New category creation through cross-industry convergence More transformative value chain convergences occur when entirely new product categories are made by combining technologies in ways that redefine market boundaries and address previously unsolvable challenges. CASE STUDY 2 Humanoid robots in China Recent breakthroughs in vision-language-action models (VLA), spatial intelligence algorithms and cross-domain systems integration have unlocked new possibilities for humanoid robotics. VLA models and AI agents now enable robots to interpret multimodal instructions, learn from diverse inputs and make autonomous decisions in complex settings. Meanwhile, innovations in spatial intelligence such as Gaussian splatting and spatial reasoning allow machines to perceive and navigate human environments with greater precision and adaptability. Finally, technologies from adjacent industries including mobility control, sensor fusion, battery systems and user interface design are being recombined to support the physical form and real-world functionality of humanoid robots. In response, Chinese firms from sectors such as consumer electronics, drones and autonomous vehicles are repurposing their existing capabilities to build humanoid platforms that can operate in unstructured, everyday environments like homes, hospitals and warehouses. Transition of core competencies to humanoids for Chinese firms TABLE 1 Company Core capabilities Humanoid-enabling capabilities Xiaomi Consumer electronics miniaturization, battery systems, user interface (UI)/user interface (UX) designPhysical design and integration of humanoid hardware with intuitive human-machine interfaces, connected to home ecosystems DJI3 Drone flight control, sensor fusion, autonomous navigationNavigation and real-time environmental awareness for mobile robots and autonomous systems Li Auto4 Vehicle autonomy systems, Mind GPT AI models Multimodal reasoning, command interpretation and autonomy in dynamic physical settings Baidu and UBTech5 AI expertise, Apollo autonomous driving platform Human command interpretation and complex environment navigation for general-purpose humanoids In China, this cross-industry convergence is driving remarkable economic efficiencies. Today, building an identical robotic arm in the US (e.g. Universal Robots’ UR5e) costs approximately 2.2 times more than in China,6 largely due to China’s closely net integrated manufacturing ecosystem. As the world moves towards increased autonomation where robots manufacture other robots,7 this cost advantage, combined with cross-industry capability sharing, is accelerating the humanoid market development – with projections showing humanoid robot shipments growing from 18,000 units in 2025 to over 1 million by 2030.8 For organizations navigating convergence opportunities, several critical questions must guide their strategic positioning thinking:1 Value chain leverage: Which positions provide maximum leverage over emerging value chains while building on existing organizational strengths? 2 Integration economics: Where does technology integration create sufficient value to justify premium pricing and market development investments? 3 Capability requirements: What new skills, partnerships and organizational structures are needed to execute effectively in converged value chains? 4 Market timing: When will technological combination solutions reach sufficient scale to deliver significant returns on strategic investments and justify the effort to migrate into new or adjacent value chains? Technology Convergence Report 10

Technology Convergence Report 2025