Orient Fund Li Rui: In the future, intelligence will become one of the main driving logics of the automotive industry

In order to conform to the trend of a new round of scientific and technological revolution and industrial change, seize the strategic opportunities of the development strategy of the intelligent and connected automobile industry, build a highland and industrial agglomeration area for intelligent and connected automobile talents, and promote the construction of an automobile power, on October 15, the 2021 Automobile Capital Forum was held in Beijing.

It is reported that the forum, with the theme of "Nugget Opportunities in the First Year of Intelligent Driving", focused on the ecological chain of intelligent networked vehicles, automatic driving technology, artificial intelligence, and information security technology. As a professional event that attracts the attention of the autonomous driving industry and the intelligent and connected vehicle industry, the forum attracted the most influential automotive industry leaders, well-known investors, industry experts and professional media in the field of autonomous driving in China.

In fact, smart cars, including driverless cars, have become a hot topic of concern to regulators and industry in recent years.

On February 24, 2020, the relevant departments issued the "Intelligent Vehicle Innovation Development Strategy", which gave a clear policy orientation for China to conform to the trend of a new round of scientific and technological revolution and industrial change, and seize the opportunity of industrial development strategy; The following month, the relevant departments issued a public document on the "Classification of Automobile Driving Automation", which divided driving automation into 0-5 levels based on the degree to which the driving automation system can perform dynamic driving tasks, and according to the role allocation in the execution of dynamic driving tasks and whether there are designed operating conditions; By the end of that year, the "New Energy Vehicle Industry Development Plan (2021-2035)" issued by the relevant departments improved the top-level design of the development of new energy vehicles, and proposed to promote the mutual integration and coordinated development of electrification, networking and intelligent technology.

"In the future, intelligence will become the main driving logic of the automotive industry, and automatic driving has been clearly defined as the direction of development in the next 10 years." At the 2021 Auto Capital Forum held on October 15, Li Rui, deputy general manager of the Absolute Return Department of Orient Fund, made such a judgment on the industrialization development trend of autonomous intelligent vehicles.

In the view of Li Rui of the Oriental Fund, the first of the two waves of autonomous driving industry up to now comes from the expectation of autonomous driving industrialization. There are a lot of iconic events, the spin-off of Waymo, GM's acquisition of Cruise, Ford's investment in ArgoAI and many more. The background is that artificial intelligence technology has made great progress, deep learning has begun to be applied on a large scale in the field of perception technology such as image recognition, and sensor technology has also developed greatly. Therefore, it can be said that the application of deep learning in automatic driving is the main reason for the first intelligent driving investment. Capital had great expectations for its industrialization prospects, but at that time it was obviously blindly optimistic, the reality is that the long-tail problem has become an important constraint on the landing of high-level automatic driving, and the specific challenges lie in the robustness of hardware, the redundancy of the system, and the perfection of testing.

The second wave comes from the commercialization of autonomous driving, not just expectations. After 3-4 years of further technology accumulation, the core technologies of autonomous driving such as lidar, chips, perception and decision-making algorithms have been further developed, and autonomous driving in specific scenarios (mines, ports, airports, etc.) has been able to initially commercialize. Capital's expectations have also been gradually adjusted to a stage that coincides with the real progress of the industry. In this context, several factors further promote the second investment enthusiasm: (1) models with low-end automatic driving functions have begun to be listed and sold well, such as Xiaopeng P7, Model3/y, etc., and traditional OEMs have also begun to accelerate their transformation and have launched intelligent brands, marking the commercialization stage of smart cars. (2) Related companies have successively achieved listing, and the activity of the capital market has increased the certainty of investment in this field. In short, the gradual commercialization of autonomous driving in the second wave of opportunities is the core reason.

From the perspective of future development, the landing logic of automatic driving lies in: from low speed to high speed, from cargo to manned, from commercial to civilian. The future landing of automatic driving requires the cooperation of technology + scene operation, and the time for commercialization will vary according to different scenarios. Structured scenes have been able to achieve L4 level unmanned operation, but each different scene has different requirements for technical details, such as mine scenes with magnetic field interference, poor GPS signals, and harsh working environment, more dust; The port scene is often subject to rain and wet weather; Airport logistics requires a high level of security. These unmanned vehicles need to be connected with the overall operation management scheduling system, and they need to have a deep understanding of the scene and make targeted changes to the algorithm to achieve the real landing of automatic driving, and at the same time, they must be able to continuously obtain data through operation to continuously carry out technical iteration.

Compared with the mass production car, the test vehicle can be equipped with more sensors, the collected data is more comprehensive, a better simulation test environment can be established, and the algorithm development progress is more advanced. Progressive routes can be launched on a large scale, accumulating more perception and driving data (scene traffic conditions, etc.), these data (such as calibrated image data, cornercase driving conditions, etc.) and decision algorithms can be migrated to L4 applications. Individuals are more optimistic about the gradual development route, but the premise is that this manufacturer has a strong research and development strength comparable to L4 level. ”

While analyzing the development prospects of the autonomous vehicle industry, Li Rui of the Oriental Fund also expressed his views on the industrial chain and competition pattern.

For the industrial chain, Li Rui of oriental fund suggested paying attention to the investment opportunities of AI chips, algorithms + data, etc. to build the most core links of automated vehicles. "1) AI chip as the highest value of the bicycle hardware link, the choice of the corresponding chip is to choose the corresponding ecology. Domestic AI chips are expected to gradually become the choice target of independent substitution. 2) Algorithm + data level, is the highest value of the software link, software, algorithms, data is not only closely related to functional security, but also involves data security, the government also gives a large degree of policy, planning, testing support. At present, domestic startups have a certain scale, and companies in this field also have opportunities. ”

The second concern is the lidar in the sensor, the current semi-solid/solid-state lidar solution appears, mechanical lidar from the expensive hundreds of thousands of dollars to 600-1000 US dollars range, and can pass the car gauge detection, thus providing good conditions for mass production loading.

Finally, the chassis execution unit, because it involves driving safety, in the past, braking and steering were monopolized by the international Tier1, and the process of independent replacement was relatively slow. Intelligent driving requires the assistance of the wire control system in the execution link, and each OEM has the need to integrate the chassis domain and the chassis actuator, so it also creates a better opportunity for domestic substitution. ”

For the competitive landscape, Li Rui of the Oriental Fund said bluntly, "OEMs and technology companies will be a competitive relationship." The head car company will most likely develop its own intelligent driving system. On the one hand, automakers have historically belonged to the manufacturing industry with low profit margins, and the supply chain optimization has been relatively mature, and there is little room for continued cost reduction. Under the wave of intelligent electric, OEMs are expected to usher in higher bicycle value and profit margins, while capturing users through data and systems. Therefore, head OEMs have a strong power to develop core technologies.

The head of the traditional OEMs and the head of the technology companies are mainly in competition, and there will be business cooperation with other technology companies, and even these technology companies may be acquired by the head of the traditional vehicle factory. If the head of the technology company, if the capital force is strong enough, I think it is not ruled out that there will be a possibility of building a car. Traditional OEMs outside the head will cooperate more with technology companies, but dominance and benefit distribution mechanisms are the focus of attention.

In addition, the new forces of car manufacturing, and technology companies are mainly competitive relations, their own software capabilities are relatively strong, the organization and incentive mechanism is flexible, and the independent research and development of automatic driving technology has a great supporting effect on the valuation of its secondary market, so it will take less external cooperation, and even the automatic driving chip will start to carry out independent research and development. ”

(Caution should be exercised when there is a risky investment in the market)