近两年，具身智能赛道群雄逐鹿，在部分领域已走向规模化应用。 安永大中华区业务主管合伙人毕舜杰表示："创新是企业成长潜力的核心源泉，而资本赋能则是实现高 增长的关键支撑。作为人工智能与现实世界交互的重要载体，具身智能正推动智能技术从'感知'走向'行 动'，越来越多的创新企业积极布局、加速进场。" 据安永对该行业的观察，具身智能的应用场景涵盖工业制造、家庭服务、危险作业等多个领域，中国企 业在这些场景的商业化落地方面表现突出。 何兆烽认为，具身智能行业当前面临的最大瓶颈是商业化落地的规模化和可持续性。"尽管技术已在实 验室或小范围内取得突破，但如何将这些技术高效、低成本地应用于实际场景并实现盈利，仍是关键问 题。"为此，安永方面建议行业企业聚焦垂直场景、降低落地难度，同时优化核心组件采购、提升设备 集成度、降低单位成本；引入开放平台和模块化设计以提升扩展性。 安永大中华区上市服务主管合伙人何兆烽举例道，在14家具身智能企业中，包含了主攻家庭服务、教育 场景的通用机器人及制造业协作机器人的企业，突破关节模组、多模态传感器等核心部件的企业；此 外，人工智能领域推动空间智能、AI蛋白质整体解决方案落地，生命科学领域以可 ...

9月24日，在高通2025年度骁龙峰会上，宇树科技创始人兼首席执行官王兴兴与高通公司研发副总裁兼全球AI研发负 责人侯纪磊展开现场对谈，就通用机器人的发展阶段、技术痛点、协作方向等议题，王兴兴发表了自己的观点。 图片来源：高通 但固定动作仅为起点，王兴兴表示下一阶段的核心突破方向是"根据人类指令实时完成任意动作"。宇树科技正全力推 进这一技术，若进展顺利，今年年底即可实现；即便稍慢，预计明年上半年也能让智能机器人响应任意指令、生成对 应动作。 在任意动作实时生成的基础上，王兴兴认为机器人还需跨越"与物理环境任意交互"的门槛，并预测到明年下半年有望 实现"在未知场景下自主干活"：比如在未提前规划的会议室里，听到"拿水"指令，机器人就能自主找水；收到"整理桌 子"需求就规划操作流程，真正接近"趋生智能"的关键节点。 实用化还需攻克"高成功率"难关。王兴兴强调，80%-90%的成功率是不够的，通用机器人需达到99.9%的成功率，而 且要具备精细操作能力，比如完成手机拆解这类复杂任务，才能真正满足实际需求。 02 线缆问题成机器人可靠性"绊脚石" 在机器人技术落地过程中，通信领域的线缆问题被王兴兴视为"亟待突破的关键 ...

张鹏科技商业观察 . 聊科技，谈商业。 以下文章来源于张鹏科技商业观察 ，作者张鹏 上个月又飞了一趟硅谷，与具身智能领域的科学家和创业者们进行了一些交流。 总结起来一个核心的体感是： 具身智能这个宏大的故事，还需要我们 有个「 五年耐心 」。 这 个判断，源于对它当下所处阶段、核心瓶颈以及未来演进路径的拆解。 一些核心观察： 超 13000 人的「AI 产品市集」社群！不错过每一款有价值的 AI 应用。 邀请从业者、开发人员和创业者，飞书扫码加群： 进群后，你有机会得到： 拿一个尚不成熟的通用机器人，硬塞进一个以精准和效率为核心的工业产线里， 这件事现 在的挑战其实非常大 。 现在通用机器人，本质上是用「通用性」在交换「精准性」和「效率」。把 远未成熟的人 形机器人塞进产线，把「通用性」用到最需要高精准、高效率的场景中，多少是有些错位 的。 一个合理预期是， 一到两年内， 具身智能 有望迎来它的「GPT-3.0 时刻」，即 在实验室 环境下，圈内人将看到机器人的通用模型（大脑+小脑）明显的技术突破，并对主流技术路 线达成共识。 用于机器人训练的「现实数据」生产，存在三个局限： 一是规模上不去 ；二是 成本下不 ...

Core Viewpoint - The company, Geekplus (02590), has reported impressive financial results in its first earnings forecast since going public, indicating a significant improvement in operational efficiency and a move towards profitability [1][4]. Financial Performance - For the first half of 2025, the company expects revenue between 995-1,030 million yuan, representing a year-on-year growth of 27-32% [1]. - The net loss for the reporting period is projected to be between 45-55 million yuan, a substantial reduction of 90-92% compared to the previous year [1]. - Adjusted net loss is expected to be between 10-20 million yuan, also significantly narrowed by 90-95% [1]. Business Expansion - The company has established a wholly-owned subsidiary, Beijing Geekplus Embodied Intelligence Technology Co., Ltd., to enter the embodied intelligence sector, focusing on mechanical picking and general robotics [2]. - This new business aims to leverage the company's existing AMR product line and strong brand influence to accelerate the commercialization of embodied intelligence technologies [2]. Market Position and Strategy - Geekplus is positioned to capitalize on the trillion-dollar robotics market, with a development path likely starting from warehousing and expanding into smart manufacturing and commercial services [2]. - The company has a global presence, covering over 40 countries and serving more than 800 end customers, with a customer repurchase rate of 74.6% in 2024, significantly above the industry average [2]. Technological Advancements - The company has a strong focus on R&D, with 1,867 patents and a research team comprising 41% of its workforce, making it one of the largest R&D teams in the industry [3]. - The technological barriers and commercial capabilities of Geekplus position it favorably for future growth and valuation increases [5]. Industry Trends - Recent government policies, such as the "Artificial Intelligence +" initiative, are expected to drive the commercialization of AI applications across various sectors, providing strategic opportunities for Geekplus [4]. - The company is seen as a leader in the robotics industry, poised to advance technology towards more intelligent and versatile applications [4].

Core Viewpoint - The article presents a clear framework for classifying the development of robotic technology into five levels, emphasizing that general intelligent robots will evolve gradually, similar to autonomous driving technology. The two key capabilities determining the level of robotic development are Agency and Dexterity, which together dictate the commercial value that robots can create [1][5]. Group 1: Evolution of Robotics - The framework outlines five evolutionary stages of the robotics industry, transitioning from simple automation tools to general intelligent agents capable of performing diverse tasks in complex environments [2][3]. - The report highlights that modern AI paradigms are transforming many limitations of robots into data challenges, enabling robots to acquire new skills and accelerate deployment in the labor market [5][11]. Group 2: Levels of Autonomy - Level 0 (Scripted Motion): Traditional industrial robots that rely entirely on pre-set programs, operating in highly structured environments without any autonomy [6][12]. - Level 1 (Intelligent Pick and Place): Robots gain basic computer vision capabilities to identify and grasp items from cluttered environments, primarily used in logistics since around 2015 [6][14]. - Level 2 (Autonomous Mobility): Robots achieve significant advancements in agency, allowing them to navigate and plan actions in open, dynamic environments [6][18]. - Level 3 (Low-Skill Manipulation): Robots combine mobility with advanced manipulation capabilities, enabling them to perform multi-step tasks with lower precision requirements [6][20]. - Level 4 (Force-Dependent Tasks): The ultimate form of robotic development, where robots can perform complex tasks requiring fine motor skills and tactile feedback, currently in research [6][24]. Group 3: Challenges and Considerations - Each level of autonomy is defined by its unlocking capabilities, with a focus on commercial viability rather than just technical feasibility. Reliability and performance are crucial for creating value through action [8][9]. - The challenges faced by Level 0 robots include high capital costs, inflexibility, and the need for constant human oversight, making automation a high-risk engineering project [37][39]. - Level 1 robots encounter difficulties in adapting to novel objects and environments, requiring significant data collection and integration efforts to achieve reliable performance [60][62]. Group 4: Market Implications - The transition to general robots is expected to lead to significant labor market changes, with a gradual increase in capabilities allowing for broader task execution [5][11]. - The potential for automation in low-skill, high-turnover jobs, such as those in warehouses, presents a compelling business case for AI-driven robots, especially in light of rising labor costs and employee turnover [65][66].

Core Viewpoint - The establishment of Beijing Extreme Intelligence Technology Co., Ltd. aims to enhance the development and commercialization of embodied intelligence technology and related products, which will synergize with the company's existing logistics robot business [1] Group 1: Company Developments - Extreme Intelligence announced the completion of the establishment of Beijing Extreme Intelligence Technology Co., Ltd. on July 30, with an initial registered capital of 10 million yuan [1] - The new company will focus on research and development of embodied intelligence technology and related products, including robotic picking and general-purpose robots [1] - The group plans to apply embodied intelligence technology and related products in business-to-business scenarios, particularly in logistics and manufacturing, leveraging its existing brand and business network for rapid commercialization [1] Group 2: Business Strategy - The company is focusing on core business solutions represented by warehouse fulfillment robots to drive growth, profitability, and financial stability [1] - The establishment of the embodied intelligence company is expected to accelerate the research and application of new technologies while effectively collaborating with existing businesses to expand development opportunities [1] - The initiative aims to attract more talent to support the long-term development of the group [1]

Core Viewpoint - The aging population in China, combined with rapid advancements in artificial intelligence and robotics, will lead to a future where robots play a significant role in both the workforce and elder care [2][5][12] Population Dynamics - In 1989, China's birth rate was 24.32 million, while projections for 2024 indicate only 9.54 million newborns, suggesting that by the time the CEO of Mech-Mind reaches 60, the number of young people will be less than half of his generation [4][5] - The significant decrease in the young workforce will necessitate a shift in production and economic growth strategies [5] Robotics in Industry - Mech-Mind provides general-purpose robots equipped with "3D vision + AI brain + dexterous hands," serving various industries including automotive, home appliances, logistics, and food and beverage [5][6] - As the aging population increases, there will be a greater need for robots to fill labor shortages in tedious and physically demanding jobs [5][8] Comparison with Japan - Japan, having faced similar demographic challenges, has integrated robots into its workforce, with a notable example being an Amazon distribution center where robots outnumber human workers [6][7] - In Japan, many elderly individuals continue to work due to insufficient pensions, and companies like Amazon have adapted their systems to accommodate older workers [7] Future of Robotics in Elder Care - The potential for robots to assist in elder care is significant, with advancements allowing for more complex tasks and interactions [10][11] - Current robots can provide basic interaction and companionship, but there is still a long way to go before they can fully assist elderly individuals in their homes [10][12] Market Potential and Challenges - The home robotics market presents a vast opportunity, but challenges such as safety, cost, and skill requirements must be addressed before widespread adoption can occur [12] - The evolution of robots could lead to them taking over more menial tasks, allowing human labor to shift towards elder care services [12][13] Historical Context - The development of robotics has historically raised concerns about job displacement, yet it has also led to increased labor efficiency and the creation of new job categories [13][14]

Core Insights - Beijing JiZhiJia Technology Co., Ltd. (referred to as "JiZhiJia") has established a wholly-owned subsidiary, Beijing JiZhiJia Embodied Intelligence Technology Co., Ltd., to enter the embodied intelligence sector, focusing on mechanical picking and general robotics for B2B applications in logistics and manufacturing [1] - JiZhiJia aims to leverage its existing logistics robotics business and brand network to accelerate the commercialization of embodied intelligence technologies and products [1] Group 1: Commercialization and Global Network - JiZhiJia has strong commercialization capabilities, a comprehensive global business network, and a mature service system, which will support the development and implementation of its embodied intelligence technologies [2] - The company has been the largest AMR warehouse robotics solution provider globally for six consecutive years, with projected revenue of 2.4 billion yuan in 2024, nearing profitability [2] - JiZhiJia's existing capabilities and resources provide mature application scenarios and benchmark validation channels for its embodied intelligence technologies, facilitating global deployment and scaling [2] Group 2: Research and Development Strength - JiZhiJia possesses significant technological strength, with 1,867 patents and a research team comprising 41% of its workforce, making it one of the largest R&D teams in the industry [3] - The team has extensive experience in robot product development, particularly in perception, decision-making, control, and multi-agent collaboration [3] - The global application and replication of JiZhiJia's solutions demonstrate the robustness, reliability, and scalability of its technology in complex scenarios [3] Group 3: Data Assets and Scene Understanding - JiZhiJia has a unique advantage in scene understanding and data accumulation, with vast operational data from logistics scenarios and a deep understanding of customer pain points [4] - This practical scene recognition and data advantage will accelerate technology validation and create a self-reinforcing cycle for training and optimizing embodied intelligence models [4] - JiZhiJia's mature commercialization model, top-tier R&D resources, and substantial data assets position it well to capture opportunities in the embodied intelligence market [4]

Core Viewpoint - The establishment of Beijing Extreme Intelligence Technology Co., Ltd. marks Extreme Intelligence's entry into the embodied intelligence field, focusing on mechanical picking and general robotics for B2B scenarios in logistics and manufacturing, leveraging its existing logistics robot business and brand network for rapid commercialization [2]. Group 1: Commercialization Capability and Global Network - Extreme Intelligence has been the largest global provider of AMR warehouse robot solutions for six consecutive years, with projected revenue of 2.4 billion yuan in 2024, nearing profitability. The company operates in over 40 countries and has more than 800 end customers, including global leaders like Nike, Walmart, UPS, and DHL, providing a robust application environment for its embodied intelligence technology [4][3]. - The established commercial capabilities, global business network, and customer resources, along with a mature service system and supply chain, significantly accelerate market entry and ensure scalability for global deployment [4]. Group 2: R&D Capability - As of December 31, 2024, Extreme Intelligence holds 1,867 patents, with R&D personnel constituting 41% of its workforce, making it one of the largest R&D teams in the industry. The company has extensive experience in robot product development and strong expertise in areas such as perception, decision-making, and multi-agent collaboration [5]. - The global application and replication of its solutions validate the robustness, reliability, and scalability of its technology, showcasing its underlying software, hardware, and algorithm development capabilities [5]. Group 3: Scene Understanding and Data Accumulation - The vast amount of real operational data from logistics scenarios and a deep understanding of logistics operations and customer pain points are crucial for competition in the embodied intelligence field, serving as essential resources for training and optimizing embodied intelligence models [6]. - The practical scene cognition and data advantages of Extreme Intelligence will significantly accelerate technology validation and create a self-reinforcing flywheel for future model training, continuously expanding its competitive edge [6]. - Overall, the combination of commercialization models, R&D resources, and accumulated scene data provides a solid foundation for Extreme Intelligence's entry into the commercial blue ocean of embodied intelligence, with the potential for valuation upside through breakthroughs in this business [6].

Core Viewpoint - The article highlights the successful completion of a multi-million Pre-A+ financing round for Deep Sea Intelligence (Guangzhou) Technology Co., Ltd., which focuses on high-end deep-sea robotics technology [1]. Group 1: Company Overview - Deep Sea Intelligence specializes in the research, manufacturing, and service of high-end deep-sea robots, including work-class ROVs, underwater cable laying robots, and general-purpose robots [4]. - The core team of the company comes from leading global deep-sea equipment firms and has previously led the development of several world-class deep-sea robotics projects [4]. - The company has rapidly expanded from 3 to nearly 100 technical R&D team members within two years and has secured a significant portion of domestic commercial orders for deep-sea robots [4]. Group 2: Market Position and Achievements - Deep Sea Intelligence is the only private enterprise in China that has successfully sold deep-sea robot systems commercially, with clients spanning major sectors such as international oil and gas contractors and wind power contractors [4]. - The company has established a global presence, with customers in over 10 countries, including the UK, Egypt, UAE, Australia, Singapore, Malaysia, Indonesia, and Angola [4]. Group 3: Technological Capabilities - The company emphasizes the importance of self-research and manufacturing for core technologies, achieving self-sufficiency in all core components except standard parts and acoustic sensors [6]. - Approximately 70% of the supply chain is sourced from the Greater Bay Area, enabling rapid and stable supply and after-sales service [6]. - The team possesses a strong understanding of international standards and can effectively communicate complex professional requirements to clients [6].