Automakers Embrace the World of Robots

In 2007, the science fiction movie "Transformers" ignited the imagination of machinery enthusiasts worldwide, depicting a future where mechanical beings could transform freely. This vision captivated countless minds.

Fast forward to 2022, when Tesla unveiled its self-developed humanoid robot, named "Optimus." In March 2025, Tesla CEO Elon Musk released a video announcing that Optimus would enter the trial production stage that year.

With the continuous evolution of automotive technology, the integration of intelligence and machinery is deepening, signaling the dawn of the humanoid robot era.

Recently, UBTech, which has established its humanoid robot headquarters in Beijing's Economic-Technological Development Area, has accelerated its commercialization efforts. By the first half of this year, 20 UBTech industrial humanoid robots, known as Walker S1, will be deployed in Dongfeng Liuzhou Automobile's manufacturing factory. These robots will be applied to the production of complete automobiles, enhancing the factory's intelligence, unmanned operations, and efficiency. According to media reports, this will mark the first global batch deployment of humanoid robots in automobile factories.

On March 25, the Audi FAW New Energy Automobile Production Base also announced the successful trial operation of industrial humanoid robots performing air conditioning leakage detection tasks, officially ushering in humanoid robots into the automotive manufacturing arena.

Furthermore, the pace at which automakers are venturing into "making humans" is rapidly accelerating.

On February 23, Xiaomi announced significant progress, with its full-size humanoid bionic robot CyberOne (Tie Da) moving from the laboratory to phased implementation on the production line. Standing at 177 centimeters tall and weighing 52 kilograms, CyberOne is designed for home care and companionship.

Xpeng Motors released its AI humanoid robot Iron in November 2024. This robot, built at a 1:1 scale with humans, boasts 62 degrees of freedom for its entire body and 15 degrees of freedom for its hands. Meanwhile, GAC Group unveiled its independently developed third-generation embodied intelligent humanoid robot, GoMate, aiming to achieve mass production of self-developed components by 2025. GAC plans to lead the way in whole-machine demonstration applications within its workshops and production lines, including those of GAC Trumpchi and AION, as well as in industrial parks.

According to incomplete statistics, more than ten automakers, including BYD, Chery Automobile, Changan Automobile, Toyota Motor, Beijing Automotive Group, SAIC Motor, Li Auto, and Thalys, have expanded their businesses into the field of humanoid robots through investment or independent research and development.

"If smart cars are the crown of intelligent manufacturing, then intelligent robots will be the crown of machine intelligence," predicts He Xiaopeng, Chairman and CEO of Xpeng Motors. "I believe that in the future, manufacturers of smart cars will also be manufacturers of intelligent robots. Smart cars and intelligent robots will eventually converge into a unified industry, generating a 1+1>2 effect."

The First Year of Mass Production

Machinery, batteries, intelligence... Nowadays, automotive enterprises focused on electrification and intelligence have natural advantages in humanoid robot research and development.

Experts have noted that with the deep integration of artificial intelligence technology and the automotive industry, the homology between robot and smart car technology is growing stronger. Both are exemplary applications of embodied intelligence, and automakers' technological accumulation in autonomous driving environment perception, human-computer interaction, and motion control provides a solid foundation for robot R&D.

The industry once described advanced smart electric vehicles as "robots on wheels." In He Xiaopeng's view, with the advancement of AI automobiles, general-purpose humanoid robots have become the intersection of the automotive and robotics industries, reflected in five integration dimensions: supply chain, technology, scenarios, channels, and brands.

"Taking an automaker with full-stack self-research capabilities like Xpeng Motors as an example, from the AI software operating system and interaction algorithms to the hardware-end electronic and electrical architecture, domain controllers, chips, and the research and development of three-electric technology in the power system, the overlap of these technological reserves with the technology stack of robot research and development reaches 70%," said He Xiaopeng.

Zhang Aimin, head of the robot R&D team at GAC Group, also stated that from a technical perspective, there are many similarities between automobiles and robots. In terms of cloud services, functional services, electrical architecture, and perception technology, the two are highly compatible. The automotive system platform, encompassing underlying sensors, automotive-grade chips, real-time operating systems, and ROS-based communication middleware, can provide robust support for humanoid robots.

Specifically, algorithms such as environment perception, path planning, and decision-making control in autonomous driving technology can be applied to the navigation and obstacle avoidance of humanoid robots. Algorithms in human-computer interaction technology, like speech recognition and natural language processing, can enhance the interaction capabilities of humanoid robots. Additionally, technologies such as motor drive and sensor fusion in motion control technology can optimize the movement execution of humanoid robots.

With the fierce competition in this field, automakers' progress in "making humans" is also accelerating. Many automakers have successively released plans, and the humanoid robots they produce will gradually be put into use this year and next.

According to Founder Securities' forecast, 2025 will be the first year of mass production for humanoid robots, with 2026 ushering in an explosion of commercial applications.

Pacific Securities analyzed that automakers entering the field of humanoid robots can achieve technology homology by reusing vehicle-end algorithms on humanoid robots. It is anticipated that with the participation of more enterprises in the future, the humanoid robot industry chain will become more complete, and the possibility of commercialization will also increase.

Scenario Advantage

In terms of application scenarios, automotive enterprises also hold unique advantages. Wanlian Securities stated in a related research report that workshop factories are crucial application scenarios for humanoid robots, and automakers naturally have a first-mover advantage in the manufacture and application of humanoid robots.

He Xiaopeng also believes that the challenge of scenario implementation is actually a significant opportunity for automakers. Automotive enterprises possess tens of thousands of manufacturing process nodes and thousands of sales and service networks. These ready-made industrial and commercial scenarios provide a natural testing ground for robot technology verification.

Xpeng Motors' self-developed Xpeng Iron robot has officially entered Xpeng Motors' factories to undertake assembly and other production tasks.

BYD, on the other hand, is investing in robotics companies such as Zhiyuan Robotics while intensifying its self-research efforts. In December 2024, BYD's official WeChat public account released information on "BYD's 25th Special Recruitment for Embodied Intelligence Research Team," with recruitment directions involving humanoid robots.

Meanwhile, Shenzhen UBTech Robotics Co., Ltd., has also officially released its new generation of industrial humanoid robot, Walker S1, which has successfully entered BYD's automobile factory for practical training. It works in coordination with L4 autonomous logistics vehicles, autonomous forklifts, industrial mobile robots, and intelligent manufacturing management systems.

Changan Automobile announced at the 2024 Guangzhou Auto Show that it will carry out related industrial layouts in the field of humanoid robots, including humanoid robots and automotive ecological robots, and will release humanoid robot products by 2027.

However, it cannot be overlooked that despite the vast market prospects and high added value, the industry is still in its nascent stages and faces numerous severe challenges. Issues such as achieving human-like perception abilities and high costs still impede the large-scale development of humanoid robots.

Musk once stated that the cost of humanoid robots must be controlled between US$20,000 and US$30,000 to achieve mass production, indicating that there is still a long way to go in reducing the cost of humanoid robots.

He Xiaopeng also revealed that Xpeng Motors is developing a humanoid robot equipped with a skin perception system that can provide feedback on temperature, pressure, and even pain. In small videos of robots posted online, some people kick them to test their balance. If the robot were intelligent, it would say, "I'm in pain." However, this capability is still not yet achievable.

In recent years, robots have become increasingly prevalent in various scenarios of daily life, such as catering, logistics, and reception. Although most are still at the stage of performing simple and repetitive tasks, people's acceptance and expectations for them are high. The mass production of humanoid robots is even more anticipated, particularly in home scenarios.

The "Research Report on the Development of the Humanoid Robot Industry (2024)" released by the China Academy of Information and Communications Technology predicts that from now to 2028, China's entire humanoid robot market size will be approximately RMB 20 to 50 billion. By 2035, the market size will reach approximately RMB 50 billion. After 2045, when there are over 100 million humanoid robots in use across various industries, the entire market size could reach approximately RMB 10 trillion.

This article is originally created by China News Auto. Welcome to share it with your friends. If media needs to reprint it, please indicate the author and source before the text. Any media or self-media using any content of this article to produce video or audio scripts will bear legal responsibility.