The fluid twirl of a handkerchief, precise dance moves synced to rhythm – this wasn’t a human performer captivating audiences during China’s recent Spring Festival gala, but an intelligent robot from Unitree. Its remarkably human-like dexterity offered a tangible glimpse into the burgeoning world of embodied intelligence, a field thrust into the spotlight by its inaugural mention in China’s 2025 Government Work Report. The report explicitly called for cultivating future industries, including “embodied intelligence,” alongside biomanufacturing and quantum tech. This policy endorsement has ignited intense discussion among policymakers and industry leaders, centering on one pivotal question: How soon can humanoid intelligent robots transition from high-tech showcases to affordable, practical assistants in everyday homes?
The momentum is undeniable. Beyond the stage, intelligent robots are increasingly demonstrating real-world utility. During China’s recent legislative sessions, a humanoid intelligent robot named “Kuafu,” functioning as a reporter, navigated Beijing streets, conducting interviews using fluent Mandarin and integrated cameras, much to the fascination of onlookers. Across the nation, these machines are honing their skills: mastering household chores like dish clearing and laundry folding in Shanghai data centers, performing assembly tasks in automotive plants for XPeng, and undergoing complex multi-robot, multi-scenario training in smart factories by UBTECH. The intelligent robot is rapidly moving beyond novelty into functional application.
This surge, as highlighted by He Han, a member of China’s top political advisory body and CEO of Tianyu Digital, stems from a fundamental technological leap. “The focus has shifted,” he observes. “Previously, discussions centered on hardware. Now, breakthroughs in the ‘brain’ – advanced AI cognition – and the ‘cerebellum’ – real-time motion control – represent an evolutionary jump from ‘mechanical shells’ towards ‘digital life.'” This convergence dramatically lowers the usability barrier and explosively expands potential applications. China’s Ministry of Industry and Information Technology (MIIT) underscored this trajectory in its 2023 guidance, targeting breakthroughs in these very “brain, cerebellum, limb” technologies by 2025.
XPeng Chairman He Xiaopeng, also a national legislator, attributes the current fervor to tangible progress: relentless R&D advances, climbing intelligence levels, and clearer validation across diverse scenarios. “The industry trajectory suggests that by 2026, intelligent robots possessing basic L3 autonomous capabilities will enter a phase of moderate-scale commercial mass production,” He projected. He envisions humanoid intelligent robots evolving, over the next 5 to 20 years, into a sector with economic potential rivaling that of new energy vehicles.
Yet, the vision of a helpful, affordable intelligent robot seamlessly integrated into domestic life – tidying homes, assisting the elderly – faces significant hurdles before becoming commonplace. Representatives and advisors pinpoint price, technology maturity, and platform deficiencies as the primary gates to cross.
Zhang Yunquan, a researcher at the Institute of Computing Technology, Chinese Academy of Sciences, and a political advisor, illustrates the cost challenge using eldercare robots. “The high demands for personalized, safe, and convenient service necessitate substantial R&D investment,” he explains. “This translates into high prices, placing these intelligent robots beyond the reach of average families. Consequently, the market feedback loop essential for driving deeper technological refinement weakens.” Achieving “cheap and useful” remains elusive.
The technological hurdles are equally formidable. Core components – the high-precision actuators, sensors, and control systems constituting the “limbs” and nervous system of the intelligent robot – are critical. Nie Pengju, Chairman of Hunan Corun Motor Group and a national legislator, stresses this bottleneck: “Core components are the foundational bedrock and crucial breakthrough point for accelerating new productive forces in the humanoid intelligent robot domain. Currently, domestic development of these high-performance, high-precision parts lags. The scarcity of capable manufacturers severely hampers mass production capabilities, impeding the industry’s path to scale.”
He Han further identifies a critical lack of shared infrastructure. The absence of universal platforms for algorithm development, standardized 3D datasets, certification, scenario testing, and talent cultivation forces companies into siloed, redundant development from the ground up. “This redundancy isn’t just costly; it drastically reduces efficiency,” he states. The need for shared testing environments, analogous to open autonomous driving zones, is acute, currently constraining real-world application validation for the intelligent robot.
Beyond cost and hardware, softer challenges persist. Zhang Yunquan raises crucial questions about user experience: “How can these intelligent robots become more ‘human,’ offering genuine emotional companionship? And critically, how can they serve people effectively while rigorously guaranteeing personal information security and privacy?” These are essential societal and technical questions the industry must resolve.
Facing this vast potential market, delegates and advisors are proposing concrete pathways to accelerate the arrival of affordable, capable intelligent robots in homes.
He Han champions collaboration to avoid wasteful duplication. “We must stop ‘reinventing the wheel’,” he urges. “Encouraging the development of universal platforms – encompassing hardware, software, foundational AI models, and 3D datasets – is vital. This fosters synergistic innovation and lowers R&D barriers across the intelligent robot ecosystem.”
Zhang Yunquan advocates for a dual-track technological push. On the hardware front, he calls for intensified R&D into intelligent “memory-compute integrated” chips and specialized robot operating systems. Algorithmically, the focus should be on large-model-enhanced human-robot interaction, advancing natural language processing and AI-driven emotional recognition capabilities, all built upon uncompromising data security frameworks for the intelligent robot.
Policy support is universally recognized as a vital catalyst. Nie Pengju proposes targeted government action: bolstering domestic core component manufacturers through R&D subsidies and technical trial insurance. He also suggests establishing specialized industrial clusters for these components, leveraging agglomeration effects to drive breakthroughs in intelligent robot technology. Echoing successful precedents, He Xiaopeng recommends applying lessons from the early days of new energy vehicle policy: “Targeted support can accelerate market cultivation and unlock latent demand for the intelligent robot.”
The journey from the dazzling stage of the Spring Festival gala to the practical reality of household assistance is complex. While the technological evolution embodied by the intelligent robot is accelerating, propelled by policy recognition and intense R&D, the path to ubiquity hinges on conquering cost barriers through scalable manufacturing and component innovation, establishing collaborative ecosystems to avoid fragmented development, rigorously addressing safety and ethical concerns, and implementing smart policy frameworks. The question is no longer if humanoid intelligent robots will enter our daily lives, but how swiftly we can overcome these remaining challenges to realize their full potential as accessible, helpful companions in the home. The race to make the intelligent robot a household staple is well and truly underway.