Group 1 - The article reviews seven selected papers published in the International Journal of Robotics Research, covering various research directions in robotics such as soft actuators, human-robot interaction, dual-arm robots, multi-robot systems, and bipedal locomotion control [1][6][18][27][38][48][58] Group 2 - A new low-profile soft rotary pneumatic actuator was designed, addressing the limitations of traditional soft pneumatic actuators in confined spaces and providing a compact solution for applications in wearable devices and biomedical devices [2][4] - The THÖR-MAGNI dataset was introduced to overcome data bottlenecks in social navigation and human-robot interaction, featuring multi-modal data collection and extensive scene coverage [6][11][14] - The FMB benchmark was developed to standardize robotic manipulation research, offering a diverse set of objects and a modular framework for imitation learning [18][20][24][26] - A framework for dual-arm robots to manipulate deformable linear objects in constrained environments was proposed, achieving high success rates in complex tasks [27][30][31][34] - A real-time planning method for large heterogeneous multi-robot systems was introduced, significantly improving computational efficiency and robustness in dynamic environments [38][40][45] - A survey on communicating robot learning during human-robot interaction highlighted the importance of a closed-loop communication framework to enhance collaboration [48][50][53][55] - Reinforcement learning was applied to bipedal locomotion control, demonstrating significant advancements in adaptability and robustness in complex environments [58][60][62]

Core Insights - The humanoid robot industry in Zhangjiang, Shanghai, is expected to experience explosive growth by 2025, with significant improvements in application capabilities [1] - China has captured 63% of the global humanoid robot supply chain, with domestic core components achieving technological breakthroughs and increasing localization rates [1][2] - The recent U.S. tariff increase of up to 245% on Chinese imports poses new challenges for the industry [1] Industry Development - Eleven major humanoid robot manufacturers in China initiated mass production plans last year, with six companies planning to produce over 1,000 units by 2025 [2] - The establishment of the first full-size humanoid robot open-source platform "Qinglong" and the OpenLoong community aims to set industry standards for healthy development [2] - Major companies like Haier, Midea, and Gree are accelerating their investments in the humanoid robot sector [2] Event Overview - The "2025 Zhangjiang Humanoid Robot Developer Conference" will be held on May 29, focusing on technology implementation and ecosystem building [3][5] - The event will gather over 200 leading companies and 1,000 experts to discuss new trends and directions in the industry [3][5] Zhangjiang's Competitive Advantage - Zhangjiang has attracted 73 humanoid robot-related companies, forming a complete industrial chain from core components to application terminals [7] - Government policies are increasingly supportive, with initiatives aimed at establishing Zhangjiang as a hub for humanoid robot development [7] - The conference aims to showcase Zhangjiang's entrepreneurial environment and industry ecosystem to attract technology, talent, and capital [7] Conference Highlights - The conference will emphasize "open-source, openness, and innovation" to promote high-quality development in the humanoid robot industry [12] - Various forums and competitions will facilitate in-depth discussions on technology breakthroughs and commercialization paths [15][21] - The event will feature multiple developer forums focusing on key areas such as core components, data sets, and application scenarios [21][24][25] Innovation and Entrepreneurship Competition - The "2025 Zhangjiang Humanoid Robot Innovation and Entrepreneurship Competition" will include five tracks aimed at identifying commercially viable projects [29][30] - The competition will cover real-world application challenges in industrial, household, and dynamic environments [32] - Winners will receive substantial rewards, including cash prizes and opportunities for government collaboration [34][35] Exhibition and Ecosystem Building - An innovation product exhibition will showcase cutting-edge humanoid robots and core components, providing a platform for interaction [36][37] - The exhibition will feature four thematic areas, highlighting advancements in humanoid robots, embodied intelligence, and core components [37] Conclusion - The humanoid robot industry is poised for unprecedented growth, with Zhangjiang leveraging its complete industrial chain, open innovation ecosystem, and strong policy support to seize opportunities in the upcoming production year [38]

合成数据是放大人类示范的 "放大器"，是打通具身智能"最后一公里"的关键钥匙。与真实场景数据的采 集面临的成本高、周期长、场景覆盖率低等问题不同，具身仿真数据构建的是场景足够丰富、物理交互 真实及人在环的数据，能够有效打通机器人进入物理世界现存的技术难点。 近日，在第二届中国人形机器人与具身智能产业大会上，光轮智能（北京）科技有限公司联合创始人兼 总裁杨海波，发表了主题为 《合成数据：具身智能的必经之路》 的演讲，杨海波认为，在实现通用人工 智能（ AGI）的过程中，AI必须具备与复杂物理世界交互并深入理解的能力。 杨海波强调，具身合成数 据有三个必备条件：一是要有足够真实的物理交互能力、二是要有人的示范在环、三是场景足够丰富。 光轮智能（北京）科技有限公司联合创始人 &总裁 杨海波 杨海波在会上指出，物理交互是让机器人真正 "学会做事"的基础，就像学游泳不能光在岸上看，必须跳 进水里感受浮力和阻力一样。 要让机器人像真人一样学习，必须模拟出真实的物理反馈，比如物体的重 量、材质软硬、碰撞时的反应、关节如何活动，以及动作之间的逻辑联系（比如按下开关灯才会亮）等 物理属性，细节构成的动态交互体验，才能帮机器人在 ...

Core Viewpoint - The research highlights the unique jumping mechanism of entomopathogenic nematodes (EPNs) and its potential applications in soft robotics, demonstrating that EPNs can jump up to 20 times their body length and generate approximately 10,000 watts per kilogram of power, surpassing traditional soft robots [19][21]. Group 1: EPNs Characteristics and Research Methodology - EPNs are unique organisms capable of jumping distances 20 times their body length, which is comparable to a human jumping to the height of a three-story building [1]. - To obtain representative EPN samples, researchers cultivated EPNs in a controlled laboratory environment using wax moth larvae as hosts [2]. - The study involved the development of a physical model named SoftJM to simulate the jumping mechanism of EPNs, validated through soft robotics [4]. Group 2: Experimental Design and Findings - Researchers designed a specific experimental setup to induce jumping behavior in EPNs, utilizing vertical filter paper as a jumping platform and controlling environmental conditions [6]. - High-speed cameras recorded the jumping process at frame rates of 10,000 to 30,000 frames per second, capturing key parameters such as body shape changes, jump height, speed, and acceleration [8][9]. - The jumping process of EPNs was divided into three stages: formation of a ring structure, formation of a knot structure, and expansion of the ring structure, which facilitated energy storage and release [19][20]. Group 3: SoftJM Models and Performance Validation - Four types of bio-inspired physical models (SoftJM) were designed to mimic EPNs' jumping mechanism, with variations in stiffness and structure to enhance energy storage and release [11][22]. - The SoftJM 4 model, enhanced with a carbon fiber skeleton, achieved a jump height of up to 25 times its body length, showcasing superior jumping performance [22]. - The performance of these models was closely linked to their aspect ratio and stiffness, indicating that lower aspect ratios and higher stiffness improve energy storage and release efficiency [24]. Group 4: Numerical Simulation and Mechanism Exploration - A Cosserat rod model was employed for numerical simulations to analyze the jumping dynamics of EPNs, confirming the accuracy of experimental results [15][25]. - The simulations revealed that knot instability plays a crucial role in the jumping process, allowing EPNs to efficiently store and release energy [25][27]. - The findings provide theoretical support for optimizing soft robot designs by adjusting aspect ratios and stiffness to enhance jumping performance [27]. Group 5: Future Directions - The research team emphasizes the potential of reversible knot instability in improving soft robot jumping performance and suggests exploring its applications in crawling, swimming, or grasping tasks [28].

Core Viewpoint - The article discusses the rapid development of Chinese robotics companies, particularly in the pool cleaning robot sector, driven by capital investment and innovative strategies for international expansion [1][2][4]. Group 1: Pool Cleaning Robots - The pool cleaning robot market has seen significant investment, with companies like Yuan Ding Intelligent completing a strategic financing round of nearly 1 billion RMB, supported by global industry leader Fluidra [2][4]. - Yuan Ding Intelligent's Aiper brand ranks among the top three in the global pool cleaning robot market, achieving a market share of over 35% on platforms like Amazon, with revenue growth nearing 10 times from 2020 to 2024 [4][5]. - Another company, Xingmai Innovation, has rapidly captured 85% of the high-end market segment for pool cleaning robots priced above $1,400, with sales reaching several hundred million RMB within six months [5][6]. - The global pool market is projected to grow from approximately 30 million pools in 2021 to nearly 40 million by 2026, with the automated pool robot market expected to reach $2.16 billion by 2029, reflecting a CAGR of 10.9% [8]. Group 2: Lawn Mowing Robots - The lawn mowing robot market is characterized by low penetration rates, with less than 10% market saturation, indicating significant growth potential as technology replaces traditional equipment [9][12]. - The Segway Navimow, developed by a subsidiary of Ninebot, achieved revenue of 861 million RMB in 2024, marking a 283.92% increase year-on-year, and becoming the first brand to cover over 100,000 households globally [9][12]. - The market for lawn mowing robots is supported by a large number of gardens, with approximately 250 million gardens worldwide, and a strong cultural emphasis on lawn maintenance in the U.S. and Europe [9][12]. Group 3: Logistics Robots - The demand for warehouse automation has surged due to the rapid growth of e-commerce and logistics, with Chinese companies like Geek+, Hai Robotics, and KuaiCang expanding into international markets [14][15]. - Geek+ has maintained the top market share in the global autonomous mobile robot (AMR) sector for six consecutive years, focusing on tailored solutions for different regional markets [15][18]. - The logistics robot sector benefits from China's robust supply chain capabilities, allowing for competitive pricing and high profit margins, with logistics robots achieving around 50% gross margins [18]. Group 4: Cooking Robots - The commercial kitchen robot market is projected to grow from $150 million in 2020 to $310 million by 2025, with a CAGR of 15.6%, indicating a significant opportunity for growth [19][20]. - Companies like Xianglu Robot and Bupt Technology have successfully entered international markets, with Xianglu's AI cooking robot reaching 40 million consumer interactions across various dining scenarios [22]. - The demand for standardized cooking processes in large restaurant chains is driving the adoption of AI cooking robots, which can ensure consistent quality and reduce labor costs [20][22]. Group 5: Overall Industry Trends - The international expansion of Chinese robotics companies reflects the ongoing advancements in technology and the pursuit of high-quality development within the sector [23].

1、 星宸科技：高阶机器人芯片有望于2026年量产 近日，星宸科技（301536）在其业绩说明会上表示，公司机器人芯片已初步完成产品矩阵，现已大批量出 货。技术布局方面，公司在清洁机器人打下良好基础后，正快速升级更大算力、具备多模态大模型、人机 交互的高阶机器人芯片，有望于2026年量产，未来包含各类具身智能机器人（如服务机器人、人形机器人 等），也将随市场情况及客户需求，及时提供对应芯片。市场布局方面，公司已成功切入多家头部客户， 有望在2025年跻身领先梯队，目标在2—3年内成为智能机器人行业头部的SoC芯片供应商。 2、 肯尼索州立大学推出旨在简化仓库库存跟踪的自主机器人 近日，肯尼索州立大学推出了一款可以记录库存的自主机器人——OmniScanBot。该机器人由RugGear U S LLC和堪萨斯州立大学计算机与软件工程学院合作开发，旨在简化、加快和保障仓库库存跟踪流程。与 传统的人工盘点方法不同，该机器人创新性地利用射频识别 (RFID) 技术实现了整个流程的自动化。此 外，OmniScanBot基于多传感器平台构建——包含摄像头、激光雷达和导航软件，能够绘制仓库地图并自 主规划路线。 3、 自主 ...

南京神源生智能科技有限公司董事长戴振东 表示， 人形机器人 板块 的热度飙升，对神源生智能来说是 一个巨大的市场机遇。神源生智能 是一家 专注于力测试技术研究及其产业化的高新技术企业。 公司旗 下拥有 多维力传感器、扭矩传感器 等多个系列产品。 这些传感器在人形机器人中扮演着至关重要的角 色，尤其是在力感知方面。 南京神源生智能科技有限公司董事长戴振东 资料来源：特斯拉 AI Day，Vehicle，中航证券、行行查 戴振东 表示， 人形机器人中大约会有 40多个力传感器，其成本占到总成 本 的 15%左右。因此，人形 机器人市场的快速发展，将直接带动力传感器的需求增长， 这也为 神源生智能 提供了一个更加广阔的 发展空间。 ▍ 13年运动力学积累 神源生智能 拥有出色的正向设计能力 2012年 神源生智能 成立，依托于南京航空航天大学的技术支持，公司拥有出色的正向设计能力，能够 根据 客户需求和行业特点，设计出性能优越、成本合理的传感器。此外，公司在运动力学方面的深厚积 累，也使其能够为客户提供更加专业的解决方案。 戴振东 认为 ，在机器人运动研究领域，完整全面地测定机器人运动过程的力至关重要，它是帮助机器 ...

在全球制造业智能化转型的浪潮中，中国工业机器人行业正迎来结构性变化。一方面，"智能化升级"与"设 备更新"政策红利持续释放，加速机器人向下沉市场渗透；另一方面，国产机器人技术平台不断升级，向汽 车、医药等高端场景开拓。 01 国产品牌登顶中国机器人市场第一 工业机器人作为推动全球制造业转型的关键驱动力，近年来取得了飞速的发展，中国工业机器人市场销量 从2015年的6.1万台增长至2024年的29.4万台，中国更是连续10多年稳居全球最大机器人市场宝座。随 着智能制造产业的发展，中国机器人市场进入调整期，行业迈入新增长阶段。 2015-2024年中国工业机器人整体市场销售规模推移（台） 数据来源：MIR DATABANK 注：*工业机器人产品包括：≤20kg 6-axis（包括其他垂直多关节）、>20kg 6-axis （包括其他垂直多关节） 、 据MIR DATABANK最新数据统计， 2025年Q1，中国工业机器人整体市场销量同比增长11.6%，国产机 器人渗透率同比提升2个百分点 。 埃斯顿机器人实现近20%的同比增长，占比中国机器人市场份额10.3% （注：含埃斯顿酷卓的市场份额） ， 成为首个登顶中国 ...

近日，苹果公司内部的一则重大调整消息引发了科技行业广泛关注。 苹果 CEO库克对公司人工智能部门进行了重组，其中一项关键举措是 将机器人团队从原本由 AI主管 John Giannandrea管理，转移至硬件工程高级副总裁John Ternus麾下。 这一调整不仅体现了苹果在战略布局上的新动向，也引发了外界对于苹果未来业务走向以及行业发展趋 势的诸多猜测。 ▍ 致力填补智能家居空白，苹果那些年曝光的机器人项目 苹果在人工智能领域的布局由来已久 。为实现 AI技术与产品服务的深度融合， 苹果此前那 不仅挖角谷 歌高管 John Giannandrea担任AI首席科学家，还曾计划参与OpenAI融资，并探讨将ChatGPT功能集 成至苹果设备。 但自 2011年推出语音助手Siri以来，苹果在AI领域的发展未达预期。 Siri升级多次延 期，大型语言模型开发持续投入资金却进展缓慢。 在此背景下， 今年年初有相关报道传出苹果正在寻找其 "下一个大事件"，机器人技术是其重点领域。 也因此，苹果此次组织架构调整引发业界诸多猜测：机器人团队并入硬件部门是否意味着战略地位提 升？苹果后续是否会推出机器人产品？其是否正从传统 ...

全球首场人形机器人半程马拉松，不仅是对双足运动极限的挑战，更是行业技术突破的试金石。这场赛 事让我们看到频频出现的意外情况：机器人 "累到头掉"、需要人工喷淋、慢速奔跑的"佛系半马"等现 象。赛事暴露了机器人的关键瓶颈——连续弯道中的动态平衡、高强度奔跑下的关节损耗、高能耗与轻 量化之间的矛盾，本质上都在考验机器人"关节模组"的硬实力。 当行业在追求速度与市场热度时，脉塔智能的工程师团队秉持长期主义不断思考如何助力智能机器人应 用落地，深知关节不是冰冷的传动部件，而是驱动智能体的 "神经元"。脉塔工程师通过大量的实验和打 磨，在功率密度、扭矩、动态响应、中空设计、通信方式等多方面取得了新的突破， 在五一前夕隆重发 布为具身机器人量身定制的 RMD-X V4系列智能行星关节模组 。此次发布的产品，融入了新的巧思， 将关节的扭矩、电流、位置等数据进一步融合，努力使其成为会 "会思考"的关节模组。V4系列关节模组 将帮助机器人从踉跄学步的参赛者，蜕变为刷新运动极限的领跑者。 ▍ 【产品矩阵】全场景覆盖，构筑具身硬件新基座 RMD-X V4系列一共包含7款产品，产品内部用料十足，高精度双编码器，交叉滚子轴承、自研高性 ...