Core Viewpoint - The article highlights the advancements showcased at the 2025 World Robot Conference (WRC), particularly focusing on the innovative humanoid robots developed by Zhujidi Power, emphasizing their practical applications and technological breakthroughs in robotics [1][38]. Group 1: Humanoid Robots - LimX Oli, a 165cm humanoid robot, is praised for its graceful movements and ability to perform live dance demonstrations outdoors, showcasing 31 active degrees of freedom [12][17]. - The robot's ability to switch between various walking styles, such as a brisk walk and a model walk, demonstrates its advanced walking control algorithms that unify different walking modes without separate programming [19]. - LimX Oli's performance includes executing traditional Chinese martial arts and dance, highlighting the precision of its movements and the integration of hardware and algorithms [15][14]. Group 2: TRON 1 Robot - TRON 1 is characterized as a robust robot capable of withstanding physical challenges, such as climbing stairs and jumping from heights, thanks to its advanced reinforcement learning algorithms [20][22]. - The robot can autonomously recover from falls, showcasing its high precision in joint control and communication stability, which is crucial for practical applications in research and development [24]. - TRON 1 also features a modular design that allows for the integration of various tools, such as mechanical arms for tasks like waste collection, enhancing its versatility [28]. Group 3: Remote Operation System - The humanoid remote operation system developed by Zhujidi Power allows operators to control robots in real-time using VR technology, effectively mirroring their movements [29][31]. - This system is designed with ergonomic considerations, ensuring that the robot's movements closely match those of the operator, which is essential for effective interaction and data collection [32]. - The system's ability to gather first-person perspective data is valuable for training more intelligent robotic algorithms, enhancing the overall learning process [32]. Group 4: Overall Significance - Zhujidi Power's approach at WRC emphasizes creating a closed-loop system for embodied intelligence, integrating physical design, control algorithms, and data interaction to enhance usability and adaptability in real-world applications [35][37]. - The company received recognition as one of the "Top 20 Most Investable Robot Companies in China," reflecting its practical and innovative path in the robotics industry [38].

机器人大讲堂获悉，戴盟机器人（Daimon Robotics）近日官宣成功完成 亿元级天使++轮融资。该轮融资由 招商局创投领投，东方嘉富、架桥资本跟投 ，融资资金将主要用于助力戴盟加速全球领先的视触觉感知与灵 巧操作技术的落地应用，持续引领具身智能技术的产业化进程。 值得关注的是，戴盟机器人此前已先后完成由 昆仲资本独投 的天使轮融资，以及 国中资本、金鼎资本、联想 创投和招银国际联合投资 的天使+轮融资。其天使轮、天使+轮与天使++轮累计融资金额达数亿元，不仅刷新 触觉传感领域天使轮融资的 最高纪录 ，更印证了顶级产业方和资本市场对戴盟单色光视触觉技术路线的绝对 信心。 01 高分辨率触觉成为全球具身智能的"临门一脚" 近年来，国内外具身智能行业正加速迈入大规模应用的临界点。随着人形机器人、服务机器人、特种机器人等 产品逐步走向市场，行业需求已从 "能动" 转向 "能干活" 。在工业制造、智慧物流、医疗康养等任务中， 仅 依赖视觉 的机器人在柔性物体抓取、微型部件装配、动态交互等操作中依然存在精度不足、稳定性欠佳、泛 化能力弱等问题，制约了其实际应用场景的拓展。 与此同时，大多数机器人 触觉传感器也面临技 ...

Group 1 - Zhiyuan Robotics has secured a multi-million yuan order from Fulian Precision Engineering, marking the first large-scale commercial contract for embodied robots in the industrial sector in China [3] - The order includes nearly 100 units of the Yuanzheng A2-W robots, which will be deployed in Fulian's factory, a leading company in the automotive parts sector [3] - In July, Zhiyuan Robotics also won a humanoid biped robot contract from China Mobile (Hangzhou) with a total budget of 124.05 million yuan, making it the largest single procurement in the domestic humanoid robot field [3] Group 2 - Lingdong General Robotics has completed a multi-hundred million yuan angel round financing, led by Yuanhe Puhua, to accelerate the development and industrialization of its embodied VLA large model and general-purpose robots [6] - The company focuses on creating highly generalized embodied large models and robust robot bodies, utilizing a fully self-developed architecture [6] Group 3 - The World Robot Cooperation Organization released the "2025 Trends in Embodied Intelligent Robotics," highlighting ten key trends including the development of large factories for embodied intelligent robots and the creation of large-scale high-quality datasets [9] - Key trends include the integration of physical practice and simulation, multi-modal large models for decision-making, and the establishment of open-source communities for embodied intelligent robotics [9] Group 4 - The world's first humanoid robot sports event will take place in Beijing on August 14, featuring 26 competitions that challenge robots in traditional human sports [12] - Events will include running, high jump, long jump, gymnastics, and team sports, as well as practical applications like factory material handling and hospital medication sorting [12] Group 5 - A Harvard University research team has developed a personalized wearable robot control system that significantly improves upper limb function for stroke and ALS patients, as published in Nature Communications [15] - The innovation lies in the integration of intent recognition models with physical lag models for adaptive control, enhancing the effectiveness of inflatable soft exoskeletons [15]

在人形机器人赛道加速狂奔的当下，核心零部件的性能与量产能力始终是制约行业落地的关键瓶颈。 其中， 机器人灵巧手作为人形机器人的核心功能部件 ，在其中占据了重要地位，目前也是人形机器人板块技术迭代 最快的部件。 中商产业研究院数据显示， 2024 年灵巧手市场容量为 76.01 万只，随着灵巧手功能持续完善，叠加下游人 形机器人的放量，预计 2025 年灵巧手市场容量将达 86.18 万只，对应市场规模为 19.21 亿美元 。而 根 据 QYResearch 预测， 2030 年全球机器人多指灵巧手市场规模将超过 50 亿美元， 2025-2031年CAGR 为68.7% 。 2025 年世界机器人大会（ WRC ）上， 机器人大讲堂注意到， 中国初创企业曦诺未来 就 带来 了 三 款 突破性 产品 ，包括 全球首款全自研量产的 25 自由度腱绳传动灵巧手 ， 扭矩密度最 高 可 达 322Nm/kg 的关节模组 和 推力密度最 高 可达 5.2N/g 的 微型 电缸 ， 硬核参数领先国际顶尖水平，为行业提供了一 套从技术原理到工程落地的完整解决方案。 ▍ 8 个月崛起：技术积累的 " 厚积薄发 " 全球首款 ...

Core Viewpoint - Soft robots based on fluids, particularly Ferrofluid Droplet Robots (FDR), have gained significant attention in micro-scale intelligent manipulation due to their deformability, self-recovery, and environmental adaptability, promising new paradigms in various fields such as biomedical engineering and microfluidics [1][4]. Group 1: Programmable Control Platform - A programmable control platform for large-scale FDR has been developed by a research team at Tsinghua University, enabling simultaneous control of multiple droplet robots for complex tasks [4][6]. - The platform features a compact array structure that enhances robot mobility and operates on low power, ensuring long-term stability [4][6]. - The system integrates visual feedback strategies, allowing the robot formation to track multiple independent paths [4][6]. Group 2: Distributed Magnetic Field System - A high-density distributed electromagnetic coil array has been designed to achieve large-scale programmable control of droplet robots at the millimeter scale [6][7]. - The system consists of 144 small electromagnetic coils, generating localized magnetic fields within a workspace of 113 × 113 mm², enabling independent multi-point driving [6][7]. - The modular power architecture supports a total power of up to 36 A and 400 W, ensuring efficient operation of the coils [6][7]. Group 3: Motion and Deformation Mechanism - The interaction between FDR and external magnetic fields has been analyzed, revealing that FDR can be magnetized and moved towards areas of higher magnetic field strength [7][10]. - The deformation of FDR is a result of the dynamic balance between magnetic forces and surface tension, allowing for various forms of stretching, splitting, and merging [7][10]. - The process is reversible and programmable, providing a solid physical foundation for precise control of FDR [7][10]. Group 4: Control Strategies and Performance Validation - A discrete closed-loop control strategy based on visual feedback has been proposed to enhance the path tracking performance of FDR in distributed magnetic fields [10][13]. - The system has been validated through experiments, demonstrating the ability to track predefined paths with average position errors of 0.4 mm and 1.0 mm for different movement patterns [13][17]. - Reliability tests showed that even after 1,000 cycles, FDR maintained precise tracking and effective driving, confirming the system's stability and thermal performance [17][18]. Group 5: Applications and Experimental Results - The research team conducted various experiments to showcase the capabilities of the programmable control system, including droplet sorting, coordinated movement, and integrated manipulation of motion, deformation, splitting, and merging [20][21]. - The droplet sorting experiment demonstrated a 300% increase in task efficiency when multiple FDR collaborated compared to a single FDR [21]. - The system's potential for complex multi-modal control was further illustrated through experiments involving integrated movement and transformation of FDR [25][30].

新疆大学陈涛博士团队近期在机器人控制研究领域取得重要进展，其关于六轴 机械臂先进运动控制的核心成 果，已连续发表两篇文章于核心期刊中。该研究聚焦提升机械臂在复杂任务中的控制 性能，依托中科深谷 CSPACE开源异构协同智脑系统开展研究，全程基于深谷六轴机械臂平台开展实验验证。 ▍ 深耕核心算法：高精度轨迹跟踪控制成果丰硕 研究团队致力于解决六轴机械臂应用中的关键控制难题。已发表的两篇论文，分别深入探讨了积分增强型控制 方案及迭代学习融合方案。前者通过嵌入式积分项实现"即时纠偏"，在动态场景下精度更高；后者则依赖任 务重复性实现"渐进优化"，在长期重复任务中误差抑制能力更强。这些成果为机械臂在精密操作等场景的应 用提供了有力的技术支撑。 中科深谷核心产品CSPACE开源异构协同智脑系统，具备模块化、开放性特点，可实现机器人"大小脑协 同"控制，解决智能体对环境与任务的建模、仿真、决策与控制问题，为科研提供高效、可靠、通用的综合解 决方案。 | 第59卷 第1期 西 安 交 通 大 学 学 报 Vol. 59 No. 1 | | --- | | 2025年1月 Jan. 2025 | ▍ 开源平台赋能：中科深谷六 ...

Core Viewpoint - The recent events surrounding Berlangte Robotics have raised significant concerns regarding corporate governance and the actions of its founder and chairman, Yin Rongzao, leading to calls for a restructuring of the company's management and governance [1][10][11]. Group 1: Company Governance Issues - The board of Berlangte Robotics rejected a proposal to increase Yin Rongzao's monthly salary to 2 million yuan, citing the need to address the negative impact of two audit reports that questioned the company's financial practices [6][7]. - A public letter from the investment firm Junlan Investment accused Yin Rongzao of exploiting the company's governance structure for personal gain and called for a vote to amend the company's articles of association to restore normal governance [1][10][11]. - The letter highlighted a shocking proposal from Yin Rongzao to allocate all net profits minus one yuan as his personal bonus over the next decade, which was perceived as a blatant attempt to siphon off company resources [10][11]. Group 2: Financial Performance - Berlangte Robotics has experienced a decline in sales and profitability from 2020 to 2023, with revenue dropping from approximately 447 million yuan in 2020 to about 201 million yuan in 2023, and net profit shifting from a profit of 73.8 million yuan to a loss of 125.4 million yuan [7][16]. - The company reported a slight recovery in revenue in 2024, but it still faced a net loss of 37.4 million yuan, with the first half of 2025 showing a reduced loss of 7.4 million yuan [16]. Group 3: Shareholder Dynamics - Junlan Investment, which holds 1.54% of Berlangte's shares, has taken a stand against Yin Rongzao's management practices, emphasizing the need for shareholder representation and accountability [10][11]. - The governance changes initiated by Yin Rongzao have led to concerns about the exclusion of external shareholders from board representation, effectively consolidating his control over the company [11][12].

2025 世界机器人大会（2025 WRC）博览会上，人形机器人产业链上下游企业纷纷携最新产品亮相，展示 前沿技术， 人形机器人无疑成为了全场最大焦点。 近年来，人形机器人行业发展势头极为迅猛，众多科研团队与企业持续发力，推动技术突破不断涌现。从核心 零部件到整机解决方案，从机械结构的优化升级到智能化水平的显著提升，人形机器人技术成熟度持续提高， 正逐步从概念走向现实应用。在这一进程中， 灵巧手 作为人形机器人与环境交互的最终执行端，其综合性能 不仅 是衡量整机可用性的关键指标之一，更成为下一阶段产业链竞逐的核心技术焦点。 由中国科学院院士、吉林大学教授、国际仿生工程学会名誉理事长任露泉院士，与英国曼彻斯特大学终身教 授、国家特聘专家、教育部长江学者任雷教授共同创立的北京达奇月泉仿生科技有限公司 （以下简称 "月泉 仿生"），长期致力于推动仿生科技领域的创新与突破。前不久，月泉仿生 正式对外推出了仿生拉压体灵巧手 应手 Y-Hand M1 。 该灵巧手不仅 以 38个自由度创下全球最高纪录，更在力量、速度、精度、柔顺性、灵 活性五个维度实现了五位一体的性能突破 ， 彻底革新了仿生灵巧手的技术边界，为该领域的发 ...

Core Viewpoint - The forum highlighted the significant progress of embodied intelligence technology transitioning from technical breakthroughs to industrial applications, emphasizing the need for robots to be not only capable but also practical in real-world scenarios [2][23]. Group 1: Key Innovations - The "Embodied World Model System" was introduced, featuring a 72B multimodal model that enhances robots' understanding of reality and predictive capabilities [4][7]. - The "Cross-Entity VLA Model" XR-1 was launched, enabling a single model to operate across different entities and adapt to various scenarios and skills [6][9]. - The "Humanoid Robot Full-Body Control Autonomous Navigation System" was unveiled, addressing the challenges of mobility and operational coordination [8][11]. - The "Thousand Robot Real-Scene Data Collection Plan" was initiated to gather extensive multimodal interaction data from real industrial environments, enhancing model training and iteration [10][13]. Group 2: Industry Collaboration - The forum facilitated partnerships between Beijing Humanoid and several key industry players, including Baidu Smart Cloud and China Electric Power Research Institute, to promote the practical application of embodied intelligence technology [21][24]. - A simulation training platform was launched to address the industry's lack of standardized simulation data, providing a high-fidelity training environment for developers [21][24]. Group 3: Future Prospects - The continuous breakthroughs in core technologies and the expansion of application scenarios are expected to lead to the large-scale implementation of embodied intelligence, significantly supporting the development of the manufacturing and intelligent economy [23][24].

界机器人 CONFERENCE ROROT PLACE 北人亦创国际会展中心 会议中心一层多功能厅 主办单位 中关村石景山园管委会区科委 北京石景山科技创新集团 时间 TIME 2025年8月11日 9:00-11:35 地 点 乐聚通研(北京)机器人技术有限公司 北京中关村通力科技服务有限责任公司 活用以后 0 09:00-09:05 主持人开场 909:05-09:10 北京市石景山区领导致辞 合作签约 009:10-09:15 中关村通力与乐聚通研、紫东太初、睿尔曼智能、 石景山科技创新集团"共建石景山人形机器人通用 智能基座战略合作协议"签约 09:15-09:20 乐聚通研与北极星途"多功能智能末端执行器战略 合作协议"签约 009:20-09:35 具身智能操作系统创新与实现 中 织 粉 哈尔滨工业大学教授、教育部长江学者特聘教授 009:35-09:50 从数字AI走向物理AI 夏 文华 北京大学武汉人工智能研究院副院长 909:50-10:05 生成式大模型驱动的具身智能大规模高质 量数据合成开源平台 山东大学机器人研究中心教授、国家级青年人才 9 10:20-10:35 NK 榻 上海交通大 ...