省人工智能人才培养教指委徐向民主任介绍教指委五大创新工作计划，广东工业大学、中山大学分享人 才培养经验做法，腾讯、优必选、粤科金融集团等企业代表分别介绍了自身行业领域产业动态、技术应 用、人才诉求等。会议现场设置了人工智能与机器人领域的创新产品展示，校企双方围绕展品和人才培 养合作进行深入交流。 南都讯 记者孙小鹏 为全面落实省委、省政府关于聚力发展人工智能和机器人两大战略性产业，打造高 技术含量、高增长潜力、大体量规模产业新支柱的具体部署，日前，广东省教育厅在中山大学成功举 办"产教双向赋能，智启融合发展"——高校助力人工智能和机器人产业发展大会。会议旨在充分发挥全 省高校科技创新和人才培养优势，推动教育链、人才链与产业链、创新链深度融合，为广东省打造全球 人工智能与机器人产业创新高地提供坚实的智力支撑和人才保障。 会上，省工业和信息化厅总经济师邹勇兵作《广东省人工智能与机器人产业发展情况报告》，系统剖析 产业发展现状、问题及支持政策，明确提出打造全球影响力产业创新发展新高地目标，亟须通过深化产 教融合破解发展瓶颈。东莞松山湖国际机器人产业基地创始人李泽湘教授向大会介绍了以"端到端、高 质量、高效率"为主要 ...

Core Viewpoint - The research from the Cambridge University team presents a novel approach to robotic tactile sensing using a gelatin-based hydrogel, which functions as a giant sensor with 1.72 million information channels, eliminating the need for multiple sensors [1][19]. Group 1: Research Methodology - The hydrogel is designed as a continuous sensing field, with only 32 electrodes placed at the edges, allowing for the monitoring of 863,040 different current paths through electrical impedance tomography (EIT) [2][6]. - The system can simultaneously identify six types of stimuli: human touch, contact with conductive objects, pressure from insulating objects, localized heating, damage, and multi-point touch [2][6]. Group 2: Performance and Accuracy - A complete robotic hand made of hydrogel was created, achieving a positioning accuracy of 25mm even at the fingertips, which are far from the electrodes [3][19]. - The system demonstrated an average positioning error of 24.7mm during tests with 1,080 random touch points, and using only the ten optimal configurations, the error could be controlled within 40mm [18]. Group 3: Data Processing and Analysis - The research team utilized a data-driven configuration selection method, achieving a frame rate of 33kHz when monitoring specific configurations, compared to 0.02Hz when monitoring all channels [10][17]. - Principal Component Analysis (PCA) was employed to generate "fingerprint maps" for each stimulus type, indicating distinct electrical field change patterns for different stimuli [12]. Group 4: Material and Design Innovations - The study highlights the limitations of traditional electronic skin, such as complex manufacturing and susceptibility to damage, and proposes a simpler design paradigm that enhances robustness through information redundancy [19][20]. - The manufacturing process for the hydrogel hand involved 3D printing and molding techniques, demonstrating the feasibility of creating complex 3D shapes for tactile sensing [14].

作者｜黄楠 编辑｜袁斯来 硬氪获悉，北京智在无界科技有限公司（以下简称「智在无界」，英文为BeingBeyond）近日完成数千万元融资，联想 之星领投，智谱Z基金、燕缘创投、彬复资本跟投，势能资本担任独家财务顾问。资金将用于加大核心技术研发投 入，加速现有模型迭代与产业化验证，以持续提升技术壁垒与产品竞争力。 「智在无界」成立于2025年1月，专注人形机器人通用大模型的研发与应用。创始人卢宗青是北京大学计算机学院长聘 副教授，曾任智源研究院多模态交互研究中心负责人，负责过首个国家自然科学基金委原创探索计划通用智能体项 目；多位核心成员均来自智源研究院，在强化学习、计算机视觉、机器人控制和多模态等领域有丰富的技术研发积累 和应用落地经验。 当前，数据规模与泛化能力是制约具身大脑性能提升的核心矛盾。一方面，具身智能机器人要实现高度拟人化的行动 与决策能力，依赖海量且多样化的数据进行深度训练。这些数据覆盖了日常琐碎操作、复杂环境交互等各类场景，数 据规模呈指数级攀升态势。然而，数据采集过程仍面临技术、资源等多重门槛，依赖大量人力且难度大，存储成本更 是随着数据量激增而迅速攀升。 另一方面，即便有海量的数据支持，机器 ...

据 Mordor Intelligence 数据， 2021-2024年全球割草机器人市场规模从15亿增至 26亿美元，行业复合 增长20%。 Fortune Business Insights进一步预测，到2028年割草机器人市场规模将增长至40.4亿美 元。 草坪作为欧美家庭重要的户外空间，不仅是彰显社会地位的标志性元素，更是社区社交生活的重要载体。据统 计，目前全球私人花园数量已达2.5亿个，其中欧美地区占据绝对主导地位—— 美国约1亿个，欧洲超过 8000万个，两大区域合计占比高达全球总量的72%。 近年来，随着欧美各国环保法规日趋严格和社区景观标准的提升，家庭草坪养护需求呈现持续增长态势。然而 值得注意的是， 智能割草机器人在这一广阔市场的渗透率仍处于较低水平： 北美市场渗透率尚不足6%，即 便是智能化程度较高的欧洲市场，渗透率也仅维持在10%-30%区间。这一现状与快速增长的草坪养护需求 形成了鲜明对比，折射出巨大的市场发展潜力。 ▍三重领先优势，重新 定义割草机器人感知新标准 面对这一 快速成长的市场， 九号、库犸、科沃斯、石头等知名机器人厂商纷纷加大投入， 奋力角逐智能割 草机器人市场，但 在 ...

Group 1 - The article highlights the active preparation for intelligent elderly care service robot pilot projects in Hunan, with a focus on addressing user needs and product upgrades as the elderly population exceeds 22% by the end of 2023 [1] - The AI tomato-picking robots in Qitang Town, Banan District, demonstrate advancements in agricultural technology, with over 30,000 acres of tomato cultivation generating an annual output value exceeding 400 million [5] - Meta's release of the V-JEPA 2 world model, which utilizes over 1 million hours of internet video and image data for training, marks a significant advancement in robotics, enhancing action prediction and modeling capabilities [8] Group 2 - Morgan Stanley's report indicates that China leads the robotics competition over the U.S., attributing this to advantages in rare earth resources, policy incentives, and a robust education system with over 5 million students in related fields [9] - The second Xiong'an International Robot Competition showcased 182 entries, with 83 reaching the finals, emphasizing the focus on industrial, service, and special robots, and highlighting several innovative technologies [13]

Group 1 - China Bank Shanxi Branch has developed a comprehensive financial innovation service system to support technology-based enterprises, aiming to inject continuous momentum into the innovation development of Shanxi's tech sector, with over 100 billion yuan in loans projected by May 2025 [1] - Shanxi Jiashida Robot Technology Co., Ltd. transformed from a small R&D team into a national-level specialized "little giant" enterprise, receiving 3 million yuan in credit loans during critical development phases, with credit support increasing to nearly 70 million yuan over ten years [2] - In January, China Bank announced a plan to provide no less than 1 trillion yuan in financial support for the AI industry chain over the next five years, with Shanxi Branch already providing nearly 500 million yuan to over ten AI enterprises by May 2025 [3] Group 2 - Taiyuan Fulairida Logistics Equipment Technology Co., Ltd. received 50 million yuan in funding from China Bank Shanxi Branch, enabling the development of advanced smart warehousing systems and equipment, enhancing operational efficiency and promoting low-carbon storage solutions [4] - Datong has become a significant data center hub, with over 204,000 large-scale data centers and more than 3 million servers supporting high-tech industries, showcasing the city's transformation from coal mining to data processing [5] - China Bank Shanxi Branch tailored financial products for Datong's enterprises facing funding challenges, simplifying approval processes and reducing financing costs, with nearly 500 million yuan in loans issued to a specific data service company by May 2025 [6][7]

Core Insights - New Wave Holdings Corp. has made a strategic investment of US$150,000 into Cartwheel Robotics Inc. to support the development of advanced humanoid products and solutions [1][5] - Cartwheel Robotics, founded in 2021, specializes in advanced robotics products and engineering services, with a focus on solving complex problems [2] - The company is led by Scott LaValley, who has extensive experience in robotics from notable companies such as Disney, Google, and Boston Dynamics [3] Company Overview - Cartwheel Robotics engages in various functions within the physical AI and humanoids industry, including prototyping and delivering production-ready designs [4] - The investment from New Wave is aimed at accelerating R&D initiatives and commercialization plans for Cartwheel Robotics [5] - New Wave Holdings Corp. focuses on supporting innovative companies in sectors like E-Sports, Artificial Intelligence, Blockchain, and Web3 [7]

Core Viewpoint - The Chinese government is promoting the development and application of intelligent elderly care service robots to address the challenges posed by an aging population and the increasing demand for elderly care services [1][2][3] Group 1: Government Initiatives - The Ministry of Industry and Information Technology and the Ministry of Civil Affairs have launched a pilot program for intelligent elderly care service robots from 2025 to 2027 [1] - The focus of the initiative includes improving the quality of life for the elderly, alleviating family caregiving pressure, and addressing the labor shortage in institutional and community elderly care services [2][3] Group 2: Market Demand and Demographics - By the end of 2024, the population aged 60 and above in China is projected to reach 310.31 million, with those aged 65 and above accounting for 15.6% of the total population [3] - Approximately 35 million elderly individuals are currently disabled, representing 11.6% of the elderly population, with projections suggesting this number could rise to around 58 million by 2050 [3] Group 3: Industry Growth and Trends - As of now, there are 577,900 existing companies related to intelligent robots in China, with a significant increase in registration over the past decade [5][6] - In 2023, the registration of intelligent robot companies surpassed 100,000, marking a growth of 48.61% in the first five months of the year [6][5] Group 4: Technological Advancements - Intelligent elderly care robots are being developed to perform tasks such as feeding, medication delivery, and emotional companionship, with a focus on achieving near-human caregiving capabilities [2][10] - Innovations include automated toilet care robots and bathing robots that enhance the dignity and comfort of elderly users while improving care efficiency [12][14] Group 5: Industry Participation and Events - Major companies such as SAIC Toyota, IKEA, Huawei, and Haier are entering the silver economy, showcasing their products at the Shanghai International Elderly Care and Rehabilitation Medical Expo [10][19] - The Shanghai Expo has seen a record number of exhibitors, indicating a growing interest and investment in the elderly care robotics sector [19]

2025智能机器人关键技术大会 大会接受各项商务合作，欢迎垂询具体合作形式。 张老师：13305713232（微信同号） 谷老师：18355423366（微信同号） 详细会议介绍参看往期文章： （点击蓝字跳转） 报名开启｜顶流期刊征+2025智能机器人关键技术大会盛会将至！ 展览展示｜抢位2025智能机器人关键技术大会！高曝光商务合作虚位以待，共赴解锁新机遇 征文通知｜投稿2025智能机器人关键技术大会，年底可见刊 图源｜豆包AI文生图 此外，人形机器人涉及功能安全、数据隐私、就业冲击等多重风险，可借鉴《生成式人工智能服务管理暂行办 法》的"发展与安全并重"原则，构建包容审慎的治理框架，完善分级分类监管。例如对医疗、安防等高风险场景 实施强制认证，对家庭服务机器人强化隐私保护设计。机器人的"拟人化"可能引发情感依赖或伦理困惑，这也需 要通过公众参与、跨学科对话确立技术边界，培育社会共识与伦理自觉。例如，限制过度拟人化设计，明确机器 人的"工具属性"；在研发中嵌入伦理算法，防止自主决策偏离人类价值观。 这场机器人格斗赛，既是一场技术秀，又是一面镜子，照见了人类将科幻变为现实的雄心。以扎实的技术探索为 根基，以负 ...

2025智能机器人关键技术大会 清华大学交叉信息研究院助理教授高阳 说，机器人数据采集成本高（需遥操作+物理交互），且难以用语言描述 （如游泳），形成"无智能-少机器人-少数据"的恶性循环，可利用互联网视频数据，通过跟踪视频中物体运动预 训练模型，再迁移到机器人遥操作数据微调，结合强化学习，解决仿真与现实差距。 在具身智能的通用泛化能力构建方面， 智源具身多模态大模型中心负责人、北京大学计算机学院助理教授仉尚航 认为，构建具身快慢系统是具身智能从"单一任务/本体"迈向"通用泛化"的关键路径，分别以"大脑（推理）+小脑 （控制）"的分层快慢系统和VLA端到端快慢系统两条技术路线，解决跨本体（如机械臂、人形机器人）与场景的 泛化性问题。 大会接受各项商务合作，欢迎垂询具体合作形式。 张老师：13305713232（微信同号） 谷老师：18355423366（微信同号） 详细会议介绍参看往期文章： （点击蓝字跳转） 报名开启｜顶流期刊征+2025智能机器人关键技术大会盛会将至！ 展览展示｜抢位2025智能机器人关键技术大会！高曝光商务合作虚位以待，共赴解锁新机遇 征文通知｜投稿2025智能机器人关键技术大会，年底可 ...