Investment Rating - The report maintains a "Buy" rating for the mechanical industry [1] Core Viewpoints - The humanoid robot sector is experiencing significant advancements, with companies like Unitree and Zhiwei Robotics making strides in product development and partnerships, indicating a robust growth trajectory for the industry [4][5][6] - The liquid cooling technology is becoming increasingly essential due to the rising power consumption of data center servers, with a projected increase in adoption as standards are established and high-power chips become more common [7][8] - The engineering machinery sector is expected to benefit from increased infrastructure investment, particularly with major projects like the Yarlung Zangbo River hydropower project, which will drive demand for machinery [13][14] Summary by Relevant Sections Humanoid Robots - Unitree Technology announced a new humanoid robot with enhanced flexibility, featuring 31 joints, a 19% increase from its previous model [4] - The humanoid robot industry is poised for a breakthrough in 2025, with mass production expected to address data scarcity issues and enhance practical applications [6] Liquid Cooling - The establishment of national standards for immersion liquid cooling power supplies is underway, which will guide the design and application of liquid cooling systems in data centers [7] - The shift towards liquid cooling is driven by the limitations of traditional air cooling systems in managing the heat generated by high-power chips [8] Engineering Machinery - The engineering machinery sector is witnessing a recovery in domestic sales, with excavator sales in July 2025 showing a year-on-year increase of 25.2% [14] - Major infrastructure projects, such as the Yarlung Zangbo River hydropower project, are expected to significantly boost demand for engineering machinery [13] Forklifts - Forklift sales have shown strong growth, with a total of 118,605 units sold in July 2025, reflecting a 14.4% year-on-year increase [15] - The rise of autonomous forklifts is anticipated to further enhance market growth, with a projected 39.3% increase in sales by 2025 [15] Rail Transit Equipment - The rail transit equipment sector is benefiting from increased investment in railway infrastructure, with significant growth in orders for high-speed train maintenance [16] Semiconductor Equipment - The push for domestic semiconductor equipment manufacturing is accelerating due to geopolitical tensions and trade restrictions, with a focus on critical equipment like photolithography machines [17][18] New Energy Equipment - The solid-state battery sector is expected to see rapid growth, with new equipment needs arising from the transition to solid-state production lines [21] - The market for photovoltaic equipment is also anticipated to recover as competition stabilizes and prices begin to rise [21]

（原标题：工业质检迎来具身智能时代，国产化进程加速） 21世纪经济报道记者曹恩惠、实习生朱泓慧 8月19日晚间，一场时长不过40分钟的新品发布会，却在国内工业质检领域留下浓墨重彩的一笔——AOI（自动光学检测）专有设备制造商中科慧 远发布了首款工业具身质检机器人CASIVIBOT。 在中科慧远总经理张武杰看来，这是"工业质检范式的转变"。 张武杰介绍，与市面上多数依然停留在单一视觉检测或机械臂分拣的尝试不同，CASIVIBOT是一个能模仿人类质检员动作、能理解缺陷本质、还 能与群体协同进化的具身智能体。具体而言，该机器人依托"手-眼-脑"协同的技术架构，实现工业产线的灵活操作——"眼"由三组相机组成的多光 谱感知系统，可进行大视野扫描与微米级精准检测，支持不同材质、复杂曲面及高反光工件的识别需求；"手"是机械臂与可切换的灵巧夹具组 合，模拟人类手臂的灵活动作，并可通过轨迹规划算法自主避障和调节操作路径；"脑"则是基于中科慧远在工业质检场景积累的工程检验经验， 自主研发"慧脑"AI平台，具有垂直行业大模型及百万级精标注缺陷样本数据库，具备行业内类间小样本迁移能力，可通过语言描述引导识别，实 现多模态融合检测。 ...

经观汽车 8月19日，国内AOI（自动光学检测）设备制造商中科慧远发布了其首款工业具身质检机器人 CASIVIBOT。这是工业质检从单一视觉向全面智能化升级的一个开端。 "时代正在呼唤一种新的质检范式，它必须打破传统AOI设备的边界，能够适配复杂多变的工业场景， 又具备快速导入、灵活部署、持续进化的能力。我们相信，工业质检的未来，将由具身智能机器人打 开。"中科慧远总经理张武杰表示。 中科慧远成立至今已有9年时间，在过去主要向工业产业链客户供应传统AOI设备，至今累计出货超过 千台，并积累了大量的经验和数据，所服务的行业覆盖3C、传统家居、汽车、半导体等行业。 不过，随着多个行业产业链向智能化方向升级，制造场景碎片化、产品形态多样化、工艺流程复杂化都 成为制约传统AOI设备发挥其功能的重要因素，尤其是在应对一些柔性化场景和非标准的瑕疵时，AOI 设备的固定检测边界、长周期导入的特性，让质检需求与检测设备的矛盾愈加突出。当前，"最后一公 里"质检环节还需高度依赖人工。 中科慧远发布的工业具身质检机器人CASIVIBOT，正是一款用来解决工业质检痛点的产品。与市面上 多数依然停留在单一视觉检测或机械臂分拣的尝试不 ...

北京商报讯(记者魏蔚)8月19日，中科慧远发布工业具身质检机器人CASIVIBOT。CASIVIBOT采用具身 设计，依托"手-眼-脑"协同的技术架构，实现工业产线的灵活操作。"眼"由三组相机组成的多光谱感知 系统，可进行大视野扫描与微米级精准检测，支持不同材质、复杂曲面及高反光工件的识别需 求。"手"是机械臂与可切换的灵巧夹具组合，模拟人类手臂的灵活动作，并可通过轨迹规划算法自主避 障和调节操作路径。"脑"则是基于中科慧远在工业质检场景积累的工程检验，自主研发"慧脑"AI平台， 具有垂直行业大模型及百万级精标注缺陷样本数据库，具备行业内类间小样本迁移能力，可通过语言描 述引导识别，实现多模态融合检测。 ...