Yiche original Nezha car as a rising star in the smart electric vehicle although the start is late but quietly has its own place, according to the data of the association, in January 2022, the new force pure electric SUV sales list:
The Nezha V car series delivered 6482 units, ranking first in the list for four consecutive months, becoming the top of the new force pure electric SUV sales in January 2022, ranking the top spot in the new force pure electric SUV sales, so that the market once again saw the "Nezha speed".
Nezha U Pro delivered 4527 units, just 10 months on the market, nearly 5000 units delivered in a single month, ranking in the top three sales of pure electric SUVs of the new forces for two consecutive months. Since its delivery, Nezha Automobile has delivered 106986 units, with production and sales exceeding 100,000 units, ranking first in the new camp of car-making forces.
What magic weapon or ability does Nezha have behind this series of good results that can make him gain a firm foothold in the fiercely competitive new forces? On the last day of February, Nezha Automobile held a unique Shanhai platform intelligent driving communication meeting at its design center in Beijing, and the author had the honor to participate in it and conduct in-depth communication with the intelligent driving team of Nezha Automobile.
01 From the Shanhai platform, THE TA Pilot system is blessed
First of all, "Shanhai Platform" is the industry's first full-stack self-developed intelligent safety vehicle platform that integrates high safety, high scalability, high intelligence and environmental friendliness released by Nezha Automobile. Nezha S, which will be listed at the end of 2022, is built from this platform.
Secondly speaking of Nezha intelligent driving system, this time they announced the full-stack self-developed intelligent driving system TA Pilot, TA taken from Nezha Car's English name NETA suffix; TA is also taken from the TALENT prefix, TALENTED Pilot literal translation is the talented navigator, representing that Nezha has enough confidence and expectations for his own intelligent driving.
Referring to this TA Pilot Nezha said that it will launch different levels of configuration from 2.0-5.0 to meet the needs of different stages of products, and will cover the entire series of models. Ta Pilot 2.0 and 3.0 have been deployed in the mass-produced U and V series and subsequent models.
TA Pilot 4.0 will be mass-produced on the Nezha S listed at the end of this year, realizing intelligent driving technology and user experience "generation leveling", which is also the highlight of this communication meeting.
In addition, ta Pilot 5.0, the next generation of full-stack self-developed 1000+TOPS, is under development.
So what are the highlights and unique features of the TA Pilot 4.0 system?
First of all, this system is equipped with Huawei's MDC610 chip, the computing power reaches 200TOPS, the sensor is equipped with 2 solid-state lidars; 11 auxiliary driving cameras, including 2 forward 8M cameras, 4 surround view cameras, 5 circumferential vision cameras; as well as 5 millimeter wave radar and 12 ultrasonic sensors, through follow-up communication to understand that the two solid-state lidars equipped on the Nezha S are also from Huawei, it is not difficult to guess that it is Huawei's equivalent 96-line solid-state lidar. Through live vehicle demonstrations, the two solid-state lidars will be placed under the headlights.
It is worth mentioning that although the hardware equipment is provided by Huawei, the software level is self-developed by Nezha full-stack, and the software algorithm and data are in their own hands for future corrections, upgrades and model training, which have a very important role, for an OEM, the core algorithm can be mastered in its own hands to grasp more discourse power and do not lose the "soul". This time, Nezha became the industry's first intelligent driving mass production solution equipped with Huawei's MDC platform + OEM software self-development mode, which has also become a benchmark for industry cooperation.
Next, the product functions of TA Pilot 4.0 include NNP-Nezha Intelligent Pilot Assist, NCP-Nezha City Pilot and NTP-Nezha Memory Parking;
--NNP can realize pilot assistance driving of high-speed and urban expressways within the coverage of high-precision maps;
--NCP can realize the driving assistance function of structured roads in urban areas covered by high-precision maps;
--NTP is for unstructured roads that do not cover the last 5km of HD maps.
Engineers also replied to whether these functions can be enjoyed on the market, and the NNP can be delivered, while the higher-level NCP will continue to upgrade the OTA according to the coverage of the local high-precision map.
02 Why Nezha chose full stack self-research
We know that the traditional automobile ecology can not provide a good intelligent ability and driving experience, intelligent track everyone is in the same starting line, Nezha such a new force company can be said to catch up with a "best era".
Full-stack self-development can quickly respond to users' demand for intelligent driving: The current trend of consumer group rejuvenation is becoming more and more obvious, and now young people can be said to be the indigenous people of the "intelligent" era, the requirements for intelligent experience are getting higher and higher, and it is more and more important to respond quickly to user needs and enhance intelligent driving experience. Intelligent driving is a long-term track, in order to adapt to the characteristics of long-term and rapid iteration, the full-stack self-developed way can ensure rapid iteration of products and form a long-term system capability accumulation, and finally provide customers with a leading and safe driving experience.
Full-stack self-development can improve the coherence of the driving experience and realize the path: 'Scene +'
From simple driving scenarios to complex scenes, the system often requires drivers to frequently take over or intervene, or even quit, which will bring users great anxiety, and lack of trust in intelligent driving over time. Around this point, intelligent driving requires full-scene adaptation, ensuring the consistency of the experience and thus gaining the trust of consumers.
1. Weaken the driving level and strengthen the 'Scene +' layout
Nezha Automobile's intelligent driving technology is gradually weakening the concept of driving level, and gradually enhancing the expression of the scene, realizing the coverage of the whole scenic spot to the point, extending from high speed to the main road in the urban area, parking lot, residential area, and even covering all semi-closed unstructured roads, enhancing the coherence of the cross-scene functional experience.
In the industry, the difficulty of full-scene coverage lies in solving the driving ability of the area without HIGH-precision map coverage, city ends, semi-closed and closed areas, because of factors such as high cost of map data collection and information limitations, it is impossible to achieve full coverage of high-definition maps. In order to solve this problem, Nezha Automobile took the lead in proposing the 'last 5 kilometers' scenario, and developed a solution that relies on perception ability, which is a forward-looking thinking on the difficulties of the industry and customer needs.
2. Lidar support to improve the perception ability of complex scenes
With the blessing of lidar and front-looking 8M camera, obstacle type recognition is more abundant and environmental portrayal is more accurate. In order to adapt to the complex urban traffic environment, Nezha intelligent driving system adopts the point-wise & bird-view and feature level fusion perception scheme of lidar; the last 5 km scene (no HD map coverage area) generates a local map in real time by fusing the navigation map and on-board sensors; and the parking scene uses lidar to improve safety and garage cross-layer capability Lidar has a great effect on improving the adaptability and system robustness of the above scenarios.
3. Data closed loop
The advanced nature of intelligent driving systems is largely reflected in the effectiveness of handling various corner cases. In order to improve the adaptability to long-tail scenarios, the algorithm must be continuously improved through data. Therefore, the realization of data closed loop is the current core power point of Nezha Automobile, and it is expected to achieve algorithm iteration and product upgrade through data closed loop to promote experience upgrade. Specifically, it is embodied in three points, self-developed data-driven algorithms, collect massive high-quality training data, and develop an efficient automated data processing platform. Data-driven algorithms can continuously learn data generalization scenario recognition capabilities; in terms of data sources, as more and more Nezha cars are equipped with intelligent driving systems, data will increase in geometric multiples, and it is expected that the cumulative amount of data will reach 10+ PB in 2022, and the large-scale simulation platform will reach million kilometers per day; in order to make more efficient use of data, Nezha Automobile has developed an automated processing platform, relying on large-scale heterogeneous data lakes, semi-automatic labeling of large models, ultra-high computing power GPU clusters, Large-scale simulation platform and data mining algorithm, continuously mining effective scene data in massive data, and iteratively verifying algorithms.
03 Increase investment to build a super intelligent driving team
At this event, Nezha Automobile also announced to us the investment plan in talent introduction and technology. At present, the number of smart hospitals is 600+, and the goal is to reach 1000+ within 2022;
Intelligent driving team information: the current 300+, the proportion of master's and doctoral degrees is more than 80%, the proportion of algorithm software personnel is 70%, and it will reach 500+ in 2022;
Total investment in intelligent R&D: 1000+ people, 10 billion investment;
Intelligent driving R & D investment: it is expected to continue to invest 8 to 1 billion yuan per year;
Intelligent R&D area: Shanghai and Beijing double R&D center; Intelligent driving R&D region: Shanghai is the R&D headquarters, the Beijing R&D center is based on algorithm development and data development, and both centers are currently expanding.
04 Write at the end
This activity allows us to have a further understanding of the intelligent process of Nezha Automobile, in today's foreign countries have Tesla as the only one, and there is fierce competition from "Wei Xiaoli" in China. As a rising star Nezha Automobile is constantly low-key to come up with a high-quality answer sheet, but also let us have more choices in the field of intelligent electric vehicles, there are more high-quality domestic car companies appear, the more our domestic new forces can gradually grow, for consumers, who can provide cost-effective, high intelligence and high endurance of electric vehicles who are the most worthy of our respect and support of car companies. We will see how the future pattern is.