The story of the frontier is always magical, and trillion-scale autonomous driving is inevitable.
Recently, Intel's autonomous driving subsidiary Mobileye officially landed on NASDAQ. The offering price was $21 per share, raising $861 million, up 27% at the open, closing with a market value of $23 billion (about 165 billion yuan), making it the fourth largest IPO in the U.S. this year.
On the same day, Argo AI, an American self-driving startup with a total investment of $2.6 billion by Ford and Volkswagen and once valued at $7 billion, collapsed.
In the past year, a large number of overseas companies focusing on self-driving car related technologies, especially L4 and above, have experienced varying degrees of decline in valuation or market value. Although the enthusiasm of the autonomous driving market at the global level is declining, in sharp contrast to overseas, the domestic boom for autonomous driving has not weakened, but more and more players are gearing up to try to make a big deal in this field.
The players of the domestic autonomous driving track can be roughly divided into three categories, namely Internet manufacturers who are accustomed to the frontier of the beach, such as Baidu and Huawei; and the "unicorns" who have grown up with the help of capital, such as Pony.ai and Wen Yuanzhixing; There are also new car companies that are trying to achieve corner overtaking with autonomous driving, such as Wei Xiaoli.
Standing on the same runway does not mean that you are an opponent. According to the information that has been publicly available so far, players have their own directions, either to provide full-stack solutions, or to target driverless taxis or commercial vehicles themselves.
No matter what direction, the immaturity of L3 level technology and the distant landing of L4 level are two mountains that players must face. In the process of climbing the mountain, there will always be people who will fall because they do not have the ability to make their own hematopoiesis, and there will always be people who will achieve a breakthrough from technology to application, from application to commercialization and maturity under the continuous influx and accumulation of a large amount of capital.
Internet giants rushed to enter the game
Autonomous driving will revolutionize the entire automotive industry, and the commercial value contained in it naturally cannot escape the eyes of Internet manufacturers.
Domestically, Baidu is the earliest Internet giant to enter the field of autonomous driving. If you count from the establishment of the autonomous driving R&D team in 2013, Baidu has been "deeply cultivated" in the field of unmanned driving for ten years.
In terms of commercialization, Baidu's unmanned driving business is divided into three ways. The first is to provide Apollo autonomous driving technology solutions for host manufacturers; The second is to integrate Baidu's innovation in autonomous driving through its Jidu Auto; The third is the self-driving taxi radish run.
Under the ecological layout for many years, Baidu has achieved certain results, especially the commercialization of the third road "Radish Run". At the end of 2021, Radish Run finally began trial charging operation, and since the beginning of this year, Baidu has successively opened pure unmanned autonomous driving services to citizens in Beijing, Wuhan and Chongqing.
Compared with unmanned taxis, providing technical solutions and vehicle manufacturing, the development of Baidu's two roads is not very ideal.
Baidu has cooperated with five independent brands including BAIC, Great Wall, JAC, Changan and Chery to provide L2 to L4 autonomous driving systems. However, cooperation with independent brands has not been smooth, and no car company has been able to help Baidu achieve its original planned goals.
In addition, on June 8 this year, after 15 months, Baidu's smart car brand Jidu finally released its first concept car, ROBO-01, but it did not achieve the effect of dazzling consumers.
Mass production is the ultimate proposition that autonomous driving cannot bypass, but under the influence of many external factors such as the epidemic, rising raw material prices and supply chain shortages, it is unknown whether Jidu can mass-produce and deliver on schedule.
Among the giant companies, Huawei is regarded as a major competitor of Baidu's driverless in the domestic market.
Huawei has always declared that it does not build cars, but only talks about cooperation. The first, like Baidu, provides a full-stack solution for intelligent driving including intelligent driving application software, computing platforms and sensors. The second is to provide the MDC intelligent driving computing platform, which connects the sensors and actuators of partners to support partners in developing intelligent driving software from various aspects.
At present, Huawei's intelligent automotive components business R&D team has more than 5,000 people, including more than 2,000 autonomous driving R&D personnel. It is understood that Huawei, BAIC, Changan and GAC chose the Huawei Inside cooperation model. In addition, Huawei is also helping Xiaokang Celis to build and sell cars, and from the perspective of the sales of "Ask the World", this cooperation is relatively successful.
The lively track is naturally indispensable for Ali and Tencent.
Alibaba established DAMO Academy and set up an autonomous driving department in October 2017, choosing the L4 technical route, mainly in the vertical sub-categories of logistics vehicles and passenger cars, which is highly compatible with its own business. On the one hand, Tencent has invested in autonomous driving companies, and on the other hand, it has continuously optimized its autonomous driving software, and has become a "connector" at the same time, without launching related hardware products.
In fact, in addition to the above players, Xiaomi, Didi, DJI, OPPO and other Internet manufacturers have also played automatic driving in the past two or three years, and each has different development trajectories and product and technical characteristics.
Unicorns are flocking
The wide track of autonomous driving, the large market, and the continuous rise of capital enthusiasm have also achieved many entrepreneurial enterprises.
At present, there are many "unicorns" on the entire track, such as Pony.ai, Wen Yuan Zhixing, Milli Chi Heng, momenta, Autox, Yuanrong Qixing, Qingzhou Smart Aviation, Horizon, Black Sesame, Chaoxing Future, CIIC, Mushroom Car Union and other companies.
Among these "unicorns", Pony.ai, co-founded by Peng Jun, former chief architect of Baidu's North American R&D, and Lou Tiancheng, the head of Chinese programming, has attracted the most attention from the market.
Like most autonomous driving companies, Pony.ai also provides full-stack autonomous driving technology in terms of business model, and has established cooperation with Toyota, Hyundai, FAW, GAC, Bosch and so on. In addition, I also chose Robotaxi, the most difficult but also the most imaginative business.
At present, Pony.ai's driverless taxis have entered the preliminary stage of commercialization, and related technologies have also been connected to Cao Cao's car, Ruqi and other travel platforms.
The difference is that in the choice of autonomous driving scenarios, Pony.ai aims at heavy truck logistics. Starting from the car use scenario, the truck route is basically a highway, and there are no complicated road conditions such as mixed driving, which also means that the technical research and development of self-driving trucks will be lower than that of cars, and commercialization will naturally land faster.
There is always a big difference between theory and reality, and although Pony.ai has certain technical advantages and no problems with the layout, it has not embarked on a path of commercialization that can be continuously profitable. Valuations continue to climb, losses are getting bigger, and it will rely more on external blood transfusions.
And when the blood transfusion is suddenly interrupted, you can only choose to survive with a broken arm, or package yourself for sale.
Compared with Pony.ai, with a valuation of US$4.4 billion, WeRide has more diverse business scenarios, including self-driving taxis and self-driving freight vehicles, and WeRide also has self-driving minibuses and self-driving sanitation vehicles.
Starting from subdivided scenarios, commercialization is achieved faster, revenue growth attracts capital for continuous blood transfusion, and then more complex scenarios. The players who choose this transition path are not only Wen Yuanzhixing, to be precise, it has become a helpless choice for many autonomous driving start-ups.
For example, Great Wall Holdings' autonomous driving company Milli Chi Heng is making efforts in the field of low-speed logistics distribution vehicles. "Low speed" means that even if there is a failure, the damage will not be too large. At the same time, "logistics and distribution" belongs to the business model, and the C-end is not the main user, but unmanned sales enterprises and logistics companies such as Jingdong, Meituan and Ali.
The problem is that this path is difficult to solve for long-tail data. Sensor solutions and computing platforms are getting closer and closer, and there are fewer and fewer suppliers. Therefore, the data determines the victory or defeat of the autonomous driving company. But compared to the "technology" toughness, in the "unicorns", it is more important to survive.
It is worth mentioning that Gu Weihao, CEO of Milli Chi Heng, was the general manager of Baidu's intelligent vehicle business unit, and Han Xu, the founder of WeRide, was also the chief scientist of Baidu's autonomous driving business unit.
What's more interesting is that according to statistics, between 2015 and 2018, more than 16 unmanned senior talents left Baidu to start their own unmanned field companies, and Baidu's "Unmanned Huangpu Military Academy" lived up to its name.
New car companies are accelerating self-research
In the eyes of the market, passenger car intelligence is the main way to commercialize autonomous driving. Autonomous driving has also been regarded by the outside world as the standard of cars in the era of intelligent vehicles, and is the key to the victory of intelligent electric vehicles.
Then for car companies, they must gradually take autonomous driving technology into their own hands and add important chips for future development.
At present, the self-owned brand part has achieved mass production of L2 autonomous vehicles, and traditional automakers include Changan, Great Wall and Geely. However, compared with the new forces of car making and autonomous driving startups, the performance of domestic traditional car companies in the field of autonomous driving is not outstanding.
Of course, for the new forces that do not dominate "car building", research and development capabilities are also an important support for their outstanding encirclement, and the ability to independently complete the development of the whole process of automatic driving perception and decision-making algorithms, vehicle control strategies, underlying systems and testing will greatly increase their competitiveness in the market.
From a technical point of view, Xpeng, NIO and Ideal are representatives of autonomous driving, while car companies such as Zhiji Auto and Avita have created a competitive advantage in autonomous driving through in-depth cooperation with leading suppliers.
Among them, Xpeng has always been a new force car company with a fairly active autonomous driving layout, and has been stubbornly smashing the pure self-developed route of pure vision + perception integration since its establishment.
In particular, Xpeng's urban NGP can complete intelligent navigation assisted driving on urban roads such as patrolling and following the car in the lane, navigating/overtaking lane changes, merging in/outcoming roads, recognizing the start and stop of traffic lights, passing at intersections/roundabouts/tunnels, and avoiding other traffic participants.
As for the ideal, it has switched from the earliest supplier solution, that is, from not doing L3 to the current deeply customized quasi-L3 supplier solution. NIO originally wanted to go straight to L4, but later switched back to the current fusion perception pure self-developed route.
In addition to Wei Xiaoli, new forces such as WM and Zero Run have also made arrangements in the field of autonomous driving. For example, it actively cooperates with solution providers such as Baidu and Huawei, and configures LiDAR, camera and other hardware in many places on new models, as well as road testing L3 and L4 autonomous driving systems.
As the most powerful weapon for new forces to seize the market, autonomous driving is both a halo and a shackle.
In the past two years, the "autonomous driving" accidents of new automakers have been frequent, and the fundamental reason is the misleading propaganda of car companies. The core difference between autonomous driving and automated assisted driving is whether the control is in the human or the machine. At this stage, the automatic driving technology promoted by some automobile companies should be called assisted driving technology to be precise, and cannot be called automatic driving technology.
After many public opinions, new car companies such as Wei Xiaoli have also learned to be obedient, and the previous automatic driving has become the current "intelligent assisted driving".
The frequent occurrence of accidents has undoubtedly cast a shadow over automatic driving. But this does not affect the significance of autonomous driving for new car companies, and Wei Xiaoli has not slowed down their pace, but is accelerating the layout.
However, in recent years, the investment in self-developed automatic driving has only increased, and how long can the new forces that are not rich themselves last?
In general, autonomous driving is a demand and an inevitable development in the future. And it can be believed that under the impetus of capital, policies and other aspects, players chase me, there will always be someone out, and someone will always achieve a breakthrough.
This article is from Zinc Finance