The development of intelligent vehicles is opening up an industrial space beyond automobiles for China. Intelligent vehicles, embodied robots, and some low-altitude aircraft are forming closer links in fields such as perception computing, artificial intelligence, materials, and electrification technologies, and through supply chain migration and scenario collaboration, they are constituting new related industrial clusters. This cross-carrier integration of technology, manufacturing, and applications can be summarized as "aggregated intelligence." Its significance lies in driving artificial intelligence from information processing toward perception, decision-making, and action in the physical world. The next five to ten years are an important window for China to transform its automotive industry advantages into aggregated intelligence industry advantages. Whether this opportunity can be seized depends on whether common capabilities can be effectively transferred, whether products can achieve economic deployment, and whether sustained productivity improvements can be formed.
01 ONE Whether Automotive Advantages Can Be Transferred — The Next Five Years Are a Critical Window
First, intelligent vehicles, embodied robots, and some low-altitude aircraft are forming related industrial clusters that share part of their technology foundation, supply chains, and innovation resources. AI large models and agents provide a common methodological foundation, perception and computing power constitute common intelligent hardware support, and provide transferable engineering capabilities for software tools, electric drive systems, and precision manufacturing. Aggregation focuses on underlying capabilities, while professional division of labor is still reflected in products and markets. Technical methods are interconnected, but retraining, design, and validation are needed for different devices; customers, usage conditions, safety responsibilities, and business models also each have their own boundaries.
Second, "from automobiles to embodied intelligence" has entered the stage of practical validation, showing a pattern of "common platforms and professional division of labor developing simultaneously." The integrated layout of pioneering enterprises at home and abroad, the cultivation of second growth curves, cross-border talent flows, and supply chain migration all demonstrate that automotive accumulation has cross-industry value. Independent robot companies entering automakers' factories to undertake real tasks demonstrate the path of specialized collaboration. For automotive companies, they can choose complete machines, core components, technology platforms, or scenario cooperation; for embodied intelligence companies, making good use of the automotive industry's technology chain can accelerate engineering and mass production processes.
Third, the challenges of the two industries are complementary. The automotive industry needs to transform its long-accumulated technology, manufacturing, and supply chain capabilities into new value; the embodied intelligence industry needs to solve reliability, cost, and delivery issues. Letting the automotive supply chain, scenarios, and mass production capabilities serve the industrialization of embodied intelligence, and letting embodied intelligence open new demand space for automobiles, is the basic approach to developing the aggregated intelligence industry. The benefits of integration are first reflected in lowering engineering thresholds and shortening mass production ramp-up cycles, and only then can they be reflected in revenue and profit. The second growth curve must ultimately be proven by orders and sustained delivery.
Fourth, aggregated intelligence should be "both like automobiles and beyond automobiles." Like automobiles means drawing on their scale manufacturing, quality management, and supply chain organization capabilities; beyond automobiles means looking at both domestic and international markets from the very beginning, incorporating international market demand into product definition and operating systems. Overseas markets may become regions where new technologies and new scenarios are first implemented, and internationalization should be an important development orientation for embodied intelligence enterprises.
02 TWO Automobiles, Robots, and the Low-Altitude Economy Are Forming an Aggregated Industry
New energy vehicles have already established a scale foundation. From January to July 2026, China's new energy vehicle sales reached 9.007 million units, accounting for 51.2% of total new vehicle sales; in July, sales were 1.561 million units, accounting for 60.4%. Figures include exports. Electrification has provided a mature electric drive and manufacturing foundation, while intelligent capabilities have created the conditions for automobiles to enter the aggregation era.
Humanoid robots are moving from laboratories to production line manufacturing and real-world scenarios. XPeng announced the launch of a humanoid robot production line, and BMW disclosed the continuous operation of robots in its automotive factories, demonstrating progress on both the manufacturing and application sides. Industrialization has entered the stage of practical validation; the next step depends on stable delivery, effective working hours, and customer repurchases.
The low-altitude economy is landing while being regulated. In February 2026, China released the "Airworthiness Standards for Power-Lift Aircraft." Europe had previously established dedicated airworthiness conditions for vertical take-off and landing aircraft.
The aggregation of technology and industrial capabilities gives automotive manufacturing accumulation the opportunity to support the development of robots and low-altitude products. The same underlying technology serving multiple products can expand the applicable scope of R&D investment.There is also collaborative demand in application scenarios: a single factory, park, or logistics node may simultaneously require vehicle transportation, robot operations, and drone inspections. Through task scheduling, information interfaces, and workflow collaboration, overall efficiency can be improved. Technology aggregation plus scenario aggregation is the internal driving force behind the formation of an "aggregated industry." Whether broad and sustained productivity improvements can be achieved is the core indicator for observing the quality of aggregated intelligence in the next five to ten years.
China's key advantage is transforming its complete industrial chain, large-scale manufacturing application scenarios, and technological innovation strength into deliverable and replicable products and solutions.
03 THREE From Automobiles to Embodied Intelligence — Four Types of Practice Have Entered the Validation Stage
Automobiles and embodied intelligence are exploring technology and manufacturing integration within the same enterprise. Tesla has incorporated both automobiles and Optimus into the same corporate layout, with its official website listing the Fremont factory as a production base for Model 3, Model Y, and Optimus. XPeng announced the launch of a humanoid robot production line in September this year, extending automotive-grade manufacturing and quality systems to robot production. Infineon provides semiconductor solutions around functions such as drive, perception, and connectivity for humanoid robots. Practice shows that the value accumulated in automobiles can be extended across products, and actual results are still tested by product consistency, delivery capability, and application effectiveness.
Automotive companies are exploring the second growth curve of the embodied intelligence industry through self-development, investment, and incubation. GAC incubated and established Huilun Technology in February 2026, undertaking humanoid robot R&D and market-oriented business. Independent incubation helps leverage the technology, supply chain, and scenario resources of vehicle manufacturers, while building product and operating systems around robot customers. New businesses should be invested in by stages: first find tasks that customers are willing to pay for, then validate product stability and delivery costs, and finally expand production capacity.
Automotive talent entering embodied intelligence brings not only algorithm experience, but also methodologies in electronic and electrical architecture, embedded software, supply chain development, production processes, and quality management; at the same time, they need to supplement professional knowledge in contact operations, motion control, and other areas.
The "one-to-many" opportunity in supply chains is particularly prominent in areas such as electric drive transmission, thermal management, perception computing, and material processes. The automotive industry's accumulation in electric drive, thermal management, chips, materials, and precision manufacturing can support robots, but redesign and validation around tasks are needed. XPeng announced that its humanoid robot uses the Turing AI chip, demonstrating a pathway for underlying computing power to serve different products.
The value of the automotive supply chain lies not only in procurement prices, but also in lowering engineering thresholds, shortening mass production ramp-up cycles, and improving delivery stability. Actual transfer can be divided into direct reuse, modified reuse, process migration, and supplier sharing.
04 FOUR Five Pathways for Integrated Development: From AI Technology Foundation to the "Standards, Testing, and Certification" System
Automotive intelligence and embodied intelligence share methodological intersections in model learning, environmental understanding, and simulation training. The "common foundation" should first be implemented as shareable methods, tools, and infrastructure, including simulation tools, data interfaces, training and evaluation systems, and transferable learning methods. Automobiles mainly solve mobility problems in road environments, while robots must also solve grasping, contact, and manipulation problems, so models and data need to be re-adapted to tasks. Priority should be given to links that can validate transfer benefits, forming comparable results in development cycles, sample requirements, and task success rates, achieving "laying eggs along the way" and landing segment by segment.
Storage, computing power, connectivity, and perception are important intelligent hardware foundations for the aggregated industry. In storage, attention should be paid respectively to bandwidth and power consumption, software-hardware collaboration, latency and reliability, and multimodal perception combinations. The value of "storage-computing-connectivity-perception" lies in system coordination; stably completing tasks is more important than individual parameters.
Connecting material properties, parts processing, transmission design, drive control, and complete machine operating conditions is an important pathway for supply chain integration. Joint development can be carried out around modules such as joints, end effectors, power supplies, and heat dissipation. First clarify tasks and working conditions, then determine interfaces, lifespan, and cost targets. Supply chain enterprises should "keep one eye on automobiles and one eye on embodied intelligence," allowing existing capabilities to continuously transform, giving the automotive industry new sources of demand and giving the embodied intelligence industry more mature manufacturing support.
Automotive factories are embodied intelligence validation scenarios worth prioritizing for opening up. Tasks such as material distribution, some loading and unloading, inspection, and repetitive handling make it easy to clarify cycle time, quality, and safety requirements. BMW disclosed that Figure 02 accumulated approximately 1,250 hours of operation and handled more than 90,000 parts in its 2025 Spartanburg factory pilot. This kind of continuous operation evidence is more helpful than a one-time prototype demonstration for judging industrialization progress.
The pace of application should be based on task maturity. First achieve stable operation in tasks with clear boundaries, then expand to more processes and complex environments; high-risk tasks must first meet corresponding safety and reliability requirements. Each project should record task success rates, manual intervention frequency, effective working hours, fault downtime, and comprehensive usage costs.
The automotive industry's long-accumulated experience in reliability management, process traceability, and testing and validation can support the establishment of a "standards, testing, and certification" system for embodied intelligence. Priority should be given to advancing common component testing specifications, task performance evaluation, and laboratory capability sharing, so that enterprises can reduce redundant construction and product performance and reliability can be more easily compared. Robots, automobiles, and low-altitude aircraft each have their own safety boundaries and need to be validated separately. Cross-industry cooperation should start with the reuse of test data and evidence acceptance under clear conditions, while complete machine certification is still evaluated by the corresponding systems.
05 FIVE Adhering to "Born Global" to Build a New Competitiveness Model
The internationalization experience of China's automotive industry is worth drawing on. The embodied intelligence economy should consider both domestic and international markets from its inception, incorporating international demand into product definition and delivery systems. A recent research trip to Europe offered an insight: local manufacturing upgrading demand can serve as an entry point for cooperation on China's embodied intelligence and AI solutions. It should start from specific enterprises and processes, combining local industry experience with China's engineering development and supply chain capabilities. Similar demand still requires on-site adaptation, and local delivery and maintenance capabilities must be established. Being "born global" also requires making choices: first achieve breakthroughs in markets with clear demand, available partners, and the ability to form a closed delivery loop, then expand to similar scenarios.
To evaluate the competitiveness of aggregated intelligence enterprises, three dimensions can be referenced:
First, whether the technology is "one-to-many" — whether one underlying capability can serve multiple products and reduce the development cost of new products;
second, whether the market is "international plus domestic" — whether the enterprise can understand different market demands and build delivery capabilities;
third, whether it is "AI plus manufacturing" — whether it can transform intelligent capabilities into reliable and economical physical products.
This constitutes a "new triangle model" of competitiveness. The so-called AI intensity should be reflected in task capabilities and continuous improvement effects. All three dimensions must ultimately land on business results: whether the next product can be developed more economically, whether the next customer can be deployed more smoothly, and whether each task can generate sustainable returns. This model is equally applicable to evaluating industrial layouts in automobiles, embodied intelligence, and related low-altitude fields.
06 SIX Six Policy Recommendations
First, incorporate aggregated intelligence into the overall framework of strategic emerging industry layout, and coordinate technology and platform resources under the approach of "three major industries, one common foundation." Public investment should prioritize supporting common R&D, testing, and scenario services to reduce redundant construction; support cross-industry innovation platforms in playing their roles in technology transformation, supply chain matchmaking, and scenario matching, while retaining multiple technology pathways and professional division of labor.
Second, support breakthroughs in underlying hard technologies of "storage, computing, connectivity, and perception" and key actuation components. Around chips, sensors, connectors, materials, and actuation modules, promote joint development between complete machine and component enterprises, taking actual adoption, reliability, and cost improvement as important evaluation criteria.
Third, drive industrialization through scenario opening. Support the systematic opening of scenarios such as automotive factories, logistics, and after-sales service to embodied intelligence enterprises, and encourage the co-building of data closed loops. Support for first sets and first batches should be linked with safety validation, real operation, and application effectiveness, helping products move from demonstration to sustained procurement.
Fourth, accelerate the establishment of a "standards, testing, and certification" system in the embodied intelligence field. Drawing on automotive-grade experience, formulate reliability testing specifications and risk assessment methods for common components, and promote the opening and sharing of automotive-grade laboratory capabilities. Advance the acceptance of testing evidence in links with consistent applicable conditions and sufficient validation, while complete machine certification still follows the requirements of each field. Participate in international standards cooperation through reliable products and validation experience, transforming industrial advantages into rule-building capabilities.
Fifth, support aggregated intelligence enterprises in taking the internationalization path from the very beginning. Include embodied intelligence and low-altitude products in the scope of going-global support policies, and establish compliance guidelines for aggregated intelligence going global; drawing on the lessons and experience of the automotive industry's internationalization, guide enterprises to follow a high-quality going-global path of local value creation and compliant operations from the start; support the establishment of application scenario cooperation overseas, and transform manufacturing upgrading demand in Southeast Asia, Europe, and other regions into landing opportunities for China's aggregated intelligence solutions.
Sixth, prevent emerging fields from repeating disorderly competition. Clearly distinguish planned capacity, actual output, commercial delivery, and sustained operation, guide local governments and capital to arrange investment based on application effectiveness, and maintain fair competition.
Overall, aggregated intelligence is an important opportunity for China to transform its automotive industry advantages into new industrial advantages. The supply chain, talent, scenarios, and quality systems accumulated over more than two decades in the automotive industry are the most realistic foundation for developing the embodied intelligence industry; the embodied intelligence industry, in turn, opens new demand space for automobiles. By deepening technologies that can be shared, solidifying capabilities that can be transferred, and running through scenarios with real demand, and by adhering to integrated development, scenario-driven approaches, and the "born global" path, China is well-positioned to form a globally competitive aggregated intelligence industry cluster in the next five to ten years.
Source: China Industry News
