Source: MEGA WAVE

Author: Dong Erqian

Our world is moving towards a more intense Matthew effect, where the global economic power is increasingly concentrated in a few large countries, and the competition and game between these countries is becoming more intense, especially evident in advanced technology fields represented by artificial intelligence.

With the continuous deepening of the new round of technological revolution and industrial transformation, technological innovation has become a key lever for major countries around the world to seek endogenous economic development and industrial competitive advantages, prompting countries to increase investment in research and development of new technologies and industrial layout.

As the most typical application of large-scale model reconstruction of the physical world, self-driving cars have become an important direction for major countries to layout the future.

On the other side of the ocean,$Tesla (TSLA.US)$Introducing self-driving taxis named 'Cybercab,' Tesla's driverless vehicle without a steering wheel or pedals, costing less than 0.03 million US dollars.

"This product will make tesla a company with a market cap of $10 trillion," Musk once said, "People will still talk about this moment a hundred years from now."

Carrot Run has already released the sixth generation self-driving car in 2022, with two optional modes: with or without a steering wheel, with a cost also lower than 0.03 million US dollars. However, unlike Tesla's expected launch of self-driving cars in 2026, Carrot Run's sixth generation self-driving car has already started testing.

Alphabet, Google's parent company, also stated in the second quarter earnings call that it will invest an additional 5 billion US dollars in its self-driving subsidiary Waymo in the coming years to create the world's leading self-driving technology company.

Currently, self-driving technology is at a stage where China and the USA are leading the world, while most other countries are in a stage of simply participating.

In China, the most core player in self-driving technology is$Baidu (BIDU.US)$Its subsidiary Luobo Kuai Pao had already achieved a milestone of 7 million kilometers nationwide.

Interestingly, Cybercab's various innovations seem to have already been laid out by Luobo Kuai Pao, so some netizens jokingly refer to this as a tribute to "domestic radishes". While Cybercab has no steering wheel, pedals, rearview mirrors, or drivers, Luobo Kuai Pao released its sixth generation of self-driving cars over two years ago, supporting both steering wheel and steer-by-wire modes.

Musk said that the expected cost of Cybercab will be less than 0.03 million US dollars, but Luobo Kuai Pao's sixth generation vehicle, the Yichi 06, already costs less than 0.03 million US dollars, indicating that Luobo Kuai Pao is more mature in its application and implementation.

Especially since Cybercab will not be launched on the market for another 2 years, Luobo Kuai Pao has plenty of time to further enhance its products and technologies.

In May, Baidu released the world's first automatic driving large model Apollo ADFM that supports L4-level autonomous driving applications during Apollo Day, which can ensure both the safety and generalization of technology, achieving a safety level over 10 times higher than human drivers and covering city-wide complex scenarios.

Luobo Kuai Pao even wants to seize the time window to further enter the Cybercab's territory. According to some media reports, Luobo Kuai Pao is actively expanding globally, conducting in-depth communications with several international companies, and planning to enter overseas markets.

Industrial competition is fierce, full of various strategic maneuvers. For example, the USA has frequently put the brakes on the development of China's self-driving industry, restricting our country's self-driving technology from entering based on data security issues. Against this backdrop, the mileage of Chinese self-driving cars on California roads during road tests in the USA in 2023 has dropped by about 70%.

In the context of unprecedented major changes, self-driving is a "battle that no one can afford to lose" for any party, and we will inevitably see a more intense competitive situation.

Important race track

A must-win territory.

As early as 2004, the Defense Advanced Research Projects Agency (DARPA) of the United States Department of Defense held the first unmanned driving car challenge in the Mojave Desert in California. All participating cars failed not long after the starting point, leading to the competition being jokingly referred to as the 'Desert Big Crash' by DARPA.

However, the story of self-driving cars did not stop there. Subsequent challenges have also achieved significant results. Google and many other technology companies and car manufacturers have since started to develop self-driving technology, marking the beginning of a revolution in the automotive industry.

The road to large-scale operation of autonomous driving is much longer than anticipated. One important reason is that traditional technical solutions struggle to handle complex scenarios. 'Autonomous driving has not fully achieved its goals in the past two decades. In addition to reasons such as insufficient sensor hardware and computing power, another reason may be the methods being outdated. The most likely way to achieve it is through data-driven AI model training and deployment for true self-driving,' shared Gu Weihao, CEO of Mo Zhi Xing.

In 2017, the Transformer architecture was born at Google, capable of training more powerful models. That same year, Musk poached Karpathy from OpenAI to refactor the autonomous driving code, utilizing Transformer+BEV (bird's-eye view) to hand over the perception part to the large model.

Around the same time, Baidu decided to reconstruct its autonomous driving technology stack with large models—full-system modelization began in 2017, shifting to data-driven; in 2021, Baidu Apollo gradually integrated multiple small model tasks in the system, expanded model scale, explored large autonomous driving model technology, and subsequently fully deployed large models on vehicles.

Self-driving, which integrates artificial intelligence, communication, semiconductors, automobiles, and other technologies, involves a long industry chain and enormous value creation space. It has become a must-win territory for cross-border competition and collaboration between the automotive industry and tech industry in various countries.

According to data from Zhishi Consulting, the global self-driving car market is expected to reach $1.724 trillion in 2030, with the Chinese market potentially reaching $634 billion.

"In the past two years, a new round of technological revolution and industrial transformation is reshaping the global innovation landscape and restructuring the global economic structure. Automatic driving, representing new productive forces, has become a key area of global competition," said Xu Lijuan, director of the Market Research Institute at Beijing Jiaotong University. "As a typical application scenario empowering the automotive industry with artificial intelligence, self-driving cars involve the development of related industries such as chips, operating systems, the Internet of Things, urban infrastructure, and detection and certification, which are an important track for the integration of digital economy and real economy."

Key competition

The 'Three Big Families' of global autonomous driving

The self-driving taxi market is a blue ocean, as well as a huge cake that has not yet been allocated. As the two countries with the most autonomous driving companies globally, China and the USA, whoever can first complete the commercial loop and establish related industrial chain will have the ability to define the track and export technology products to other countries.

Looking at China's active commercialization of self-driving taxis, it is clear that China indeed has a strong intention in this area. At present, only Baidu is basically capable of undertaking this task.

In May, the foreign media EV Magazine selected a global ranking of self-driving technology companies, mainly evaluating the technical maturity, research progress, and product capabilities of L4-level self-driving systems, with 3 US companies listed in the 'Leaders' quadrant. $Baidu (BIDU.US)$ being the only Chinese company selected in this quadrant.

Amid strong demand in major markets such as Japan, Europe, and North America, this world's largest auto manufacturer has increased its share buyback program to 1.2 trillion yen ($8.3 billion). $Tesla (TSLA.US)$ has released the self-driving robotaxi and as the representative of Chinese enterprises, is directly competing with Baidu and Tesla, forming a competitive situation among the 'three major players' in the self-driving industry in China and the USA. $Alphabet-A (GOOGL.US)$

The United States has taken the lead in the research and development of self-driving technology. With advanced integrated circuit technology, the U.S. has always maintained a leading position in high-end chip design, laying a solid foundation for the development of high-performance vehicle-mounted chips. Whether it's the 'Waymo camp' or the 'tesla camp,' they both lean more towards a 'single vehicle intelligence' solution, with the core capabilities being artificial intelligence algorithms and decision-making chips.

Chinese companies have found another differentiated path, with China's advantage lying in well-established infrastructure. Therefore, companies like Luobo Kuaipao are actively promoting the 'single vehicle intelligence + vehicle-road coordination' solution, building an intelligent road network that perceives data across 'people, vehicles, roads, and the cloud' through the collaboration between 'smart vehicles' and 'intelligent roads.'

In terms of data accumulation, Tesla's FSD has accumulated over 1.6 billion miles of driving miles, while Waymo's self-driving taxi service in the U.S. has exceeded 0.1 million paid trips per week. Currently, the only domestically driven enterprise capable of large-scale and normalized testing in the self-driving sector is Luobo Kuaipao, with its actual road testing and demonstration mileage surpassing 0.1 billion kilometers as of June 2024.

It can be inferred that there is still some gap in the field of unmanned driving between China and the United States, but it is not insurmountable. However, industry competition on a personal level is often more brutal than we imagine.

In July 2023, four lawmakers wrote to the U.S. Secretary of Transportation and Secretary of Commerce requesting an investigation into the situation of Chinese autonomous driving technology in the United States and how to restrict it. The letter stated, 'Technology used for autonomous driving vehicles, lidar, radar, cameras, artificial intelligence, and other advanced sensors and semiconductors can be used to collect data on the American people and infrastructure which can be shared back to China.'

Subsequently, many Chinese unmanned driving companies have successively withdrawn from the U.S. market.

As Vice Chairman of the China Association for Science and Technology and Professor at the University of the Chinese Academy of Sciences Lv Benfu said: "Currently, it is the window period for the global development of intelligent connected cars, the opportunity is fleeting, and companies need to actively embrace intelligent connected cars without hesitation in order not to fall behind in the future core technological areas."

Economies of scale

Landing applications.

What few people pay attention to is that, in this competition, China has actually taken the most crucial step.

The self-driving industry has gone through a long period of downturn. The core reason is that the low revenue cannot cover the high costs, and many companies have failed in the face of the market and consumers' big test. In this game of giants, those companies lacking technology, capital, and the ability to apply landing are ultimately difficult to succeed.

At Baidu's second quarter financial report conference call in 2024, Baidu's founder, chairman, and CEO Robin Li also revealed that its self-driving taxi business will achieve regional revenue balance.

Specifically, the driving force to achieve profitability lies in cost reduction and demand enhancement.

The high manufacturing cost of complete vehicles is currently the core reason for the high unit service cost, but an effective cost reduction path seems to have been found. As observed from the perspective of cost-effectiveness, the sixth-generation self-driving car Yato 06 is priced at only 0.2046 million yuan, with a cost below 0.03 million US dollars, a 60% decrease compared to the fifth-generation self-driving car.

The increase in demand comes from the continuous expansion of the scale and coverage area of the deployed vehicles, which heavily relies on the relevant regulatory authorities providing a positive, open, and inclusive policy environment for the self-driving cars industry, fully leveraging China's abundant scene advantages to accelerate implementation.

Professor Chen Yanyan from Beijing University of Technology's Urban Transportation College has also mentioned: "Currently, most cities in China that are developing self-driving are conducting limited commercial trial operations within specific areas and time frames. The pace of local legislation needs to be accelerated to strongly support self-driving companies in expanding their operating areas and assisting in achieving large-scale implementation."

After all, historical experience has proven that as long as we can keep up with the actual conditions in the United States in terms of technology, we can surpass them in total output value and profits based on market and scale advantages. Simultaneously, massive capital investment can feed back into the industry development, ultimately achieving technological advancements as well.

Facing competition from Tesla's Robotaxi, Chinese players like Pony.ai still have the potential to continue releasing energy.

In conclusion,

The increasingly intense competition in technological innovation between emerging economies and traditional powers is becoming more pronounced.

Developed countries like the USA, relying on inherent advantages, continue to accelerate progress while maintaining a leading position in technological innovation, but they always have various problems in terms of commercial implementation. In contrast, China, relying on infrastructure, market advantages, and policy support, continues to introduce more cutting-edge models to the market, forming a scale advantage, and constantly iterating and innovating, allowing it to "be ahead by dint of coming later".

Various problems may arise in the development of emerging industries, but the final result is always satisfactory. In various internet technology fields, there are already numerous successful cases.

With this trend, the Chinese technology industry will eventually complete a new successful path. Just like the description in the State Council's "Development Plan for the New Generation of Artificial Intelligence" released in 2017, by 2030, China's artificial intelligence theory, technology, and applications will overall reach a world-leading level and become a major innovation center for artificial intelligence in the world.

Editor/Rocky