Core Viewpoint - The transaction of humanoid robot training data between Hubei Humanoid Robot Innovation Center and Zhiyuan Innovation (Shanghai) Technology Co., Ltd. marks the first inter-company trade of humanoid robot training data in China, which is expected to accelerate the commercialization and scaling of humanoid robots [1][2]. Group 1: Training Data and Methods - The Hubei Humanoid Robot Innovation Center sold thousands of hours of humanoid robot training data, which is essential for the development of humanoid robots [1]. - Humanoid robots require extensive training, often taking two to three months to learn a single action, accumulating tens of thousands of training data points [1]. - The training methods include remote operation, simulation, and video methods, which help in generating and injecting behavior data into humanoid robots [2]. Group 2: Data Requirements and Collaboration - To achieve significant advancements, humanoid robots need to store at least 40 billion effective data points to transition from being merely functional to truly useful [2]. - The sharing and collaboration of training data among companies can significantly reduce training costs and expedite the commercialization process of humanoid robots [2]. - The effectiveness of data transmission among different humanoid robots is crucial for their scalability and application across various industries [2].

转自：中国科学报 指尖轻轻触碰，便能感知苹果的爽脆、面包的松软、丝绸的细腻，这得益于人类皮肤历经数百万年进化 形成的精妙结构。人类指尖每平方厘米分布着约240个触觉感受器，它们能够同时捕捉压力、纹理、软 硬等多维信息，并瞬间传递给大脑，从而形成精准的"触觉画像"。倘若能为机器人赋予这样"指尖触 觉"，甚至使其具备更敏锐的感知能力，那么世界将充满无限可能。 然而，团队并未满足于此。何少帅介绍，此前的研究虽实现了材料种类识别率99.4%、材料软硬度识别 率100%的准确率，并制作出单个触感器在5毫米级的阵列集成，但要赋予机器人真正灵巧的"智能指 尖"，必须在传感器的微型化与集成度上继续取得突破。 迈向亚毫米级的"触觉指尖" 指纹间距约在一毫米以内，触感器要做到这个尺度以内，才能感知如指纹般精细的纹理，这意味着研究 需迈向亚毫米级的阵列集成。 "在港科大（广州），我们不仅要以科学家的思维发现问题、提供解决方案，更要以工程师的思维交 付'实际可用的样品'，让科研与工程贯通，走向真实应用场景。"訾云龙说。 近日，一项来自香港科技大学（广州）（以下简称港科大<广州>）的突破性研究，正逐步将这份"指尖 智慧"变为现实。该校 ...

Core Insights - The article highlights the journey of a technology company founded by graduates from Harbin Engineering University, focusing on underwater robots for practical applications, particularly in ship cleaning [1][3][4]. Group 1: Company Background and Development - The company was established in 2019 in Weihai, China, by a group of young engineers specializing in underwater robotics, exploring various applications such as sea cucumber harvesting and pool cleaning before settling on ship cleaning as a viable business model [1][4]. - The fifth-generation underwater robot is currently operating in the challenging waters of the Singapore Strait, showcasing the company's growth and technological advancements over five years [1][10]. Group 2: Business Model and Market Demand - The decision to focus on ship cleaning stems from the realization that shipping companies prioritize cost savings, particularly in fuel consumption, which can be significantly impacted by marine growth on ship hulls [5][16]. - The robots are designed for efficiency and reliability rather than complex features, meeting the industry's demand for stable performance in challenging conditions [7][16]. Group 3: Technological Innovation and Testing - The geographical advantage of Weihai, with its proximity to a deep-water testing facility, has accelerated the company's R&D process, allowing for rapid prototyping and testing of their robots [8][19]. - The company has successfully completed operations with its fourth-generation robot in Singapore, with the fifth generation on its way, demonstrating a commitment to continuous improvement and innovation [10][20]. Group 4: Strategic Positioning and Future Goals - Operating in the Singapore Strait, one of the busiest shipping lanes, presents both a challenge and an opportunity for the company, validating its technology in demanding conditions [11][13]. - The company's goals align with national strategies for deep-sea technology development, aiming to expand its market globally while also contributing to national strategic needs in deep-sea exploration [14][22]. Group 5: Collaborative Environment - Weihai's strong network of marine-related industries and academic institutions fosters collaboration, enabling the company to solve technical challenges efficiently [18][19]. - Government support and the availability of a national testing facility have been crucial in providing the necessary resources for the company's growth and innovation [20].

Group 1 - The stock market showed a low opening but rebounded, with the robotics concept recovering and turning positive. The robotics ETF (159770) saw a 1.12% increase in its index during trading, with a transaction volume of 132 million yuan. Key stocks such as Tianzhihang-U, Ruishun Technology, and Top Group rose over 5% [1] - The robotics ETF (159770) experienced a net subscription of over 16 million shares during the trading session, accumulating a net inflow of 194 million yuan over the last five trading days. As of February 5, 2026, the fund's latest scale reached 9.826 billion yuan, making it the largest in its category in the Shenzhen market [1] - The ETF closely tracks the robotics index, with major industry allocations including automation equipment (42.75%), software development (11.1%), and automotive parts (10.3%). The top five constituent stocks are iFlytek, Huichuan Technology, Top Group, Zhongkong Technology, and Dazhu Laser [1] Group 2 - Longcheng Securities believes that 2026 will be a critical year for mass production of robots both domestically and internationally, with companies like Tesla, Yushun, and Xiaopeng clearly outlining their production timelines. This will create high growth opportunities in core areas of the supply chain, such as lead screws, electronic skin, and "robot brains" [2] - Li Xiang, the founder of Ideal Auto, stated that the ultimate form of a car is a robot, aiming to transform vehicles into intelligent entities that actively serve users. The new Ideal L9 set to be released this year is expected to be a pioneering product in embodied intelligent robotics [1]

2月7日，"千联竞秀·万福满城"成都·青白江写春联送祝福活动在城厢古城举行。活动现场，数百张长桌 依次排开，来自青白江区的五百余名书法家、书法爱好者及中小学生同写春联。4台人形机器人也"灵 巧"地握持毛笔，与书法家同台展开了一场别开生面的"书写竞技"。图为人形机器人写春联。中新社记 者 张浪 摄 图为人形机器人写春联。中新社记者 张浪 摄 现场。中新社记者 张浪 摄 图为活动 现场。中新社记者 张浪 摄 图为人形 机器人写春联。中新社记者 张浪 摄 图为活动 图为活动 学生写春联。中新社记者 张浪 摄 者展示人形机器人书写的春联。中新社记者 张浪 摄 图为表演 现场。中新社记者 张浪 摄 学生写春联。中新社记者 张浪 摄 现场。中新社记者 张浪 摄 图为中小 图为活动 图为中小 ...

经济观察网据人民财讯消息，上海市人民政府主题记者会2月7日举行。上海市经济信息化委主任汤文侃 表示，"十五五"时期上海将累计建成500家先进智能工厂，工业机器人应用密度达600台每万人，创建 200家市级以上绿色制造企业。瞄准智能终端、商业航天、低空经济等新赛道，再造万亿级产业新增 量。 ...

北职大机电工程学院院长黄敦华介绍，定向班为学生配备校内外"双导师"，动手实践课时占比超过 60%。在这里，学生能接触到具身智能机器人场景应用中真实的项目案例，学到最接地气的技术技能。 本报讯（记者何蕊）本市首个校企共建的具身智能机器人定向班，将在今年春季学期开班。昨天，北京 科技职业大学和京东集团签约，启动具身智能机器人领域专业建设、课程开发、实习就业等合作。企业 将把具身智能机器人4S店"搬进"大学校园，引入机器人售后维修服务产业，定向培养高端技能人才，满 足产业高速发展新需求。 ...

本报讯（记者袁璐）昨天，随着最后一根钢梁平稳就位，中关村数字经济总部园主体结构全面封顶。这 座总建筑面积达40.79万平方米的海淀最大在施TOD产业园区，从去年10月底冲出"正负零"到如今封 顶，仅用了3个月。 快速建造的背后，有一支特殊的"机器人施工队"在发力。记者在封顶现场看到，多种智能机器人正在工 地投用。北京建工六建该项目总工郑立金指着不远处的放线机器人说道。这款由项目团队自主研发的机 器人，特别适用于室外土方、基坑放线等开阔场景，只需导入图纸，就能自主完成高精度放线，将传统 4人小组的工作缩减为1人操作，效率提升一倍。 ...

中新网济南2月7日电 题：机器人护航春运 "人机协同"为水下铁路桥梁"体检" 据悉，这支队伍近年足迹遍布全国，完成400多座铁路桥梁检测，行程超10万公里。(完) 来源：中国新闻网 桥梁检测队伍正在检测钱塘江大桥。中国铁路济南局集团公司铁道战备舟桥处供图 记者 孙婷婷 当一列列满载归心旅客的列车呼啸着经过新兖线京杭运河特大桥，桥下深水之中，一场无声的"体检"正 同步展开。 2月7日，一台水下检测机器人缓缓潜向桥梁墩台后，背负着60斤装备的潜水员高佳旺紧随其后，"人机 协同"开始为桥梁进行精密扫描。 高佳旺是中国铁路济南局集团公司铁道战备舟桥处潜水队成员。此前，潜水员需身负逾60斤装备，在能 见度近乎为零的冰冷河水中，完全依靠触觉为桥墩"问诊"。 铁路桥梁水下检测被称为"水下战场"，水文环境复杂、潜在风险多样，要求潜水员不仅要精通潜水，还 需了解桥梁结构、熟练操作专业设备。 "以前人工排查一处桥墩要数小时，现在机器人先'探路'我们复核，效率翻倍。"高佳旺说，该机器人兼 具"浮游"与"吸附爬行"模式，可在江河流中悬停，通过高清摄像头和声呐快速扫描、标记隐患并实时传 输数据。 在今年的春运中，"人机协同"新模式让检 ...

人形机器人产业尚处于起步阶段，但早期优势加速了中国在研发和布局上的进展，有可能锁定其主导地 位，复制在电动汽车领域的成功策略。全球技术市场研究咨询机构Omdia的分析师苏廉节表示，中国的 早期优势来自于"政策支持、公共投资、成熟的供应链以及人工智能软硬件的进步"。 2025年全球人形机器人出货量在1.3万至1.8万台之间。中国的宇树科技去年出货超过5500台，全球最 高，另一家中国公司智元机器人以5168台排名第二。这让人联想起中国如何在电动汽车领域建立起领先 地位——早期的国家支持、产业政策催生数十家新入局者，快速拓展将它们推上全球畅销车榜前列。在 苏廉节分析师看来，"中国人形机器人供应商在设计中使用越来越多的本土零配件，这有助于提高成本 效益、保障供应链安全，并推动本土化创新和上市进度"。 美国"世界其他地方"网站2月5日文章，原题：中国正在机器人领域复制电动汽车的模式，且行之有效 中国企业控制着全球人形机器人的大部分市场份额，主导着这项重塑制造业和劳动力的前沿科技，西方 企业几乎无法与之竞争。2025年，当硅谷还在为人形机器人的未来营销策划时，中国的机器人已在忙着 出货了。去年，全球销售的人形机器人大多 ...