人民财讯8月11日电，据亿纬锂能（300014）消息，维他动力与亿纬锂能正式达成机器人业务深度合 作。双方将通力协作，基于用户需求和机器人动力电池技术发展趋势，推进具身智能产品化量产的进一 步发展。 ...

上天天基金APP搜索【777】注册即可 领500元券包，优选机器人基金10元起投！限量发 放！先到先得！ 2025世界机器人大会于8月8日至8月12日召开。会上，全球机器人产业发展趋势和具身智能企业全球化 市场路径，成为多位国内外行业人士热议的焦点。 记者在大会上采访多家企业及行业人士获悉，受技术突破、政策驱动、资本涌动、劳动力短缺问题日益 突出等多重因素影响，一方面，今年以来机器人产业增长显著，未来有望为全球经济发展带来新的机 遇；另一方面，具身智能企业正积极探索出海路径，全球化成为关键词。 中国有望引领机器人产业发展 机器人产业未来有望为全球经济发展带来新的机遇。多位国际行业人士表示，机器人密度与产业智能化 水平，将成为衡量国家竞争力的核心指标，这将为人们创造更多与机器人协同工作的机会，推动社会经 济的全面发展。 2025年是物理人工智能年，物理AI潮流正推动机器人产业快速发展。英伟达Omniverse和仿真技术副总 裁雷夫·勒巴雷迪安（Rev Lebaredian）表示，智能机器人的出现将从根本上改变人类与物理世界的互动 方式：它们会把计算机技术与价值数百万亿美元的物理世界相连，释放出巨大的生产力，这对 ...

Investment Rating - The industry investment rating is "Overweight" [3][24]. Core Viewpoints - The 2025 World Robot Conference (WRC) showcased over 1,500 exhibits from more than 200 companies, highlighting the rapid commercialization of humanoid robots, with over 27 new models launched in July and August [6][7][12]. - The humanoid robot industry is entering a critical phase of technological breakthroughs and commercialization, with a projected global demand of approximately 2 million units by 2030 [6][20]. - The report emphasizes the importance of core technology innovation and cost control for companies to succeed in a competitive environment similar to the smartphone industry [12][20]. Summary by Sections Humanoid Robots - The WRC featured 50 humanoid robot companies, showcasing innovations in both industrial and consumer applications [7]. - The BOM cost for high-end humanoid robots remains high, ranging from 200,000 to 500,000 yuan, but lower-priced models like the Unitree R1 start at 39,900 yuan, facilitating market entry [19]. - Government subsidies for purchasing robots are available, with businesses eligible for up to 250,000 yuan and individuals for 1,500 yuan [19]. Key Recommendations - Recommended companies include Sanhui Electric, Jingpin Special Equipment, Zhejiang Rongtai, Beite Technology, Hanwei Technology, and Ruantong Power [4][6]. - The report suggests tracking the supply chain of humanoid robots, including companies like Sanhua Intelligent Controls, Top Group, and others [20]. Other Industry Insights - In the photovoltaic sector, the penetration rate of N-type technology is accelerating, strengthening the competitive edge of leading companies [21]. - The energy storage market is expected to grow due to favorable policies and increasing demand from both generation and user sides [21]. - The semiconductor equipment market is projected to reach $140 billion by 2030, with a focus on domestic alternatives [22]. - The automation market, particularly industrial consumables, is expected to grow from 40 billion yuan to 55.7 billion yuan by 2026 [22].

在2025世界机器人大会的江淮前沿技术协同创新中心展区内，一台结构精巧的绳驱机器人正在流畅地完 成抓取、传递等一系列精细动作，吸引众多观众驻足观看。这是由江淮前沿技术协同创新中心自主研发 的全省首款绳驱S1机器人，其独特的驱动方式和精巧设计，标志着我省在具身智能领域取得重要突 破。据了解，绳驱S1机器人全身共46个自由度，配备激光雷达、毫米波雷达、深度相机、视触觉传感 器与六维力传感器等多种传感设备，可实现多维度场景与交互感知。区别于常见的刚性臂机器人，S1 机器人采用的是7个自由度的绳驱臂。通过在电机与关节之间增加绳索进行传动，利用绳索自身的柔性 震动消减冲击力，使得末端接触力呈现递减的趋势。（安徽日报） ...

Core Insights - GAC Group showcased its second-generation humanoid robots, GoMove, GoSide, and the third-generation GoMate, at the 2025 World Robot Conference, highlighting applications in healthcare and security [1] - GAC Group's humanoid robots were included in Morgan Stanley's global top 100 humanoid robots list earlier this year [1] - The company has achieved self-design and manufacturing of key components for its humanoid robots and plans to establish a global demonstration area in Guangzhou by the end of 2023, with large-scale production expected to start in 2027 and projected revenue exceeding 1 billion yuan by 2030 [1] Product Highlights - The third-generation humanoid robot GoMate features an industry-first variable wheel-foot configuration, allowing it to operate in both two-wheel and four-wheel modes, reducing energy consumption by over 80% compared to similar products and achieving a 6-hour battery life [2] - GoMate's self-developed drive unit weighs only 40 grams and is 80% smaller than the industry average, while its dexterous hand integrates visual and tactile sensors for precise object manipulation [2] - The robot is equipped with proprietary visual autonomous driving algorithms and advanced AI perception technologies, enabling rapid response to complex commands and seamless switching between remote and autonomous control [2] Market Outlook - The Chinese humanoid robot market is projected to reach 38 billion yuan by 2030, with a compound annual growth rate exceeding 61% [2] - The year 2025 is anticipated to mark the beginning of mass production for humanoid robots, positioning this sector for high growth, particularly for companies with established automotive supply chains that can leverage component commonality [2]

Market Overview - The A-share market showed strong performance with all three major indices rising, and the Shanghai Composite Index reached a new high for the year [1][2] Lithium Battery Industry - The lithium battery supply chain saw significant gains, with Tianqi Lithium and Ganfeng Lithium experiencing substantial stock price increases, and lithium carbonate futures contracts hitting the daily limit [4][6] - The price of battery-grade lithium carbonate rose by 8,000 yuan, averaging 78,000 yuan per ton [6] - The recent suspension of mining operations at CATL's Jiangxiawo lithium mine has raised concerns about supply, potentially impacting monthly supply by approximately 0.8 million tons, which accounts for 8% of domestic supply [7][8] Technology Sector - The technology sector, particularly AI and computing power supply chains, performed well, with significant gains in stocks like New Yisheng and Zhongji Xuchuang [2][9] - The humanoid robot sector surged following the World Robot Conference, with companies like Shuangyi Technology and Chaojie Co. seeing their stock prices hit the daily limit [9][12] Financial Sector - The financial sector showed active performance, with gains in brokerage, insurance, and diversified financials, while the banking sector experienced a slight decline [2] Defense and Consumer Goods - The defense and military industry, as well as the liquor and real estate sectors, contributed positively to market sentiment [2]

机器之心原创 作者：张倩 人类手部是自然进化的杰出产物，从早期灵长类的简单结构演变而来，兼具力量与灵活性。 在「用进废退」原则的驱动下，拇指逐渐变长且肌肉发达，能够与其他手指精准对握，实现精细操作。手部由多个关节组成，具有高度灵活性，可完成捏、握、 夹等多种动作。此外，手部皮肤富含触觉感受器，能够感知压力、温度和纹理等信息，为操作提供精准反馈。 凭借卓越的灵巧性和精细控制能力，手部能够执行 多样化任务，成为人类与物理世界交互的关键工具。 人手构型演化过程 以人为参照的机器人系统一直是人类的梦想，如果能在机器人身上复刻人手的这些能力，具身智能的未来会是什么样子？ 但让「灵巧手」像人手一样灵活并非易事。人手在非结构化环境中表现出卓越的灵巧性、柔顺性和鲁棒性，能完成精细操作和力量型任务，其优势源于肌肉骨骼 的串并混联结构、高冗余肌肉骨骼刚柔耦合驱动、皮肤感知多元信息、肌腱驱动的被动柔性和变刚度特性，以及神经机理控制等。 然而， 人手的复杂结构和功能机制难以完全解释和模拟，存在功能解释模糊、物理器件模拟不足、非线性驱动机理难以推导、多尺度操作泛化困难等问题 。此 外，生物与人工系统在驱动、传感和控制上的本质差异，导致 ...

本文一作为陈骁，香港中文大学 MMLab - 上海人工智能实验室具身智能中心联培博士生，研究方向是三维计算机视觉和具身智能，导师为薛天帆教授。个人主 页：xiao-chen.tech/。 研究背景 当人类走入陌生房间时，会通过移动和观察来掌握室内结构。想象机器人被扔进一个陌生场景：有的房间堆满障碍，有的走廊九曲十八弯，它能像人类一样主动 探索未知空间吗？ "主动探索" 这一智能基石，何以成为技术盲区？ 经典方案往往依赖人工预设的轨迹、视角与指令，而现有探索策略在陌生复杂场景中频频失效：机器人既可能在废墟救援时因全局规划缺失而卡死墙角，又容易 在障碍密集的客厅中反复碰撞进退维谷。当机器人在此类复杂环境下运转时，感知 - 决策 - 行动闭环如何挣脱被动依赖桎梏？ 这正是下一代机器人跨越 "智能鸿 沟" 的核心挑战。 如何让机器人在完全未知的复杂房间里自主探索？ 针对移动机器人在复杂未知环境中 "探索 - 建图" 的泛化难题，香港中文大学与上海人工智能实验室联合提出系统性解决方案：研究者们搭建了全球规模最大的 " 探索 - 建图" 基准 GLEAM-Bench—— 该数据集涵盖上千个室内场景，并在此基础上设计了通用 ...

具体而言，RynnRCP包括RCP框架和RobotMotion两个主要模块。RCP框架旨在建立机器人本体与传感器的连接，提供标准化能力接口，并实现不同的传输 层和模型服务之间的兼容。RobotMotion则是具身大模型与机器人本体控制之间的桥梁，能将离散的低频推理命令实时转换为高频的连续控制信号，实现平 滑、符合物理约束的机器人运动。同时，RobotMotion还提供了一体化仿真-真机控制工具，帮助开发者快速上手，支持任务规控、仿真同步、数据采集与回 放、轨迹可视化等功能，降低策略迁移难度。 具身智能领域飞速发展，但仍面临开发流程碎片化，数据、模型与机器人本体适配难等重大挑战。达摩院将MCP（Model Context Protocol）理念引入具身智 能，首次提出并开源了RCP（Robotics Context Protocol）协议以推动不同的数据、模型与本体之间的对接适配。达摩院打造了名为RynnRCP的一套完整的机 器人服务协议和框架，能够打通从传感器数据采集、模型推理到机器人动作执行的完整工作流，帮助用户根据自身场景轻松适配。RynnRCP现已经支持 Pi0、GR00T N1.5等多款热门模型以及S ...

具身智能的核心突破在于赋予机器"理解"与"规划"的能力。传统工业机器人依赖人类编程的精确指令，执行直接而单一的动作。随着多模态大模型和具身大 模型的发展，机器人开始具备理解模糊任务指令、进行动作规划并执行的能力，形成"任务理解-动作规划-执行"的复杂全链路架构，为突破柔性制造的瓶颈 提供了可能。 【环球网科技报道 记者 李文瑶】"个性化消费时代，多品种、小批量、高柔性的制造需求，与传统工业机器人的结构化封闭环境、单一指令模式产生了根本 性矛盾。"在2025世界机器人大会（WRC）现场，聆动通用CEO、科大讯飞机器人首席科学家季超如此点出当前制造业自动化的核心痛点。 效率与通用性的 " 悖论 " 这一矛盾正推动机器人控制技术经历一场静默革命——从过去基于精确模型（Model Base）的编程式控制，转向基于数据驱动学习（Learning Base）的具身 智能。其背后，是以Transformer架构为代表的大模型技术向机器人领域的深度渗透。 拥抱通用性并非没有代价。季超坦言，具身智能机器人在单一场景的效率目前可能仍低于专用机器人。他描绘了一个清晰的坐标轴：横轴是通用性，纵轴是 效率，二者呈现显著的负相关。"通用 ...