How Does Xiaopeng, A Latecomer With Outstanding Achievements, Break Through The Encirclement Differently?

Around 2015, driven by the "new four modernizations" of automobiles and the entrepreneurial trend driven by Tesla, a number of new forces of China's Internet car making were born. In the voice of doubt, many entrepreneurs have broken their dream of making cars and left the scene in a gloomy way. However, some leading enterprises, represented by Weilai, Xiaopeng and ideal, have gradually stepped out of their own development path.

In 2021, a new round of car making is coming. Compared with the previous round of car building boom starting from scratch, Baidu and Xiaomi are able to absorb experience, and the purpose of cross-border car making in the name of "company" is more integration with the original business.

It is worth noting that whether it is weixiaoli, which stands out from the first round of new forces of car making, or a new round of cross-border car making forces of ICT enterprises, they have a common point: electrification is the foundation, and will eventually tend to be homogeneous. The differentiated competition of intelligent vehicles is the new opportunity brought about by the reform of automobile industry.

What does cross-border vehicle manufacturing enterprises bring to the automobile industry? How to evaluate the mode differences among different forces? What is the core of future intelligent electric vehicles? How do car companies arrange automatic driving? How to solve the problem of neck sticking?

In response to the above problems, on June 7, in the large-scale video interview column automobile new business geography jointly produced by China electric vehicle 100 people's Association and 21st century economic report, Yang diange, director of development planning department and director of discipline Office of Tsinghua University, founding president and professor of School of vehicle and transportation of Tsinghua University, Xia Heng, co-founder and President of Xiaopeng Automobile Co., Ltd, Xia Yiping, CEO of Jidu automobile, launched a wonderful thought collision.

Xiaopeng automobile is one of the first batch of "outstanding students" of China's first new automobile manufacturing forces. Its sales volume is firmly in the first echelon of new automobile manufacturing forces, and has successfully landed on the New York Stock Exchange. He Xiaopeng, chairman and CEO of Xiaopeng automobile, has many years of Internet entrepreneurship background. The co founders of the company, Xia Heng and He Tao, have many years of working experience in traditional automobile enterprises and have a strong Internet + automobile gene.

Later, Jidu automobile was established at the beginning of this year, which is a joint venture between Baidu and Geely Automobile. The main purpose of Baidu's launch of Jidu automobile is to implement the AI technology accumulated in the past seven or eight years. By working with Geely to make up for the vehicle manufacturing experience, we can push the model to the market as soon as possible.

Although the birth time, background and concept are different, both Xiaopeng and Jidu regard intelligent and automatic driving as their core competitiveness.

"If the biggest innovation in the past 20 years has been in mobile phones, the biggest carrier of innovation in the next 30 years will be smart cars." Yang diange said.

In Yang diange's view, the main difference in the future of automobiles is not the difference brought by electrification, but the difference brought by intelligence. For all auto companies, data will become a very core capability.

Each path has its own advantages and disadvantages

"21st century economic report" (hereinafter referred to as "21st century"): how to look at this round of cross-border car building boom?

Yang diange: the cross-border vehicle manufacturing of ICT enterprises is driven by the market first. There is demand in the market, but the technology drive is more important behind it. If the biggest innovation in the past 20 years has appeared on mobile phones, the biggest innovation carrier in the next 30 years will be smart cars.

Smart cars will generate much larger data than mobile phones, and at the same time will change our way of travel. This is a once-in-a-century development opportunity. In such a wave of technology, I believe that these visionary and visionary enterprises will not give up the opportunity.

Of course, the entry of these enterprises into the smart car industry at this stage also shows that they think it is feasible to build smart cars on the basis of electric vehicles. Especially, Weilai, Xiaopeng and ideal have proved to you that when new energy vehicles and smart cars arrive, we will have the opportunity to follow these traditional Toyota, Mercedes Benz BMW and other big car companies are on the starting line, and we have a chance to win this victory. It is because of this phenomenon, to millet, baidu more confidence, bravely into the car track.

21st century: what impact will the entry of ICT enterprises such as Huawei and Xiaomi have on Xiaopeng?

Xia Heng: the electric industry chain is becoming more and more mature, coupled with the development of China's Internet, the accumulation of intelligent experience and potential energy in recent years, which brings opportunities to a large number of new car manufacturing enterprises. Compared with the automobile industry of a century, even many traditional enterprises in China still have a long way to go.

In the future, the competition of intelligent electric vehicles should first cultivate the differences of intelligence. It is difficult to establish the difference of electric power, but the difference of intelligence is the most fundamental content of this automobile revolution. In the automatic driving, intelligent cockpit and other aspects of the intelligent need to continue to cultivate, to find out the differences that can be recognized by consumers.

Second, it must be novel in design and high in appearance. Intellectualization is very important. At the same time, appearance must be able to resist fighting, and future design must be the focus of competition among enterprises.

Third, electric vehicles and smart cars are changing from vanguard products to more and more popular products. Quality and quality are very important. The quality and safety of products are always concerned by all consumers.

"21st century": some people say that the second half of the Internet is a combination of traditional enterprises. How can baidu cooperate with Geely to do a good job in concentration?

Xia Yiping: the cooperation between Baidu and Geely is OEM mode, but it is not at all. The cooperation between the two sides is very in-depth. In addition to the vast architecture is a very advanced and open platform, the cooperation with Geely has more to measure.

First, in the era of software defined automobile, is there a relatively open platform that can support Baidu to make a deeper change in the process of intelligent and unmanned landing, or drive the hardware changes on the whole platform.

Second, how to solve the problem of supplier's willingness and cost. Geely's supply chain capability can solve this problem very well. In addition, we revere product safety, Geely's quality is very good. Baidu hopes that quality and quality control can continue Geely's system, and solve the problems that may be faced in the beginning of car building with the help of their logistics and factory production capacity.

Of course, in essence, product research and development is still guided by user needs. From the product appearance design, the definition of user requirements to the definition of all the functions in the car are completed by concentration. But all the selection and supplier selection will be completed together with Geely's supply chain team.

Both sides adhere to the concept of win-win cooperation. Baidu takes advantage of Geely and Geely also shares Baidu's ability and experience in intelligence. For example, Baidu's supplier of high-level automatic driving needs can also be used by Geely in the future.

The two sides are very open cooperation mode. In this process, Baidu will give full play to Geely's existing maximum advantages to ensure the platform and supply chain requirements of intelligent and high-level automatic driving technology landing, so as to ensure the smooth development of products.

21st century: the cooperation mode between Geely and Baidu is different from that of Weilai, Xiaopeng and ideal. It may even be different from that of Huawei and Jihu. What do you think of the difference?

Yang diange: Wei Lai, Xiao Peng, Li Xiang, Li Bin, he Xiaopeng and Li Xiang are outstanding Internet elites who make cars in their personal capacity. In fact, it is the super product managers of the Internet that bring the concept, capital and talents of the Internet. However, a new group of enterprises, no matter Xiaomi, baidu or 360 or Didi, enter the automobile industry, which is the behavior of the company, not the individual behavior. It means that what they are doing is not only cars, but also driving the business and ecological integration of the original companies with intelligent cars.

In the future, we will see more combination of smart cars and mobile phone ecology, smart cars and data security, information security, and smart cars and mobile travel. We will see the integration and coupling of this format. This is the difference brought about by the new changes of smart cars.

There are many choices in the development path. All roads lead to Rome. We should pay more attention to our own strength and technical advantages. However, the choice of different roads will encounter different problems and difficulties.

The advantage of the cooperation between Baidu and Geely is that it can give full play to Baidu's technology accumulation in intelligent cars and automatic driving, and make full use of Geely's R & D capabilities in the production, manufacturing and even design of traditional automobiles, so that Baidu's automatic driving technology can be applied in industrialization with the fastest time, the smallest cost and the lightest assets.

In this mode, it will encounter difficulties. The development of intelligent vehicles, especially in the future, depends on the ability of rapid iteration. The data are constantly collected, the automatic driving technology is continuously iterated, and the automobile production and manufacturing are improved. The speed of automobile production, manufacturing, iteration, upgrading and modification is very fast. In this process, the two companies may encounter problems between cooperation, and further more complex ecological problems. Whether to form an automatic driving ecology around Baidu or Geely is a key and core issue for the two companies.

This problem has been encountered in the development process of Xiaopeng automobile. Xiaopeng automobile mastered the complete design and development ability in the early stage, but the production and manufacturing is the OEM of Haima. Why do we have to get the qualification and build our own factory at such a high price? It is still necessary to fully control the ability of the whole link from design to production to manufacturing. With this ability, we can ensure its fast iteration ability, ensure its advantages in competition, and give full play to the advantages of intelligent vehicle data-driven.

21st century: what is the essential difference between the new car manufacturing enterprises and the traditional automobile enterprises? How to see the transformation of traditional automobile enterprises? Any suggestions?

Xia Heng: the automobile industry is a long-distance race. In the era of intelligent electric vehicles, if enterprises want to survive, they must have their own long board. They should consider how to continue to reflect the differences and their own advantages in personalized places, and strengthen this advantage through various forms. Without its own characteristics and strengths, it is difficult for consumers to identify so many brands in such a fierce competition.

Second, unlike the Internet, automobiles can not have any short boards. No matter in R & D, production or marketing, there can be no short boards in all aspects. The automobile industry is a reverse "barrel theory", which requires both long and short boards. Once there are short boards, customers will be lost.

In the long run, we can win by quality. Innovation can make us fly high, and quality can make us run far. In this process, we must do a good job in safety and quality, which is our commitment to car owners of Xiaopeng.

Finally, in such a tide of innovation, luck is also very important. No matter who wins in the end, it is a good thing for China's intelligent electric vehicles and the development of the whole industry. Let's do a good job in the industry together. What we can gain is our consumers.

Xia Yiping: there are more differences in talent and organizational structure between the two sides. No matter the traditional automobile or the new car manufacturing enterprise, the core or the car, can not do without the car manufacturing process, quality control, and supply chain. These are the foundation. Only by doing these things solidly can we make good products. Without good products to the users, there will be no strong resonance and emotional connection.

Even if we talk about the dispute of software defined automobile and future intelligence, no matter who is going to do this, we should first of all be steady and steady, have the spirit of innovation and breakthrough, and at the same time, we should respect the traditional automobile industry, revere the experience brought by the traditional automobile industry, fully understand, digest and utilize the advantages of the traditional industry, be user-oriented and product oriented, Create the best products.

In the past, the methodology of automobile quality was biased towards mechanical manufacturing. In the future, software and hardware will be combined. Especially when software occupies a higher position in the future automobile, quality is not only hardware, not only manufacturing, but also the quality of software and user experience.

Therefore, from the traditional manufacturing quality, parts quality, vehicle quality or production quality in the past, it is constantly changing and evolving to how to ensure the software safety and meet the user's requirements for product experience, which also requires both sides to learn from each other.

For many new enterprises, the biggest challenge is whether they can produce integrated talents, so that people in the Internet can understand cars better, and people in the automobile industry can better understand the Internet and software. This is also the necessary basic condition for the continuous survival, evolution and iteration of intelligent electric vehicle enterprises or intelligent vehicles in the future.

What is the core of competition in the future?

21st century: how to see the competition of intelligent electric vehicles in the future? Will the future competition be chip operating system and ecology?

Xia Heng: Xiaopeng automobile attaches great importance to independent research and development. In the supply chain system, Xiaopeng automobile is still willing to cooperate with the strong to make good products together. If we have such a good partner as NVIDIA, Xiaopeng will join forces to do it instead of having to do it by ourselves. The purpose of Xiaopeng is to make consumers willing to use Xiaopeng's products through independent research and development, which is the most fundamental starting point of Xiaopeng's independent research and development.

However, in terms of automatic driving, Xiaopeng automobile insists on independent research and development, because everyone has no good experience in the field of automatic driving, and the iteration speed of relying on suppliers is very slow; Secondly, we worry about the quality and safety of products.

Xiaopeng's independent research and development is still in the field of automatic driving, and can also do better automatic driving. More than 95% of Xiaopeng's cars are equipped with self-developed advanced automatic assistance system. Most customers who buy Xiaopeng cars have bought automatic driving system, and the added value in this respect is very high.

And data is the most critical to automatic driving. Through data to drive the transformation of intelligent cars, leading the future travel. From the beginning of its birth, Xiao Peng believed that this round of AI revolution must be data driven.

Xia Yiping: the core of intelligence is the transition to unmanned, which is the ultimate competition. The future competitiveness may include chips and core algorithms.

In the future, automatic driving is not a problem of computing power at the end, but also the computing power of the background AI cluster. It is not something that a chip or an operating system can solve. With the same number of algorithms and the same amount of data, the future competitiveness depends on the iterative ability.

What needs to be solved in the next few years is to improve the server-side performance, and to be able to solve the massive data training in a very short period of time in a very cheap and fast way. This is also the basic requirement for unmanned driving to enter the fast lane, quickly iterate and upgrade rapidly in the future.

If the real L4 and L5 intelligent driving in the future is to be large-scale, safe, legal and compliant, it is necessary to think about the technology, products, solutions and ecology behind them, and constantly improve their core competitiveness from the perspective of development and ecology in the next product iteration process.

Yang diange: whether it is the competition of product function and performance or the competition of ecology, we should look at it in stages. If the core of competition is ecology, it is the ultimate form of competition. What will win in the end? If these enterprises are left in the world, what will enterprises rely on to win the competition? It depends on ecology. However, at this stage, it is still the most important for enterprises to ensure their survival. Only when they survive can they develop. At this stage, the survival may be more because of the product and its functions and performance. If these are done well, the enterprise can survive, develop and be qualified to finally enter the stage of ecological competition one day. At different stages, the key points of competition will change. We can't say that the ultimate competition is ecology. We must also seek ecological competition when we are very weak and seeking survival.

21st century: how does Xiaopeng define the automatic driving level of products?

Xia Heng: instead of defining products according to L2, L3 and L4, Xiaopeng automobile defines our products according to their own version numbers xpilot2.5 and xpilot3.5. For Xiaopeng, the current sensor density has a lot of L4 related functions and capabilities, some even higher.

The most essential difference is that Xiaopeng only provides a driving assistance system, not a pure unmanned driving. When the time is not ripe, Xiao Peng does not encourage drivers to ignore traffic rules.

There are 100000 traffic accidents in China every year. There are very clear regulations on how to punish traffic accidents and how to make legal procedures. People are not omnipotent. People will have accidents when they drive. But once the machine has problems, how should we face them? The current laws and regulations are completely blank. Before these are clear, Xiaopeng automobile will develop the functions of L4, but even in this case, consumers should always look ahead and be ready to take over.

Xiaopeng automobile is one of the first batch of "outstanding students" of China's first new automobile manufacturing forces. Its sales volume is firmly in the first echelon of new automobile manufacturing forces, and has successfully landed on the New York Stock Exchange- Visual China

21st century: when Jidu automobile launched L4 products in 2023, could its functions be implemented?

Xia Yiping: 100% of them will land. Baidu even has the actual scene of operating unmanned vehicles at a fee. At present, nearly 400-500 L4 unmanned robotaxi are running on the street every day. This year, it is planned to invest nearly 3000 robotaxi to promote the implementation of the whole unmanned technology.

In essence, set degree is more from the perspective of commercialization. First of all, we are very confident in technology, including what form and safe state or safe way the products built with L4 as the underlying core competence are provided to users. Users can use our products safely and safely, and how to deal with them more intelligently after being intervened in the midway.

The premise is to implement our products with user safety as the core guidance and laws and regulations as the core guidance. We hope to quickly bring the best technology to the market with the most civilian price and civilian products, so that we can use them. This is our plan for making products, especially for the first car in Jidu.

21st century: how to treat the neck jamming technology of intelligent electric vehicles? What needs to be done by car companies themselves, and what needs to be done by cooperation?

Xia Yiping: data is certainly very critical, for users to ensure the privacy of data, can not be abused. It is necessary to use some desensitized data to make relevant algorithm improvement or experience improvement, which is the core competitiveness in the future. But the problem is that in the face of massive data, how to train these data is a very challenging problem.

In the whole field of intelligent driving, domestic autopilot chips have appeared. From the perspective of development in the next three to five years, there is no special neck sticking factor in technology. Whether it is the core electric vehicle "three electric", or the so-called domain controller, OS and chip, domestic enterprises are not "stuck in the neck" in the core technology or technology chain in this field, and it is possible for China to create a world-class company in this field.

Yang diange: the main difference of future automobiles is not the difference brought by electrification, but the difference brought by intelligence. At the end of last year, Tesla's market value was equal to the total market value of Toyota, Volkswagen, Mercedes Benz and BMW. Battery, motor, electronic control, these can not form a real gap, the real gap is from the data. Tesla collects the so-called coner case data in millions of vehicles all over the world. These data in turn train its automatic driving technology, which can make Tesla's automatic driving technology progress much faster than traditional car companies. This is the most competitive advantage of autopilot.

In the future, data is the core for all automobile enterprises. With data, we can continuously improve the algorithm and iterative function. When it comes to iteration, it is more about the iterative upgrade of product software functions. The key of software function upgrade is data, which is driven and improved by big data. In the future, for all automobile enterprises, data is a very core ability.

The problem of neck sticking should be considered from the perspective of the state and enterprises. The automatic driving of intelligent vehicles involves national strategic security, the whole industrial chain and technical links. If there is neck sticking technology, the state must do it and have independent ability.

From the perspective of enterprises, whether to do chips, computing platforms, operating systems. The specific choice is closely related to data-driven. If big data is the core element of the iterative development of smart cars, when enterprises choose the technical route and do not do what technology, a core factor that needs to be considered is to choose or give up the independent development of a technology or parts. Whether it will affect their ability to fully grasp the core data is a very important judgment.

(interns Shi Xinran and Yang Liu also contributed to this article.)

?