Core Viewpoint - The humanoid robot industry is transitioning from a speculative phase to a critical commercialization stage, with a focus on practical applications and sales metrics rather than just market dreams [1][4][11]. Group 1: Industry Development - The humanoid robot industry has evolved from skepticism to market enthusiasm, and is now entering a pivotal phase where commercial orders are essential for sustainability [1][4]. - By 2025, a specific humanoid robot company is expected to ship 1,000 units, with the first four months of the year already achieving four times the total shipments of the previous year [7]. - The primary market is shifting its focus from market dreams to sales metrics, indicating a significant change in how the industry is evaluated [11]. Group 2: Commercialization Challenges - The industry must overcome three main bottlenecks: insufficient data accumulation in software, unresolved hardware issues, and the need for valuable application scenarios [16][17]. - Unlike the automotive industry, which had a clear path for innovation, the humanoid robot sector is still debating the necessity of replacing human labor with robots, highlighting the need for concrete application scenarios [17]. Group 3: Strategic Insights - The humanoid robot industry may experience rapid growth similar to consumer electronics, with the potential for annual sales to reach several billion units in a much shorter timeframe than the automotive sector [19]. - Companies must act quickly to seize strategic opportunities in the humanoid robot market, as the entry barriers are lower compared to the automotive industry [19][30]. - Successful companies in emerging industries often capture high-value positions during brief windows of opportunity, as demonstrated by several market leaders with relatively low pre-IPO revenues [21]. Group 4: Investment Strategies - Companies are encouraged to adopt a dual approach of capital and business strategies to effectively position themselves in the humanoid robot market [25]. - Key capabilities for companies include conducting extensive research, making swift decisions, identifying critical partnerships, and optimizing resource allocation [30]. - The industry is characterized by a dynamic interplay between industrial and financial capital, with the former needing to leverage its advantages during periods of market volatility [33][35].

Core Viewpoint - The article emphasizes the strategic importance of data in the advancement of robotics towards autonomy and generalization, highlighting the launch of the "White Tiger" dataset to address data bottlenecks in the industry [1][2][35]. Group 1: Dataset Characteristics - The "White Tiger" dataset is the world's first heterogeneous robot dataset exceeding one million samples, collected from real-world applications and covering various humanoid and robotic platforms [1][2]. - The dataset has received official certification from the China Academy of Information and Communications Technology, marking it as the first authoritative certified dataset in the field of embodied intelligence in China [2]. - It includes data from multiple robot platforms, such as the Qianlong, PortaGrip, and others, with specific proportions of data representation from each platform [7]. Group 2: Data Collection and Quality - The dataset was created to overcome the challenges of isolated data collection and unstandardized formats, establishing a unified data collection system adaptable to various robotic platforms [5]. - A comprehensive data quality control system was implemented to ensure high standards in the dataset, enhancing its reliability for training and evaluation [28]. Group 3: Application Scenarios and Tasks - The "White Tiger" dataset is structured around five major application scenarios, significantly improving the robots' environmental perception and cross-scenario generalization capabilities [8]. - It features a multi-dimensional task system that allows for the structured breakdown of tasks into atomic skills, facilitating the training of robots across various scenarios [12]. Group 4: Interaction and Skill Development - The dataset includes a diverse range of interaction targets, covering over a hundred types of real objects, which enhances the robots' manipulation learning boundaries [14]. - It systematically labels over a hundred atomic skills, providing essential operational representations for complex task behavior understanding and generation [26]. Group 5: Temporal Data and Behavior Modeling - The dataset records task execution processes across short, medium, and long time scales, supporting hierarchical behavior modeling from low-level action control to high-level task planning [16]. Group 6: Future Implications - The "White Tiger" dataset aims to establish a robust foundation for general intelligence in robotics, addressing key challenges in data volume, engineering standards, application breadth, and intelligence depth [35]. - The initiative invites collaboration across industry, academia, and research sectors to foster an open ecosystem for advancing general robotics towards higher intelligence and stronger generalization capabilities [35].

超声扫描是医学影像中的轻量级选手 ——无辐射、低成本、便携且能实时成像，常被用于软组织诊断。但传 统手持超声扫描有个大问题：操作者的技术水平和经验直接影响成像质量，特别是超声探头与皮肤之间的接触 力控制至关重要。 接触力太小，声耦合不足，看不清组织；接触力太大，又会改变组织形状，可能导致误诊；而接触力不稳定则 会让图像持续变化。即使是经验丰富的医生，也很难在整个扫描过程中精确维持一个恒定的接触力。 针对这一问题，澳门大学的研究团队提出了一种创新解决方案 ——基于混合主动-被动力控制的超声扫描机 器人系统。他们在《Cyborg and Bionic Systems》期刊发表的最新研究表明，该系统不仅能在变化的表面 上维持恒定接触力，还能适应呼吸等生理活动引起的震动干扰，使超声图像保持高度稳定性和重复性。 ▍ 混合主动 -被动力控制：取二者之长 传统的力控制系统主要有两种方式：主动力控制和被动力控制。前者通过传感器反馈调整执行器位置来控制力 大小，后者则依靠特殊设计的弹簧或柔性机构来提供恒定力。 主动力控制方法虽然灵活，但由于需要反馈控制，在面对患者突然震动时可能存在超调问题，带来安全隐患。 被动力控制虽然反应快 ...

人形机器人什么时候能像人一样劳动？当全球 60余支精英战队齐聚上海， 告别炫技式的 "走跑跳"表演，转 而挑战真实工作技能 时，这个问题的轮廓正在变得清晰。与单纯追求视觉表现力不同，中国人形机器人产业 正以 "好用、管用" 为导向，加速商业化落地进程。 从长三角到珠三角，从京津冀到成渝地区，中国各大城市正展开史无前例的具身智能争夺战。 5月29日， 2025张江具身智能开发者大会暨国际人形机器人技能大赛拉开帷幕，以 "劳动最光荣" 为主题的 2025国际 人形机器人技能大赛 吸引各大 主流媒体争相报道，社交平台上相关话题持续升温，现场数千观众慕名而来， 见证这场前所未有的技术角逐。 作为全球首个打造 技能链条最完整、技术最全面、行业最前沿、应用领域最广的专业大赛 ，本届赛事以 "开 源、开放、开创"构筑了技术与应用的交汇平台： 开源 ， 通过技能大赛平台汇聚全球参赛团队共享技术成果 推动协同创新； 开放 ，设置工业制造、家庭服务、商业物流等多元化应用场景，吸引产业链上下游深度参 与； 开创 ，将 28个 高难度实用场景搬上国际舞台，联合需求方共同推动场景落地驱动产业发展并吸引优质 资源落地张江。 ▍ 具身智能 ...

Core Viewpoint - The event "Soft and Hard Collaboration, Intelligent Integration" focused on the development of the embodied intelligence industry, highlighting the importance of collaboration among various stakeholders to drive technological breakthroughs and ecosystem building [1][4][20]. Group 1: Policy and Industry Trends - The inclusion of "embodied intelligence" in government work reports signifies a strategic push for digital innovation and transformation in Shanghai, particularly in the Baoshan District, which has developed a trillion-yuan industrial cluster for robotics and intelligent manufacturing [4][20]. - The collaboration between BOE Technology Group and Shanghai Real Estate Minhong Group aims to create a digital economy demonstration base, facilitating the landing of key technologies in the embodied intelligence sector [4][20]. Group 2: Company Innovations and Applications - Fourier's two main business lines include medical rehabilitation products and general humanoid robots, with successful commercialization and partnerships enhancing application in various sectors [11]. - Intelligent technology company Lingxin Qiaoshou focuses on bionic humanoid robots, utilizing advanced technologies to create versatile products for industrial and research applications [15]. - Weitai Robotics emphasizes breakthroughs in tactile perception and hand-eye coordination, showcasing new sensors that significantly enhance robotic capabilities in precision tasks [19]. Group 3: Ecosystem Development and Collaboration - The event featured a demand release and supply-demand matching session, facilitating connections between industry representatives and innovative project teams to promote collaboration [19]. - The future of embodied intelligence is expected to be driven by the synergy of technology, application scenarios, and capital, fostering a new wave of industrial transformation [20].

Core Viewpoint - The 2025 Zhangjiang Embodied Intelligence Developer Conference showcased advancements in humanoid robots and embodied intelligence technologies, emphasizing open-source innovation and collaboration among over 200 companies and 5,000 industry experts [1][3]. Group 1: Conference Overview - The conference featured a three-part model: summit, competition, and exhibition, highlighting the latest developments in humanoid robots and embodied intelligence [1]. - A total of 87 robot companies participated, with four core display areas focusing on humanoid robot bodies, embodied intelligence, industry chain components, and developer ecosystems [3]. Group 2: Humanoid Robot Innovations - The humanoid robot body exhibition included 21 companies, showcasing advancements in structural design, motion control, and human-robot interaction technologies [1]. - The Qinglong humanoid robot, developed by the National Local Co-built Humanoid Robot Innovation Center, is the world's first full-size open-source humanoid robot, featuring a height of 185 cm, weight of 85 kg, and 43 degrees of freedom [6]. - Fourier's GR-2 humanoid robot, standing at 1.75 meters and weighing 63 kg, includes 53 degrees of freedom and a self-developed 12-degree-of-freedom dexterous hand for precise grasping [9][11]. Group 3: Embodied Intelligence Technologies - The embodied intelligence exhibition featured 12 companies, focusing on large models for environmental perception, behavior control, and human-robot interaction [3]. - Companies like Guanglun Intelligent and Songying Technology showcased advanced motion control algorithms and hardware systems for humanoid robot development [3][78]. Group 4: Industry Chain Components - The industry chain exhibition included 34 companies, focusing on key components such as servo motors, controllers, and dexterous hands, with some domestic components surpassing their international counterparts [3]. - Companies like Aoyi and Yiyou Technology are recognized for their contributions to core components in the humanoid robot industry [3]. Group 5: Developer Ecosystem - The developer ecosystem area featured 20 companies, including Baidu Intelligent Cloud and Shanghai Jiao Tong University, supporting data collection, model training, and deep learning for humanoid robot applications [3]. - This area aims to facilitate the integration of emerging technologies into traditional sectors, enhancing the practical application of humanoid robots [3]. Group 6: Future Prospects - The successful hosting of the conference demonstrated the application achievements of humanoid robot technology across various scenarios and the overall development trend of the industry [3].

Core Viewpoint - UnixAI, a developer of embodied intelligent robots, has successfully completed a Series A+ financing round, raising several hundred million yuan to accelerate the development of multimodal embodied intelligent models and general-purpose robotic bodies for various commercial services and consumer scenarios [1][13]. Group 1: Company Overview - UnixAI was established in April 2024 with the goal of "making intelligent robots serve every household" [1]. - The company focuses on building a high-efficiency, generalizable imitation learning framework and a low-cost, modular stable body [1]. - UnixAI aims to achieve a threefold flywheel effect through the integration of products, supply chains, and data, promoting the widespread adoption of embodied intelligence [1]. Group 2: Core Team - The founder and CEO, Dr. Yang Fengyu, is a Yale University graduate and the first author of the world's first tactile fusion multimodal model, UniTouch [3]. - Dr. Yang has published 15 papers related to robot vision and touch, and has developed the first multimodal tactile perception system, which supports complex operations under zero-shot training conditions [3]. - Chief Scientist Professor Wang Hesheng is a leading expert in robotics, with over 50 related patents and a theory that addresses precise control of robots in unstructured environments [5]. Group 3: Technological Innovations - UnixAI has developed a generalizable imitation learning framework called "UniFlex," which decouples perception and action models to enable efficient imitation learning with minimal demonstration data [7]. - The company has also created the world's first multimodal perception model, "UniTouch," which can perform delicate tasks by generating tactile feedback from visual input [9]. - UnixAI has built a reasoning architecture, "UniCortex," that supports long-range task planning in complex environments and among multiple agents [9]. Group 4: Product Development and Market Entry - Within four months of its establishment, UnixAI successfully developed two products: the wheeled dual-arm robot Wanda and the bipedal humanoid robot Martian, which were showcased at the 2024 World Robot Conference [11]. - Wanda is designed for household tasks such as laundry and burger-making, while Martian features performances like dancing and boxing [11]. - The company has initiated pre-sales for Wanda 2.0, priced at 88,000 yuan, and has secured several million yuan in orders since its product launch, demonstrating strong commercial viability [13].

Core Viewpoint - The 2025 Zhangjiang Embodied Intelligence Developer Conference and International Humanoid Robot Skills Competition successfully took place, focusing on the theme "Embodied Intelligence, the Future is Here," showcasing over 200 key enterprises and 5,000 industry experts, emphasizing the acceleration of embodied intelligence technology into real-life applications [1]. Group 1: Event Overview - The event featured a new model combining a summit, competition, and exhibition, aimed at fostering technology implementation and ecosystem development [1]. - The conference was guided by the Shanghai Municipal Economic and Information Commission and supported by various institutions, including Shanghai Jiao Tong University and Agricultural Bank of China Shanghai Branch [1]. Group 2: Government Support and Policies - Shanghai's government officials highlighted the region's efforts to create an industrial cluster for embodied intelligence and humanoid robots, with policies to enhance supply chain cooperation and application across various sectors [2]. - The release of 16 policy measures aimed at supporting the high-quality development of the embodied intelligence industry was announced, focusing on innovation, application promotion, and talent support [2]. Group 3: Competition Highlights - The International Humanoid Robot Skills Competition is recognized as the most comprehensive and influential professional competition globally, showcasing the highest technical standards in the field [5]. - The competition includes five tracks covering 28 challenging scenarios, attracting over 60 top teams from various institutions [5]. Group 4: Application Challenges - The humanoid robot application scenario challenge demonstrated real-world application capabilities, with teams completing tasks in industrial and domestic settings [6]. - The embodied intelligence application scenario challenge showcased the technology's versatility across various environments, including industrial and commercial applications [6]. Group 5: Innovation and Collaboration - The establishment of the Embodied Intelligence Data Alliance aims to enhance data sharing and collaboration across sectors such as services, transportation, and healthcare [10]. - A partnership plan for a universal embodied intelligence development platform was launched, involving multiple organizations and universities to promote standardization and collaboration [10]. Group 6: Developer Engagement - The event included nine professional developer forums focusing on key areas such as data collection and application scenarios, promoting the development of humanoid robot data sets [11]. - Various activities, including signing ceremonies and workshops, were organized to enhance interaction among developers and showcase Shanghai's welcoming attitude towards global talent [13]. Group 7: Exhibition and Industry Growth - The humanoid robot innovation product exhibition attracted over 80 companies, covering 3,000 square meters and showcasing the latest technological achievements [15]. - Shanghai is committed to expanding the embodied intelligence industry, with initiatives like the establishment of a heterogeneous humanoid robot training ground and the launch of a super factory [15]. Group 8: Future Development - The event emphasized the importance of driving industry development through practical applications and the establishment of a supportive environment for enterprises [16]. - Shanghai aims to continue fostering innovation in the embodied intelligence and humanoid robot sectors, contributing to the advancement of autonomous technology and industrialization [16].

1、Circus SE收购代理AI公司FullyAI 近日，食品服务行业机器人开发商Circus SE宣布完成了对代理人工智能公司FullyAI的收购。该公司表 示，此举是其在机器人和软件领域构建统一智能层的使命中的一个重要里程碑。据悉，Circus SE公司研 发出的CA-1机器人应用了嵌入式人工智能 (embodied AI)，而此次收购完成后，公司表示FullyAI将的软 件将处理来自 CA-1 自动烹饪机器人的数十亿个数据点，从而解锁一个能够实时连接每个流程的自我改进 智能层。。作为收购的一部分，FullyAI 的工程团队和管理层已加入 Circus SE。其中，Niklas von Weihe 曾与Meta的高级 AI 团队密切合作，此次加入 Circus SE 担任 AI 总监。 2、"荆楚"人形机器人今年在汉量产 近日，华中科技大学丁汉教授团队、武汉产业创新发展研究院、武昌区属国有科技投资平台首义科创投集 团和逸飞激光共同签署人形机器人产业化项目合作协议。此次合作宣布华中科技大学丁汉教授团队研制 的"荆楚"人形机器人签约落户武昌。早在今年2月，身高170厘米、拥有一双灵巧手、可站立可行走、能模 ...

盈连 力控砂带机 由电动机、砂带传动系统、力位补偿器组成 ， 是一种用于表面处理的高效工业设备， 与传 统砂带机相比，配有完全自主研发的力控系统 ， 可以保证全程打磨力一致，自动补偿机器人轨迹和工件误 差。 根据力控位置不同分为两类：整体式浮动砂带机、模块式浮动砂带机。 5 安全便捷：配备防护罩、急停装置及集尘接口， 后期维护成本低， 保障操作安全并改善工作环境。 二 产品参数 广泛应用于金属加工、木工制造、模具修整等领域。其核心功能是通过砂带的高速旋转对工件进行磨削、抛 光、去毛刺及表面精加工， 使其表面更加平整、光滑和美观，满足不同工业领域对工件表面质量的要求。 尤 其适合处理大型或重型工件，如机械部件、汽车零件、家具构件等。 一 核心优势 1 高效节能：最高5.5kw的磨削效率，比传统立式砂带机打磨效率提高50%； 2 智能力控技术 : 高精度的力控制系统，可保证接触力精确与稳定，同时还可根据工艺自动补偿砂 带耗材的 磨损； 3 高稳定性： 设备整体重量 100.0kg 以上， 落地式结构提供强固支撑，减少振动，确保加工一致性 ， 可固 定于工装台或地面安装 4 操作灵活： 打磨轮及耗材可以非常方便拆卸 ...