Core Viewpoint - The integration of medical and engineering fields is becoming a key driver for breakthroughs in high-end medical equipment and biopharmaceutical innovation, particularly in Zhuhai, which is positioned as a strategic hub in the Guangdong-Hong Kong-Macao Greater Bay Area [2]. Event Details - The "Medical-Engineering Integration Annual Conference 2025" will be held from December 26 to 28, 2025, in Zhuhai, Guangdong, aiming to promote high-quality development in the medical-engineering integration sector [2]. - The conference theme is "Jointly Inviting the Bay Area: Medical-Engineering Integration to Inspire the Future," focusing on creating a resource platform for collaboration among government, associations, hospitals, universities, research institutions, investment entities, and enterprises [2]. Organizing Institutions - The conference is guided by the China Medical and Health Culture Association and organized by Beijing Institute of Technology (Zhuhai), Zhuhai Health Bureau, and the Medical-Engineering Integration Branch of the China Medical and Health Culture Association [3]. Conference Highlights - The event will feature top scholars and industry leaders sharing insights on cutting-edge research and industry trends, fostering a collaborative ecosystem for project signing and talent introduction [60]. - A dedicated Zhuhai Medical-Engineering Integration Liaison Office will be established to provide ongoing support and services, including policy consultation and funding connections [60]. Main Content - The agenda includes leadership speeches, high-quality thematic forums, academic dialogues, and an award ceremony for outstanding orthopedic innovation projects [61]. - Resource matching sessions will be organized, including project roadshows and expert evaluations, along with site visits to local leading enterprises [61]. - Policy and service promotion will cover key industry policies in biomedicine, high-end medical equipment, and artificial intelligence [61].

11月19日，PNDbotics正式发布了 旗下 全尺寸人形机器人 Ada m 挑战高难度 "查尔斯顿舞"（Charleston Dance）的 舞蹈 视频。这段 兼 具力量感与高频动态的演示，不仅是 全球 范围内 全尺寸机器人 首次 成功 挑战 这一 复杂步态的 里程碑式 实践 ， 也 向行业集中展示了 PNDbotics 长达 十年的全栈自研 与 技术积淀 实力。 在全行业竞速商业化的浪潮中， PNDbotics 呈现出一套差异化 的研发逻辑： 不急于靠博眼球抢占市场，而 是选择用接近成年人尺寸的机器人进行技术验证与开拓， 系统 检验了 人形机器人在极致动态平衡、瞬时高功 率输出及全身极端工况协同控制上的能力，为后续真实场景应用夯实地基。 ▍ 查尔斯顿舞 背后 ：全尺寸平台的 "极限演习" 对于全尺寸人形机器人而言，查尔斯顿 舞 并非单纯的展示，而是 一次对物理极限的严峻考验 。由于全尺寸 机器人 与成人相近 ， 其 运动具有 大惯量特性 ， 这 对关节峰值扭矩、驱动系统能量密度及结构耐久性 都提 出了 指数级 增长的 要求 。 从技术验证逻辑来看， 查尔斯顿舞蹈的高强度特征，对机器人的动态平衡、核心稳定 ...

以下文章来源于雷峰网 ，作者高景辉 雷峰网 . 别的机器人赛跑都在公路上，这场比赛机器人直接去越野了，而且场地还是在香港中文大学岭南校区的一个小径 上。这里路面狭窄，左右只有一米，而且山间小道，路面湿滑，极影响机器人与地面之间的摩擦力，还有坡度不 定带来的平衡问题，以及树叶遮挡带来的光线变化等等，不管是双足机器人还是四足机器人，如果它们能开口的 话，肯定会直接说，我不走了。 洞见智能未来，共与产业变迁 " 唯有智能，才能让机器人活过来。 " 作者丨高景辉 编辑丨马晓宁 "这也太难了吧。"比赛之前，主办方采访正在备赛的选手，几乎每一个人都叫苦连天。 就连比赛的讲解员、香港中文大学窦琪教授，比赛直播一开场也给大家打了一剂预防针："这场比赛的四个赛题， 瞄准的都是具身智能领域难以攻克的难题。" 难，就难在考题都太 真实 了。 垃圾分拣题目里面，机器人面对的三个物体，香蕉模型的难点是柔软难抓取，塑料瓶的难点是透明不容易识别， 纸盒的难点是不受力易形变，哪一个都不是省油的灯。 至于另一边的浇花题目，不仅要浇在白色的花上才得分，而且机器人的手拿水壶远没有人类那么方便，一开始就 被教授们认为是最难的题目。还有吊桥题目，一不 ...

Core Insights - The article discusses the insights from the sixth China Robot Industry Annual Conference, emphasizing the evolution and challenges of embodied intelligence in the robotics sector [1][2][3]. Group 1: Industry Overview - The global annual installation of industrial robots is approximately 600,000 units, with China accounting for over 50% of this market share, showcasing significant domestic achievements in robot manufacturing [5]. - The number of robotics-related companies in China surged from over 300,000 to nearly 1,000,000 within five years, indicating a booming industry [5][9]. - The concept of "embodied intelligence" has gained traction since 2022, driven by major companies like Tesla and NVIDIA, which has redefined the robotics landscape [5][11]. Group 2: Market Dynamics - Despite the enthusiasm in the market, there are concerns about potential bubbles, as the financing scale has exceeded 10 billion in the past two years, driven by unrealistic growth expectations [6][9]. - The article highlights the dual nature of bubbles, suggesting that while they can fuel innovation, they also pose risks if companies fail to establish viable business models and technological barriers [9][11]. Group 3: Strategic Insights - The article outlines three major challenges for the robotics industry: the lack of closed-loop scenarios, misjudgment of technological thresholds, and severely overstated valuations [3][11]. - It emphasizes the importance of core technologies and the need for companies to possess unique technological advantages to survive in a competitive market [9][11]. Group 4: Pathways to Success - The article presents three strategic pathways for different players in the robotics ecosystem: 1. "Brain-first" approach focusing on comprehensive technological disruption, exemplified by companies like Tesla [12][16]. 2. "Body-first" strategy for established robotics firms to integrate intelligence into existing products [12][16]. 3. "Vertical intelligence" approach for startups to focus on niche markets and specific applications [12][16]. Group 5: Survival Strategies - The article outlines three core survival strategies for startups in the robotics sector: 1. Define specific application scenarios to address practical challenges [19]. 2. Recognize the distinct paths required for industrial versus consumer applications [20]. 3. Balance between national strategic goals and market realities to ensure sustainable growth [21]. Group 6: Future Outlook - The article anticipates that future entrepreneurial and investment opportunities will center around breakthroughs in key technologies and the realization of value in real-world applications [23]. - It highlights the importance of continuous evolution and integration of core modules such as decision-making, control, and perception in robotics [25].

Core Insights - The article highlights the significant progress of Zhuoyu Technology in the core component field of embodied intelligence, marked by a new round of financing exceeding 100 million RMB and a record monthly order amount surpassing 50 million RMB [1][9]. Financing and Investment - Zhuoyu Technology completed a financing round led by Changzhou Guangyang Bearing and Cuiwei Fund, with participation from other institutions, including Suzhou Xingjun and Wuxi High-tech Zone Investment Control Group [3]. - The funds will primarily be allocated to capacity construction and R&D, including the establishment of a highly automated production line for 300,000 joint modules and a production line for 500,000 fourth-generation frameless torque motors annually [3]. Technology and Product Development - Zhuoyu Technology adheres to a "full-stack integration" technology strategy, focusing on core self-research and has developed a comprehensive product matrix covering various core drive components [4]. - The company’s product offerings include frameless torque motors, integrated joint modules, servo drivers, encoders, planetary reducers, and high-end servo control products, with some products achieving low cogging torque fluctuations below 1% [4]. Market Performance - The company achieved a record monthly order amount of over 50 million RMB in November-December 2025, reflecting strong market demand and customer recognition of product stability and reliability [9]. - The growth in orders is attributed to the acceleration of high ROI applications in industrial manufacturing and warehouse automation, indicating a robust business growth model [9]. Regional and R&D Expansion - Zhuoyu Technology has established a dual R&D center strategy with the recent establishment of its Beijing R&D center, complementing the existing Suzhou center [12][14]. - The Beijing center will focus on lightweight robotic arms and dexterous hand hardware applications, while the Suzhou center will enhance core drive platform capabilities [14]. Industry Context and Future Outlook - The article notes that the domestic embodied intelligence industry is experiencing rapid growth, supported by favorable policies and market conditions, with a focus on achieving self-controlled industrial ecosystems [15]. - Zhuoyu Technology is positioned to strengthen its market position in the core component sector through enhanced production capacity, order growth, and innovation from its dual R&D centers [15].

引言 过去一年，具身智能成为机器人产业最热的关键词之一。人形机器人、服务机器人、特种机器人密集发布，但 一个现实问题也越来越清晰地浮出水面： ▍ 从 " 卖机器人 " 到 " 卖大脑 " ：产业逻辑正在发生变化 智能能力深度绑定硬件 算法围绕单一场景优化 产品更新依赖重新设计整套系统 机器人看起来更 " 聪明 " 了，但离真正的自主智能，仍然存在结构性差距。 近日， Robo Summit 机器人发展论坛（第二期）暨首届朝阳智能机器人生态大会在北京朝阳举办。千诀科 技创始人兼 CEO 高海川受邀发表主题演讲，并与银河通用创始人兼 CTO 王鹤、智平方副总裁邱巍等嘉宾共 同启动 " 朝阳区智能机器人生态伙伴建设计划 " ，围绕具身智能的技术演进与产业协同展开深入探讨。 在演讲中，千诀科技创始人兼 CEO 高海川分享了一个颇具产业争议的判断 —— 机器人智能的下一阶段竞争，不在"哪种形态跑得更像人"，而在"机器人大脑是否真正独立、通用、可演 化"。 这一判断，正在重新划分机器人产业的技术路线。 长期以来，机器人产业的主流商业模式是 整机驱动 ： 但随着应用场景快速扩张，这种模式的边界越来越明显： 每进入一个新行业 ...

Core Insights - The article emphasizes that technological breakthroughs and cost optimization of core components are crucial for the industrialization of humanoid robots [1] Group 1: Six-Dimensional Force Sensor - Fourier has announced its first six-dimensional force sensor aimed at humanoid robot applications, expected to be priced at a "thousand-yuan" level, which could alleviate the industry's cost challenges and promote large-scale production of humanoid robots [3][10] - The six-dimensional force sensor acts as the "electronic skin" of humanoid robots, measuring forces and moments in three spatial directions, enhancing system control precision and dynamic response speed [6][7] - Currently, the cost of a single six-dimensional force sensor exceeds 80,000 yuan when four are used per robot, with high-end models reaching several tens of thousands of yuan [8][10] Group 2: Cost Challenges and Market Demand - The high cost of core components directly leads to elevated pricing of end products, hindering the large-scale promotion of humanoid robots across various applications [10] - The introduction of Fourier's six-dimensional force sensor, potentially priced below 5,000 yuan, could significantly lower the industry's entry barriers and address the "high price, low configuration" issue [10][15] Group 3: Technological Development and Future Prospects - Fourier has been involved in the force feedback technology sector for nine years, developing a comprehensive force feedback technology system that adapts to personalized rehabilitation and interaction needs [12] - The new six-dimensional force sensor utilizes optical structural technology, optimizing production processes to achieve significant cost reductions while maintaining industrial-grade precision and durability [14] - The anticipated widespread adoption of the six-dimensional force sensor could lead to a downward trend in pricing across the industry, enabling more companies to engage in technological research and product innovation [15]

Core Insights - The article discusses the integration of Large Language Models (LLMs) with Behavior Trees (BT) to enhance robotic task execution and adaptability in the presence of external disturbances [1][2][12]. Group 1: LLM and BT Integration - LLMs are utilized to interpret user commands into behavior trees that include task goal conditions [2]. - The combination of LLMs and BT allows for fewer calls to LLMs while managing external disturbances through an action database [2][12]. - A human-in-the-loop learning mechanism is proposed to refine the knowledge generated by LLMs, ensuring safety and adaptability in robotic operations [5][7]. Group 2: Human-in-the-Loop Learning Mechanism - The mechanism involves designing a context for LLMs that includes prompt engineering, manipulation primitives (MPs), and an action database [5]. - User interactions guide LLMs to correct and enhance the generated action knowledge, which is then added to the action database after user confirmation [7][12]. - The generated action knowledge consists of preconditions, postconditions, and a set of MPs, implemented in BT format [7]. Group 3: Task Evaluation and Performance - Eight tasks were designed to evaluate the proposed method, categorized into three difficulty levels: Easy, Medium, and Hard [9]. - The proposed method achieved a success rate of over 80% across the tasks, significantly outperforming baseline methods that lacked human interaction [12]. - The adaptability of the generated action knowledge was tested against external disturbances, achieving a success rate exceeding 70% [14]. Group 4: Generalization and Future Improvements - The generated action knowledge demonstrated good generalization capabilities, with success rates over 70% for certain tasks involving new objects [17]. - However, some tasks had success rates below 40% due to the inapplicability of MPs parameters to new objects, indicating a need for fine-tuning before application [17]. - Overall, the proposed human-in-the-loop learning mechanism enhances robotic learning performance, enabling robots to complete tasks and respond to external disturbances effectively [18].

在工业自动化向 "高频切换、连续运转、智能监控" 迈进的当下，传统连接方式却成了产能提升的 "绊脚石"。无论是机械插头这类通用连接部件，还是快换盘专用的针脚接触式电模块，亦或是旋转 传 输 场 景 的 电 滑 环 ， 都 因 物 理 接 触 的 固 有 缺 陷 ， 难 以 适 配 某 些 现 代 化 产 线 需 求 。 而 SMW- Autoblok 凭借深耕自动化连接领域数十年的技术积淀，以电感耦合为核心的无线耦合技术，从根 源 上 打 破 传 统 局 限 ， 为 全 球 车 企 、 3C 巨 头 、 精 密 制 造 企 业 、 测 量 实 验 室 提 供 了 " 无 磨 损 、 高 可 靠、强适配" 的一站式连接解决方案。 一、传统连接方式的共性痛点： 磨损、低效与高维护成本 自动化产线的稳定运转，离不开能量与信号的可靠传输，但传统接触式连接却始终受三大核心问题 困扰： 1.物理磨损引发的 "故障连锁反应" （1）通用连接场景（机械插头）： 频繁插拔导致插头磨损变形，不仅需定期停机更换部件，更易 因接触不良引发信号中断、供电不稳 —— 汽车零部件产线的机器人焊接偏差、3C 组装线的检测 数据失真，多源于此 ...

机器人摔倒是个大难题，尤其是 "头重脚轻"的机器人，一不小心就可能造成昂贵的损伤。过去，为了防止摔 倒，工程师们要么限制其性能，让它畏首畏尾；要么任其"硬着陆" 。 这些方法都治标不治本。 但是，如果换个思路呢？ 与其想尽办法避免摔倒，不如把 "摔倒"本身，变成一门可以学习和控制的艺术。 就在最近，来自迪士尼研究院（ Disney Research）的一项最新研究，彻底颠覆了我们对机器人摔倒的认 知。他们提出了一种名为"机器人速成班：学习柔软且风格化的摔倒"（Robot Crash Course: Learning Soft and Stylized Falling）的全新方法。 这项研究的核心思想是： 让机器人不仅能摔得 "软"，最大限度减少冲击和损伤，还能摔得"帅"，在倒地后摆 出一个用户指定的、充满艺术感的姿势。 想象一下，一个机器人在舞台上出现失误，它没有僵硬地倒下，而是顺势一个翻滚，最后以一个帅气的卧倒姿 势结束，不仅没出糗，反而秀了一波操作。这简直是把 Bug玩成了绝活！ 这项研究成果，不仅能让机器人在娱乐、影视等行业大放异彩，更能为机器人的安全和快速恢复提供全新的解 决方案。一个能控制自己摔倒姿 ...