US, China split on humanoid robots: AI smarts vs. supply chain control

The global humanoid robot market is entering a new phase as companies move from technology validation—proving robots can work—to commercial trials and field deployment. According to TrendForce's analysis published July 20, 2026, the competitive battle is no longer primarily about hardware specifications but about two complementary questions: which companies can build sustainable business models, and which ones will control the supply chain.

The US strategy hinges on AI ecosystem dominance. NVIDIA is advancing Physical AI platforms including Cosmos (a world model), Isaac Lab (a simulation environment), and GR00T (an embodied AI model). Google DeepMind has launched Gemini Robotics to strengthen robotic understanding of complex environments. OpenAI is participating in embodied AI model development through investment and direct research. Together, these firms are creating a complete vertical stack that spans large language models, robotics platforms, simulation environments, and AI chips. The underlying belief is that humanoid robots moving into complex, unstructured environments will succeed based on their ability to understand context, plan multi-step tasks across scenarios, and learn continuously from real-world data. This reflects a shift in competitive focus: from demonstrating that hardware works to building AI that generalizes. Tesla's Optimus Gen 3 exemplifies this approach, undergoing real-world factory validation with the goal of improving generalization so the same AI can quickly adapt to different work environments. Figure AI is deepening partnerships with large enterprises to train its models on large volumes of real factory data. Boston Dynamics combines decades of motion-control expertise with AI to strengthen autonomous operation in complex settings. Apptrovik generates high-quality training data through extensive real-world deployment, feeding that data back into the Gemini Robotics training pipeline to continuously refine models.

China is pursuing a markedly different path, one that mirrors the electric vehicle playbook: building a complete, integrated supply chain and rapidly lowering costs, then achieving economies of scale through high shipment volumes. This strategy prioritizes the localization of core components—servo motors, reducers, and lithium batteries—to cut costs and compress development cycles. By controlling the supply chain end-to-end and shifting manufacturing risk to scale, China aims to deploy humanoid robots at significantly lower unit cost and in far greater numbers than competitors pursuing a premium AI-first strategy. The divergence reflects fundamentally different bets on where competitive advantage will ultimately reside: in the sophistication of the AI (US) or in the cost and reach of deployment (China).

The humanoid robot industry has reached an inflection point where the focus has shifted from proving technology works to proving it can work at scale and profitably. The US-China divergence reflects deep structural differences in how each region believes competitive advantage will accrue. The US strategy centers on what it calls the "AI ecosystem"—a tightly integrated stack spanning large language models, robotics-specific platforms (Cosmos, Isaac Lab, GR00T, Gemini Robotics), simulation environments, and AI chips. This ecosystem approach assumes that as robots enter complex, unstructured environments, their value will increasingly depend on their ability to understand context, plan multi-step tasks, generalize across scenarios, and learn continuously from data. Tesla's Optimus Gen 3, Figure AI's factory partnerships, Boston Dynamics' motion-control expertise, and Apptrovik's real-world data collection all point to a shared underlying model: build a data flywheel where real-world experience continuously refines AI models.

China's approach mirrors the playbook it used to dominate electric vehicles: rapid supply-chain integration, cost reduction, and scale deployment. By localizing core components (servo motors, reducers, lithium batteries), China aims to shorten development cycles, compress costs, and achieve economies of scale through high shipment volumes. This strategy trades some near-term AI sophistication for the ability to deploy robots at much lower unit cost and in far greater numbers. For enterprises and governments watching this space, the implication is that the future market may bifurcate—with the US producing higher-capability but more expensive robots tailored to complex, variable tasks, and China producing cost-optimized robots for standardized, high-volume deployment scenarios.

No comments yet. Be the first to share your thoughts!