is not difficult to imagine that whether it is brand release, technology research and development, or exit methods, the battle situation in the smart electric vehicle market will become more exciting, and the leaders of car companies will have to get involved first.
Return to Home>.
