Source: Wall Street News

Since Musk visited China at the end of April this year, relevant$Tesla (TSLA.US)$The news of FSD's entry into the Chinese market went unabated. High-grade intelligent driving technology, which was once a marketing highlight, is gradually evolving into the standard for new energy vehicles, becoming the next growth factor expected by the market.

So are car companies in the domestic market ready for the era of advanced intelligent driving?

1. High-level intelligent driving begins to accelerate

The current characteristic of the domestic smart car industry is that the penetration rate of high-level intelligent driving is accelerating, and at the same time, it is beginning to sink to mid-range and low-end models.

(1) Accelerating penetration rate of high-level intelligent driving

The 15% penetration rate of emerging products is an important watershed. It represents that emerging markets are gradually being accepted by the mainstream population, and subsequent development will begin to accelerate.

Take the penetration rate of smartphones and new energy vehicles as an example. The smartphone penetration rate in the US also developed rapidly after reaching 17% in 2009, and directly rose to 27% and 42% in 2010 and 2011. New energy vehicles also showed the same trend, and the penetration rate began to accelerate after breaking through 15%.

Advanced intelligent driving ushered in a singular penetration rate at the beginning of this year. The loading rate of L2+ advanced assisted driving functions in new energy passenger vehicles has reached 13.3%, and high-level intelligent driving will enter an accelerated stage.

(2) Low-end models are also expected to be equipped with high-level intelligent driving systems

The reduction in hardware costs has allowed high-level intelligent driving to penetrate lower-cost models. Take lidar as an example. The price of lidar in the early days was mostly 2000-3000 US dollars, but now the mainstream domestic price has dropped to 500-1,000 US dollars.

In the first two months of this year, the average delivery price for high-grade intelligent driving models (based on standard NOA, automatic navigation and assisted driving) dropped from 359,000 yuan last year to 337,900 yuan.

Currently, the market anticipates that Xiaopeng may be equipped with a high level intelligent driving system in its low-end model Mona. At that time, advanced assisted driving will be further reduced.

2. In terms of NOA progress, traditional car companies are one step behind the new forces of car building

Domestic intelligent driving is mainly divided into three stages:

(1) The basic L2 level enables intelligent driving in closed or semi-enclosed scenarios, including only auxiliary functions such as automatic emergency braking and adaptive cruising.

(2) Expressway NOA (Autonomous Navigation Assisted Driving) can achieve intelligent driving in scenarios such as high-speed overpasses and ring roads. Functions include autonomous vertical ramps and active lane changes.

(3) In the urban NOA (Autonomous Navigation Assisted Driving) stage, intelligent driving can be realized in the city, and can handle the interaction of cars to vehicles, cars to people, vehicles and other traffic participants in everyday traffic scenarios.

At present, the first two stages have been successfully passed. In the third stage, the number of cities developed by NOA and the performance of smart driving algorithms directly determine the smart driving strength of current car companies.

Looking at it now, the new car builders clearly have a first-mover advantage, while traditional car companies continue to lag behind.

(1) New car builders are a big step ahead

In terms of the city's NOA progress, the new car builders are generally divided into four echelons, and the gap between them is not small. It will take half a year to 1 year to catch up with the average of the previous tier:

1) The first tier has achieved nationwide NOA promotion

Currently, the Huawei department (including Quanjie and Avita) and$XPeng (XPEV.US)$/$XPENG-W (09868.HK)$It forms the first tier of the electric vehicle industry.

As of May of this year, the two companies' NOA (Navigation Assistance) systems have covered all cities across the country, supporting urban main roads, branch roads, and national, county, and rural roads.

According to the target number of NOA cities for the full year 2024 set by major NEV companies, in the short term, no other domestic NEV company can shake the leading position of these two companies in the industry. Take Xiaopeng as an example. Xiaopeng's intelligent driving system is at the top of the new car builders in terms of hardware and software.

In terms of hardware, Xiaopeng Intelligent System has up to 32 sensors, making it the car company with the largest number of sensors among the new car builders, forming a comprehensive combination of camera+millimeter-wave radar+ultrasonic radar+lidar+high-precision positioning;

In terms of software, the algorithm iteration process for intelligent driving is from 2D direct view+CNN - BEV+Transformer - end to end.

Currently, most new car builders still use the BEV (Bird's Eye View) +Transformer algorithm. Through a bird's-eye view and the new neural network architecture Transformer, it is possible to convert between different 2D and 3D sequences;

Meanwhile, Xiaopeng has already begun introducing AI smart driving technology, doubling intelligent driving capabilities through large end-to-end models. Xiaopeng's three-model neural network xNet+ large model xPlanner+ big language model xBrain not only enhances the scene understanding and logical reasoning ability of autonomous driving systems, but can also conduct data training through continuous learning of human driving strategies, thereby improving the comfort and safety of intelligent driving.

2) The second tier has already started NOA in some cities, and is expected to be implemented nationwide by the end of the year

The second tier includes$Li Auto (LI.US)$/$LI AUTO-W (02015.HK)$,$NIO Inc (NIO.US)$/$NIO-SW (09866.HK)$Car companies such as Zhiji, and Jiyue have made good progress in urban development even though they have failed to achieve nationwide NOA promotion. As of May of this year, the number of NOA cities they have developed is 110, 726, 4, and 5, respectively.

The four NEV companies' NOA goals for this year are all to be able to keep up with the first tier by the end of this year and achieve NOA nationwide promotion.

3) The third tier has not yet achieved the city's NOA, but it can be achieved before the end of this year

The third tier includes Xiaomi cars,$ZEEKR (ZK.US)$und$LEAPMOTOR (09863.HK)$Until now, car companies have not been able to actually promote urban NOA, but they still pay great attention to the field of intelligent driving, and it is expected that they will reach an official launch in the short term.

Among them,$XIAOMI-W (01810.HK)$Automobile is expected to take the lead in implementing NOA systems in 10 cities by the end of May this year, and plans to increase the number 10 times to 100 cities by the end of the year; Zero Sports and Extreme Krypton plan to officially launch urban NOA in the second half of 2024, and are expected to land in the first batch of cities by the end of the year.

As for new car builders, such as Nacha, which are at the end of the hierarchy, there is clearly not much investment in the field of intelligent driving, and it may be difficult to make progress in the field of urban NOA this year.

(2) Traditional car companies are falling behind again

Currently, none of the traditional car companies have completed NOA landing in the city. BYD Tension and$GWMOTOR (02333.HK)$/$Great Wall Motor (601633.SH)$It is also only in the NOA testing phase in some cities, and has barely reached the level of the third tier of new car builders.

The main reason why traditional car companies are lagging behind is that they did not pay attention to intelligent driving in the early days.$BYD Company Limited (002594.SZ)$/$BYD COMPANY (01211.HK)$Chairman Wang Chuanfu still had reservations about autonomous driving in early 2023.

However, BYD has quickly turned around, and its smart driving engineers already have more than 4,000 people. The company said it will also invest 100 billion yuan in smart driving research and development in the future, and plans to deploy the self-developed advanced intelligent driving assistance system “Tenjin Eye” in all models of more than 200,000 models.

The second half of electric vehicles has already switched to intelligence. Xiaopeng and Huawei are already ahead, but the level of competition in smart driving will not be lower than when it was electrified; however, this time it will still bring about a major reshuffle.

Editor/jayden