In the long history of technological change, many great leaps look like moves across industries, while sharing the same underlying DNA.
A profound shift is unfolding in technology: automakers and leading component suppliers with years of experience in intelligent electric vehicles are entering embodied-AI robotics systematically and at scale.
At intelligent-vehicle exhibitions, companies such as XPeng, GAC, BYD, and Xiaomi have given center stage to humanoid robots developed internally or with partners, and have even placed them on vehicle production lines for practical training. This convergence illustrates a compelling argument: moving from advanced intelligent vehicles into robotics can be more than an aggressive expansion; it can be a highly coherent evolution in the age of Physical AI.
1. Different forms, shared DNA: the engineering convergence of cars and robots
Strip away their physical appearance and examine them as systems, and intelligent EVs and robots look like different expressions of the same species:
An intelligent EV is a two-ton, four-wheeled heavy mobile robot capable of traveling at more than 100 kilometers per hour.
A humanoid robot is a lightweight mobile platform weighing tens of kilograms, with one or two legs, moving at roughly five kilometers per hour.
The engineering challenges intelligent vehicles have tackled over the past decade provide fertile ground for taking robots from laboratory showpieces to large-scale commercialization:
Computing—the “brain”: Handling urban traffic at more than 100 kilometers per hour has driven the development of powerful automotive computing platforms and end-to-end neural networks, such as systems that convert images directly into control commands. Transferring a driving brain that handles rapidly changing traffic to a robot moving at five kilometers per hour promises a powerful advantage in response speed and redundant safety.
Perception—the “senses”: Bird’s-eye-view representations, occupancy networks, and mass-produced high-resolution CMOS cameras all aim to digitize the complex physical world in three dimensions. Reusing these visual approaches could give robots entering factories or homes advanced spatial perception and three-dimensional obstacle avoidance from the outset.
Manufacturing spillovers—the “heart and skin”: Lightweight battery packs and rigorous battery-management systems developed for vehicles address robots’ need for long endurance, low weight, and thermal safety. The automotive industry’s mature high-strength aluminum die-casting and tooling processes also offer a route from expensive 3D-printed robot skeletons to industrial production cycles measured in minutes.
2. The supply chain as an incubator: reusing more than 50% of resources
Pure-play robotics companies can struggle to commercialize because they lack scale. Annual demand for only a few thousand actuator joints cannot readily support chip tape-out costs or precision-machine depreciation.
Intelligent vehicles, by contrast, serve a consumer market producing tens of millions of units annually. The proposition here is that cars and humanoid robots share more than 50% of supply-chain resources, allowing automakers’ purchasing power and process expertise to push robot hardware costs down by as much as 90%.
The following table compares the core supply-chain capabilities that intelligent EVs and robots can share:
Shared Supply-Chain Capabilities: Intelligent EVs and Embodied-AI Robots
| Core system | Capabilities built for intelligent EVs | Applications in robotics | Shared capabilities and cost benefits | Representative supplier categories |
|---|---|---|---|---|
| Main control—the brain | High-performance automotive driving-compute platforms and access to advanced semiconductor manufacturing, including high-compute autonomous-driving chips. | Robot control boxes and high-performance automotive-grade AI computing modules, such as AI Cube platforms. | Spreading R&D costs: automotive AI chips can ship in the millions, distributing high tape-out costs and making ready-made, reliable control modules available to robotics companies. | Automotive electronics groups such as Desay SV |
| Visual perception—the senses | High-resolution CMOS cameras, automotive LiDAR, depth cameras, and fisheye cameras. | Stereo vision in robot heads and surrounding environmental-sensing arrays. | Large cost reductions: adoption in intelligent EVs has pushed automotive LiDAR and camera costs from tens of thousands of dollars toward hundreds, allowing robots to benefit from mature, inexpensive sensor supply chains. | Automotive optics suppliers such as Hesai and Sunny Optical |
| Energy management—the heart | High-density, fast-charging cells, including large cylindrical and lithium-iron-phosphate cells; cell-to-chassis integration; automotive battery-management systems. | Compact, lightweight battery packs embedded in robot torsos, with high power density and roughly 2–3 kWh of capacity. | Safety and endurance: battery supply chains tested against demanding automotive collision, water-ingress, and fire-prevention requirements provide a route away from tethered power toward independent operation. | Battery groups such as CATL and BYD |
| Physical actuation—limbs and joints | Thermal-management assemblies, chassis suspension, power-steering motors, and precision steering shafts. | Rotary and linear actuator assemblies, frameless torque motors, and planetary roller screws. | Strong overlap in processes and customers: frameless motors and screws share core materials—rare-earth permanent magnets and copper wire—and precision-manufacturing processes with automotive propulsion and thermal-management systems. Automotive Tier 1 suppliers can adapt existing production lines. | Automotive Tier 1 suppliers such as Tuopu, Sanhua, and Xinquan |
| Precision transmission—the musculoskeletal system | Automotive gearboxes, constant-velocity joints, and high-strength aluminum body structures. | Planetary, harmonic, and RV reducers, plus topology-optimized, lightweight robot frames. | From individual machining to production-line casting: automotive suppliers can use existing thousand-ton-class integrated die-casting machines and precision gear-grinding equipment to manufacture robot structures and reducers in volume, drawing on established yield-management expertise. | Precision manufacturers such as Leaderdrive, Shuanghuan, and GSP |
3. Capital as a signal: the leap from EVs to robotics
In business, technology is the skeleton, the supply chain is the muscle, and capital—the real money committed by major companies—is the blood.
China’s private markets, local government guidance funds, and automaker financial reports point to a concentration of investment: capital that once poured into new-energy vehicles, batteries, and autonomous-driving algorithms is moving systematically into embodied-AI robotics.
This is the industrial-capital transition taking shape in China:
1. Industrial giants: building cars with one hand, robots with the other
China’s leading automakers are becoming major financial backers of robotics.
Xiaomi established Beijing Xiaomi Robotics Technology Co., Ltd., making embodied intelligence a major bet in intelligent manufacturing.
XPeng has not only introduced its own humanoid robot but also anchored its long-term strategy in becoming an embodied-intelligence ecosystem company.
BYD has invested in prominent humanoid-robotics startups such as AgiBot and is also pushing robot deployment on its highly automated vehicle lines in Xi’an and Shenzhen.
These groups have more than capital: they understand hardware production at scale, supply-chain management, and real applications. Compared with conventional venture investment, their commitment can serve the completion of their own industrial systems rather than a purely financial exit.
2. A second act for local government guidance funds
Over the past decade, local governments in places such as Hefei, Shenzhen, and Changzhou used investments in NIO, BYD, and CATL to develop globally recognized EV and new-energy clusters. These experienced state-backed industrial investors are now seeking to repeat that success.
Beijing, Shanghai, and Shenzhen have announced robotics guidance funds on the scale of tens of billions of yuan.
Their approach extends beyond financing. Robotics parks are planned beside EV clusters in the Yangtze and Pearl River deltas, linking the two supply chains and encouraging automotive Tier 1 suppliers to conduct joint R&D with early-stage robotics companies.
3. A second venture for the winners of the EV era
Leading venture investors, industrial funds, and executives who prospered in EVs see robotics as a potential ticket into the next trillion-scale industry. Their thesis is that EV competition is maturing and margins are thinning, while robotics is approaching a breakout and can inherit 100% of the infrastructure built through a decade of EV investment.
In this view, investing in robotics means buying the technological dividends accumulated by the EV industry at a discount.
4. The longer-term evolution: from data incubation to industrial feedback
In this era shaped jointly by technology, supply chains, and capital, companies with intelligent-vehicle roots can evolve at remarkable speed.
The advantage could become a reinforcing cycle of data and applications:
[Capital from automakers and local funds] → [Faster transfer of EV technologies into robotics] → [Robots working on vehicle production lines] → [Higher efficiency and lower costs]
Data incubation: intelligent vehicles operate in the real world every day, collecting long-tail scenarios involving gravity, friction, physical collisions, and unexpected obstacles through shadow-mode systems. The thesis is that Physical AI trained on such data gains practical knowledge of the world, creating a barrier difficult for laboratory-only robotics companies to match.
Industrial feedback: as robots with automotive roots mature, their first proving grounds can be the automakers’ own intelligent factories. On production lines, they can perform demanding tasks such as precisely aligning self-tapping nuts, inspecting vehicle bodies, and hazardous assembly, further increasing automation within the automotive industry itself.
Conclusion
In the longer-term development of Physical AI, the strongest cross-industry entrants are not merely chasing a fashionable theme. They recognize the convergence of technology and capital.
A brand with deep intelligent-vehicle expertise can bring its assets in driving computation, visual perception, advanced electric powertrains, and global manufacturing supply chains into robotics. That transfer has a strong engineering and investment rationale.
Building vehicles is a starting point for accumulating real-world AI capabilities. The movement of capital signals industrial conviction. Robotics extends those capabilities outward, toward the broader physical world.