作者 | 第一财经 祝嫣然 继大模型后，具身智能成为今年科技界的新热点，被认为是新一波人工智能浪潮中的重点方向。 国家发展改革委政策研究室副主任李超27日在发布会上表示，"速度"与"泡沫"一直是前沿产业发展 过程中需要把握和平衡的问题，这对于具身智能产业来讲，也是一样的。 近年来，在创新引领和需求释放的双重作用下，以人形机器人为代表的具身智能产业规模，正在以超 50%的增速跨越式发展。根据市场调研机构预测，2030年将达到千亿元市场规模。 李超表示，当前人形机器人在技术路线、商业化模式、应用场景等方面尚未完全成熟，随着新兴资本 加速入场，我国目前已有超过150家人形机器人企业，这个数量还在不断增加，其中半数以上为初创 或"跨行"入局，这对鼓励创新来讲是一件好事；但也要着力防范重复度高的产品"扎堆"上市、研发 空间被压缩等风险。 面对机遇与挑战并存的局面，关键在于合理引导。目前，国家发展改革委已会同有关部门进行深入研 究，围绕强化政策牵引、加快技术攻关、推动成果应用等开展系统谋划。 2025.11. 27 本文字数：2393，阅读时长大约4分钟 三是推动训练与中试平台等基础设施建设，促进全国范围内具身智能技术、 ...

Group 1 - The China International Import Expo (CIIE) serves as a platform for showcasing global products and insights into industry transformation trends, highlighting the integration of advanced technologies with consumer lifestyles [1][2] - The future mobility exhibition features the M1 passenger drone, which can carry five people and is equipped with an intelligent sensing system, indicating the potential for low-altitude transportation in China [1] - The smart home sector is experiencing significant growth, with companies like Samsung and Panasonic showcasing innovative products that cater to the increasing demand for quality living among Chinese consumers [1] Group 2 - The "AI + Future Factory" exhibition by Mitsubishi Electric demonstrates the application of AI in manufacturing, showcasing intelligent operations that can accurately detect defects and make autonomous decisions [1] - The CIIE highlights the synergy between global cutting-edge technologies and China's vast market, complete industrial system, and improving innovation environment, aiming to foster new growth points and shape new industrial forms [2]

Group 1 - The China International Import Expo (CIIE) serves as a platform for showcasing global products and insights into industry transformation trends, highlighting the integration of advanced technologies with consumer lifestyles [1][2] - The future mobility exhibition features the M1 passenger drone, which can carry five people and is equipped with advanced sensing systems, indicating the potential for low-altitude transportation in China [1] - The smart home sector is experiencing significant growth, with companies like Samsung and Panasonic aiming to develop products that cater to the increasing quality of life demands from Chinese consumers [1] Group 2 - The "AI + Future Factory" exhibition showcases Mitsubishi Electric's intelligent systems capable of precise defect detection and autonomous decision-making, reflecting advancements in manufacturing technology [1] - The CIIE highlights the synergy between global innovative concepts and China's vast market, complete industrial system, and improving innovation environment, aiming to foster new growth points and shape new industry forms [2]

Core Insights - The smart connected vehicle industry is undergoing a significant transformation, reshaping product forms, industrial patterns, infrastructure, and mobility methods [1] - The integration of artificial intelligence in the automotive sector is becoming a core variable for enterprise survival and development, with a focus on leveraging current advantages for future leadership [2] - Despite progress, the domestic smart connected vehicle industry faces challenges such as insufficient policy coordination, technological silos, and supply chain risks [3] Industry Development - In the first half of 2025, China's new energy vehicle sales reached 6.937 million units, a year-on-year increase of 40.3%, with a market penetration rate of 44.3% [1] - The market share of new passenger cars equipped with L2-level assisted driving functions has exceeded 62%, indicating a leading level of intelligence globally [1] - The automotive industry's transformation towards electrification, intelligence, and low carbon is driving profound changes in the supply chain and industrial ecosystem [1] Technological Trends - The development path of automotive intelligence is being reshaped, with end-to-end large model-based autonomous driving technology becoming a focal point for industry layout [1] - The concept of "embodied intelligence" is emerging, where vehicles are redefined as integrated systems combining a "super brain" and "physical body" [2] - The transition from decentralized to centralized control in vehicle systems is essential for the development of L3 and above autonomous driving [2] Challenges and Recommendations - The smart connected vehicle sector faces challenges including standardization, technological maturity, cost control, and safety risks [2][3] - Recommendations include accelerating the definition of L3-level responsibility boundaries, unifying standards for vehicle-road-cloud integration, and establishing an industry-level autonomous driving data pool [3] - Emphasizing self-research and open collaboration is crucial for industry growth, as no single company can dominate the multi-faceted fields involved in smart connected vehicles [3]

Core Viewpoint - The establishment of the "Artificial Intelligence Embodied Intelligent Agent Application Framework and Interface Specification" standard aims to address the challenges faced by the embodied intelligence industry, including high adaptation costs, poor capability reuse, and high development thresholds due to the lack of unified standards [1][2]. Group 1: Industry Challenges - The embodied intelligence sector is experiencing rapid development but is hindered by issues such as platform heterogeneity, interface fragmentation, and closed functional modules [1]. - The absence of a unified standard leads to three prominent problems: high adaptation costs, poor capability reuse, and high development thresholds [1]. Group 2: Standard Development - The standard is jointly drafted by Beijing Human-Robot Innovation Center, China Electronic Technology Standardization Institute, and other industry stakeholders, aiming to provide a framework and interfaces applicable to humanoid robots, intelligent service robots, and industrial collaborative intelligent agents [2]. - The standard systematically constructs a layered application framework and standardized interface system for embodied intelligent agents, particularly for applications requiring cross-entity access and capability reuse [2][3]. Group 3: Technical Specifications - The standard innovatively defines a modular framework structure consisting of "basic model layer - basic module layer - open function layer," clarifying the functional boundaries and relationships of each layer [3]. - It specifies four core interfaces: control interface, feedback interface, management interface, and application development interface, promoting interoperability and capability reuse among different vendors' AI capabilities and robotic hardware platforms [3][5]. Group 4: Impact on the Industry - The formulation of this standard is a significant achievement in collaborative innovation within the embodied intelligence industry, enhancing system universality and scalability while lowering the integration costs [5]. - It aims to provide stable interfaces for upper-level task development and behavior orchestration, significantly improving development efficiency and supporting the national artificial intelligence standard system [5].