Mech-Mind Robotics
中国工业 AI 与 3D 视觉机器人引导平台
Mech-Mind 具备真实的工业 AI 部署深度和可信的技术集成;但按独角兽价格看,公开披露仍只支持继续研究。
封面要素
公司概况
Mech-Mind Robotics 是一家 2016 年成立的中国工业 AI 公司,围绕机器人引导商业化 Eye-Brain-Hand 栈。公司把 Mech-Eye 感知硬件、Mech-Vision 和 Mech-Viz 软件、Mech-DLK 深度学习工具组合起来,在工厂和仓库里自动化抓取、放置、拆垛、机床上下料、装配和检测流程;但在经济性和控制权披露上,公开信息仍远少于公开市场投资者希望看到的程度。
- 成立时间
- 2016-09-12
- 创始人
- Shao Tianlan, Fu Ao, Ding Youshuang
- 创立地点
- Beijing, China
- 总部
- Beijing, China
- 产品
- Mech-Mind 销售 Mech-Eye 工业 3D 相机和激光轮廓仪、Mech-Vision 感知软件、Mech-Viz 机器人规划软件、Mech-DLK 深度学习工具,以及集成式 Eye-Brain-Hand 机器人工作站。
- 客户
- 重点面向汽车、电子和 EV 电池制造、物流与仓储、金属或机械运营商,通常借助系统集成商和机器人品牌生态落地。
- 商业模式
- 面向工业自动化项目销售硬件、软件与服务混合部署,而不是纯独立 SaaS。
- 阶段
- late-stage private
- 融资情况
- 最新公开融资是 2025 年 8 月约 CNY500 million 的 Series E-II 轮;官方材料称累计融资已达到约 USD300 million,或超过 RMB2 billion,媒体报道则把公司置于独角兽区间,并提到可能赴香港 IPO。
执行摘要
主要优势
- 一体化 Eye-Brain-Hand 技术栈把 3D 传感、AI 软件和机器人引导工作流放进同一平台。
- 公开部署证据显示,公司已进入汽车、电子、物流和金属等工业场景,触达范围有分量。
- 创始人技术背景扎实,且多次获得头部资本和国资背景资本支持,增强可信度和扩张能力。
主要风险
- 公开收入、毛利率、烧钱速度和现金披露太薄,无法有信心按独角兽价格承保。
- 全球机器视觉龙头和平台生态竞争,可能挤压定价和渠道入口。
- 出口管制、中国工业周期和香港 IPO 时点不确定,可能压缩估值结果。
未决问题
- 经审计收入、毛利率、烧钱速度、现金余额和债务仍未披露。
- 客户集中度、续约和赢单 / 输单指标没有公开资料。
- 当前准确估值、清算条款和 2025 年后的股权结构经济性仍未验证。
目录
01公司概况
1.1 身份、布局与产品架构
应首先把 Mech-Mind 视为具身智能的赋能平台公司,而不是单一机器人 SKU 供应商。官方英文和中文公司页面都显示,公司成立于 2016 年,核心是「Eye-Brain-Hand」栈:工业 3D 相机充当眼睛,AI 软件和多模态推理充当大脑,灵巧末端执行器充当手。KrASIA 和 AsiaICT 也强化了同一叙事,并补充说,公司的商业化逻辑是把这些组件标准化,让集成商和终端用户能覆盖大量制造和物流任务,而不必为每个流程重做定制开发。运营布局也需要说清楚。中文公司页面把北京作为研发中心、上海作为销售和交付基地;中英文联系页面则显示更广的网络,覆盖德国、日本、韩国、美国,以及多个中国城市。一些目录和 IPO 传闻报道突出雄安或河北地址,但更适合把这些地址理解为法律实体或协会地址,而不是唯一运营总部。供后续章节复用的标准身份应是:一家处于后期的中国私营机器人平台公司,北京和上海在运营上重要,现场团队全球分布,并拥有标准化的「Eye-Brain-Hand」产品架构。[CO001, CO002, CO003, CO004, CO007, CO008]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 备注 |
|---|---|---|---|---|
| 成立时间 | 2016-09-12 / 2016 | 2016 | 高 | Aiqicha 与公司材料在 2016 年成立日期上相互一致。 |
| 创始团队表述 | 清华海归创始团队 | 2016 | 中 | 法律记录不能证明正式高校孵化;只能支撑清华海归团队叙事。 |
| 运营总部布局 | 北京研发,加上海销售和交付 | 2026-05-26 | 中 | 部分第三方页面突出与法律或协会主体相关的 Xiong’an 或 Hebei 地址。 |
| 商业模式 | 由 3D 视觉、AI 软件和灵巧手组成的具身智能平台 | 2026-05-26 | 高 | 应把公司视为面向集成商和终端用户的赋能栈,而不是单一机器人本体 OEM。 |
| 当前阶段 | 后期私营;公开 36Kr 资料显示 E 轮,官方英文文案仍写 Series C+ | 2026-05-26 | 中 | 公开元数据在不同来源中存在滞后,后续章节应明确保留这种不一致。 |
| 累计融资 | 公开声称 RMB 2B+;较旧英文页仍写 Series C+ 时为 USD 300M | 2026-05-26 | 中 | 金额反映不同披露时点,而不是一张已调和的股权结构总额。 |
| 最新披露轮次 | CNY 500M 融资投资方组合 | 2025-08-25 | 高 | MarketScreener 识别出投资方组合;已抓取材料中,官方公司页面没有发布匹配的条款清单。 |
| 最新披露员工数 | 600+ 名全球员工 | 2026-05-26 | 中 | 除公开公司资料外,未找到经审计或工资单支持的企业员工数。 |
| 最新披露部署规模 | 24,000+ 台,覆盖近 50 个国家和地区 | 2026-05-26 | 高 | 较旧第三方页面引用 10,000+ 台相机或 15,000+ 次安装,意味着口径更窄或披露时点更早。 |
| Fortune Global 500 客户信号 | 100+ 家客户 | 2026-05-26 | 高 | 该说法在当前公司自有页面和 Automate Show 参展商资料中反复出现。 |
| 当前估值 / 收入 / ARR | 留存来源中未公开确认 | 2026-05-26 | 低 | 不要用过时的独角兽表述或私人数据供应商占位符替代已验证披露。 |
本表把证据扎实的身份和规模标记,与仍不透明的指标分开,尤其是估值、收入或 ARR,以及完整调和后的当前融资总额。
[CO001, CO002, CO007, CO009, CO019, CO020]公司身份串起几条线:清华海归创始团队、模块化 “Eye-Brain-Hand” 技术栈、集成商赋能、全球办公室,以及后期私募融资。
[CO002, CO007, CO008, CO009, CO010, CO019]1.2 创始人、领导层与治理
公开领导层记录显示,公司由创始人驱动,技术可信度较强,但对机构尽调来说仍偏薄。Mech-Mind 官方团队页面列出 Shao Tianlan 为创始人兼 CEO,并披露其在 Tsinghua University 软件学院和 Technical University of Munich 的学习经历,这与 36Kr、KrASIA 对其工业机器人创始人与市场匹配度的报道一致。中文创业公司资料页又补充两位公开可见的联合创始人:商务侧的 Fu Ao、研发管理侧的 Ding Youshuang;官方团队页面还突出 Professor Jianwei Zhang 为创始技术顾问兼首席科学家。KrASIA 还提到 Xu Tingting 担任业务和市场副总裁,这一点有用,因为它显示创始人之外还有国际化班底。治理披露弱于领导层披露。本次检索到的官方页面没有给出完整董事会名单、委员会架构或详细股东权利框架。公开可见的是合规姿态:公司的反腐败声明称其设有诚信合规委员会、监察团队和正式举报渠道。这是正面治理信号,但不能替代董事会构成、投资人权利或控制权经济安排的披露。因此,后续章节应把 Shao、联合创始人、Zhang,以及合规架构作为已验证的公开领导层锚点,同时把正式董事会和控制权细节保留为未解决的尽调问题。[CO003, CO004, CO005, CO006, CO014, CO015]
| 人员 | 公开角色 | 背景或职责范围 | 重要性 | 披露备注 |
|---|---|---|---|---|
| Shao Tianlan | 创始人兼 CEO | 清华大学软件学士;Technical University of Munich 机器人学硕士 | 公司叙事的主要战略和技术锚点 | 创始人与市场契合公开可见,但已抓取材料未披露董事长职位和持股细节 |
| Fu Ao | 联合创始人兼商务 VP | 36Kr 项目材料公开将其列为商业侧联合创始人 | 帮助确认 Mech-Mind 并非只围绕单一高管搭建 | 角色公开,但当前汇报线和职责没有完整展开 |
| Ding Youshuang | 联合创始人兼 R&D 管理 VP | 36Kr 项目材料公开将其列为工程管理侧联合创始人 | 显示 CEO 之外早期已有技术管理梯队 | 留存材料中的公开履历深度有限 |
| Prof. Jianwei Zhang | 创始技术顾问兼首席科学家 | 中国工程院外籍院士、German National Academy of Engineering Sciences 院士 | 为技术叙事增加外部科学声望和连续性 | 官方团队页列名,但没有详细说明其日常运营角色 |
| Xu Tingting | 业务与市场副总裁 | KrASIA 引述其谈及审慎海外扩张和本地化 | 说明全球化有机构化人员配置,而非完全由创始人驱动 | 披露来自媒体,而不是官方团队页 |
本表是局部公开领导层图谱,不是完整管理层或董事会名册。正式董事会构成、委员会成员和投资人权利披露仍是开放尽调事项。
[CO003, CO004, CO005, CO006, CO014, CO015]1.3 融资形成、投资人和阶段信号
Mech-Mind 的融资历史相当扎实,但源材料包清楚显示,公开阶段标签滞后于公司真实状态。官方中文材料称累计融资已超过 RMB 2 billion,并列出 IDG Capital、Meituan、Sequoia China、Source Code Capital、Intel Capital、Qiming Venture Partners 等投资人。独立中文报道补上了带日期的节点:Aiqicha 记录了 Pre-A、A/A+、B、B+ 以及更早的 2021 年 C 轮;36Kr 2021 年 9 月报道补充了由 Meituan 和 IDG 领投、接近 RMB 1 billion 的 C 系列融资;36Kr 项目页则记录了截至 2025 年的 D、D++ 和 E 轮标签。MarketScreener 随后报道 2025 年 8 月公司完成近 RMB 500 million 的联合融资;香港财经媒体称公司已在 2025 年 9 月秘密递交香港 IPO,目标募资 USD 200 million。关键尽调点在于,并非所有公开数据集都同步更新。英文官方关于页面仍写着 Series C+ 和 USD 300 million 累计融资,Tracxn 仍以 Series C 标签呈现,并只统计六轮、USD 200 million。因此,更合适的标准阶段描述是「有公开证据支持 2024–2025 年 D/E 轮融资步骤和 IPO 准备的后期私营公司」,而不是「当前 Series C 公司」。[CO019, CO020, CO021, CO022, CO023, CO024]
| 利益相关方 | 公开角色 | 参与证据 | 重要性 | 尽调要求 |
|---|---|---|---|---|
| Meituan | 领投方或具名投资方 | 官方中文公司资料和 36Kr 2021 融资报道提及 | 释放战略科技背书和大型平台支持信号 | 确认当前持股比例以及任何董事会或观察员权利 |
| IDG Capital | 领投方或具名投资方 | 官方中文公司资料、36Kr 2021 融资报道和香港 IPO 报道提及 | 长期机构支持方,并在外部报道中持续可见 | 核实 IDG 是否仍主导治理影响力,或已被稀释 |
| Sequoia China | 具名投资方 | 官方中文资料和 36Kr 融资报道提及 | 是顶级风投反复参与的重要证明 | 澄清该持股是否已纳入当前 HongShan 架构,以及还保留哪些权利 |
| Source Code Capital | 具名投资方 | 官方中文资料和多条 36Kr 融资记录提及 | 显示多轮融资中的持续支持 | 需要当前持股和 2024-2025 轮次后的任何按比例跟投情况 |
| Intel Capital | 战略投资方 | 官方中文页面和 Aiqicha 保留的较早融资历史提及 | 为机器人叙事增加硬件和生态验证信号 | 判断 Intel 仍有财务参与,还是主要作为历史验证 |
| Qiming Venture Partners | 具名投资方 | 官方中文公司资料提及 | 扩大了投资方集合,不只依赖最常被引用的 Meituan 或 IDG 名称 | 当前持股和角色未公开披露 |
| China Xiong’an Group(产业客户) | 国资关联投资方 / 股东 | 36Kr 项目历史、MarketScreener 2025 融资和香港 IPO 报道提及 | 重要性在于把后期资本与地方国资支持连接起来 | 确认这是少数股权成长资本,还是更深层迁址或治理安排的一部分 |
| Hebei SOE Reform Fund | 国资关联投资方 / 股东 | 36Kr 项目历史、MarketScreener 和香港 IPO 报道提及 | 强化 2024-2025 期间国资背景资本参与的判断 | 澄清该基金是否附带政策条件、董事会权利或 pre-IPO 治理预期 |
| China Growth Capital | 成长阶段投资方 | 2025 年 8 月 MarketScreener 融资简报提及 | 显示 2025 年投资方组合并非只有国资关联资本 | 需要轮次经济条款,以及该投资方是否与老股流动性一起进入,或只投新股 |
本表不是完整股权结构表,只捕捉与资本形成、治理解读和 IPO 准备背景最相关的投资方和准战略利益相关方。
[CO019, CO021, CO022, CO023, CO024, CO025]从留存来源包梳理公司最有支撑的规模、融资和风险标记,以数字呈现。
这些 KPI 混合了当前公司页面披露,以及有时间戳的融资和法律风险快照。当前估值和收入被有意排除,因为留存公开来源没有确认它们。
[CO013, CO011, CO012, CO019, CO024, CO032]1.4 全球规模、里程碑和未解风险
公司概况中最强的证据在商业化规模,最弱的证据在财务透明度。当前官方页面称 Mech-Mind 已部署超过 24,000 个单元,服务 100 多家 Fortune Global 500 客户,覆盖近 50 个国家和地区,全球员工 600+。这些数字与更早但口径更窄的第三方信号大体一致:Automate 的协会资料提到在 50+ 个国家有 10,000+ 台工业 3D 相机、1,500+ 个客户;KrASIA 和 36Kr 在当前官方 24,000+ 说法之前报道过 15,000+ 个安装案例。这一变化说明规模确有增长,同时相机、安装案例和总单元之间的分母也在切换。里程碑记录也可复用。Aiqicha 保留了自 2017 年以来的早期融资节点;KrASIA 报道 2025 年 3 月东京机器人实验室开设;官方 iREX 2025 和 AW 2026 信息显示公司借大型国际展会发布产品,并展示更广的具身智能场景;QbitAI 的 Davos 报道则把 Mech-Mind 置于中国机器人出海玩家框架中。主要负面信号不是灾难性经营失败,而是不完整披露叠加可衡量的法律摩擦:Aiqicha 对雄安实体列出多项纠纷和诉讼相关记录。再加上估值、收入、ARR 和完整董事会披露均未公开确认,后续章节应把 Mech-Mind 定位为商业化已有规模、但仍部分不透明。[CO012, CO013, CO028, CO029, CO030, CO031]
| 日期 | 事件 | 类型 | 金额 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2016-09-12 | 公司成立 | 成立 | 成立日期 | Shao Tianlan 和清华海归创始团队 | 将公司锚定为 2016 年创业公司,而不是更晚出现的具身 AI 新进入者 |
| 2017-05 | Pre-A 融资 | 融资 | 数千万 RMB | Huachuang Capital / China Creation Ventures 领投方 | Aiqicha 保留的公开记录中第一项可见机构融资里程碑 |
| 2019-04 | A 轮和 A+ 轮融资完成 | 融资 | RMB 100M 级 | Aiqicha 摘要引用的 Mech-Mind 和投资人 | 显示业务在更广泛 AI 机器人资本热潮前,已越过种子实验阶段 |
| 2019-08 | Intel 投资记录 | 合作 | 战略投资 | Intel Capital | 增加生态可信度和硬件相邻的战略信号 |
| 2020-02 | B 轮主融资记录 | 融资 | 登记摘要披露的轮次 | Sequoia China | 显示早期规模化过程中持续获得顶级风投支持 |
| 2020-11 | B+ 轮记录 | 融资 | 接近 RMB100M | Source Code Capital 和 Sequoia China | 在 Meituan 领投资本到来之前,延长了 2021 年前的融资台阶 |
| 2021-09-27 | 36Kr 报道 C 系列融资 | 融资 | 接近 RMB1B | Meituan、IDG Capital、Sequoia China、Source Code Capital 等投资方 | 确立 Mech-Mind 为同代 AI + 工业机器人公司中融资最充足者之一 |
| 2023-08 | 36Kr 项目页显示 D 轮融资 | 融资 | 金额未披露 | Galileo Capital | 标志从 C 轮时代成长资本转向后期私募融资 |
| 2024-12 | 36Kr 项目页显示 D++ 融资 | 融资 | 金额未披露 | China Xiong’an Group 生态客户 | 证明公司仍能获得新增资本和国资关联支持 |
| 2025-03 | 36Kr 项目页显示 E 轮融资 | 融资 | 金额未披露 | Nanxiang Venture Capital 与 Hebei SOE Reform Fund 等国资背景投资方 | 支撑公司到 2025 年已越过 Series C/C+ 阶段的判断 |
| 2025-03 | 东京机器人实验室开设 | 规模 | 1000 sqm 设施,含 400 sqm 展示和培训区域 | Mech-Mind Japan | 表明公司在最重要的工业机器人市场之一加深本地化投入 |
| 2025-08-25 | MarketScreener 报道投资方组合融资 | 融资 | CNY500M | Broad-Ocean Motor;CICC 关联基金;Haihe 基金;China Growth Capital;Xiong’an 基金等 | 留存材料中最新一项清晰量化的融资事件 |
| 2025-09-24 | 香港 IPO 递表报道 | 治理 | 秘密递表;据报道目标约 USD200M | Mech-Mind 和香港财经媒体 | 提升未来刷新中对经审计披露和修订图谱跟踪的必要性 |
| 2026-03-12 | AW 2026 产品和应用发布潮 | 产品 | 10+ 个演示单元和全球产品首发 | Mech-Mind | 显示公司仍在把资本转化为产品宽度和国际商业信号 |
| 2026-05-26 | 登记法律风险快照留存 | 反向 | 18 起商业纠纷;3 条立案记录;6 条开庭公告;4 条诉讼关系 | Xiong’an 主体的 Aiqicha 资料 | 创建一个可复用的反向数据点,但不把它夸大成更广泛的监管失败 |
这是第 1 章的单一记录年表。留存来源包支持的日期会明确保留;最后一条反向记录是按访问日期记录的登记快照,而不是单个有日期的法院事件。
[CO001, CO021, CO022, CO023, CO024, CO026]公司从创立到 2026 展会周期的资本形成、全球化和可信度建设里程碑,高层级梳理。
现有来源包只有月份精度时,日期保留到月份。最后一项法律风险来自按访问日锚定的当前工商注册快照。
[CO001, CO002, CO003, CO021, CO024, CO026]1.5 展示材料
02市场分析
2.1 市场边界、纳入支出与替代方案
应把 Mech-Mind 分析为工业 AI 和 3D 视觉基础设施供应商,而不是完整机器人 OEM。公司自有产品组合和第三方资料一致描述了一套由工业 3D 相机、视觉软件、深度学习工具和无代码机器人编程组成的栈,用于料箱拣选、拆垛、机床上下料、精密装配和检测等任务。这意味着公司参与的是架在机器人手臂和工作单元之上的感知、规划和部署层,而不是工厂或仓库自动化资本开支中的每一美元。纳入支出因此包括 3D 传感器、视觉和 AI 软件、部署工具、面向具体流程的配置,以及本地集成支持。排除支出则包括大多数机器人手臂硬件、输送线、AS/RS 钢结构,以及无关的工厂 IT。最接近的替代方案是人工搬运、更简单的 2D 视觉、硬编码机器人单元,或完全绕开柔性 3D 感知的固定自动化。因此,市场边界必须从困难流程出发,而不是从最宽泛的公开自动化类别出发。[CM001, CM002, CM003, CM004, CM005, CM006]
| 细分 / 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 与 Mech-Mind 的相关性 |
|---|---|---|---|---|
| 工业 AI + 3D 视觉工作单元 | 3D 相机、视觉软件、深度学习、机器人编程、部署支持 | 大部分机器人臂硬件和泛工厂资本开支 | 工厂自动化、集成商、配送中心(DC)运营 | Mech-Mind 的核心切入点 |
| 机器视觉和检测系统 | 相机、图像处理、质量软件、测量工具 | 非视觉自动化和无关工厂软件 | 质量、制造工程、工厂运营 | 重要相邻市场 |
| 仓储自动化 | 机器人、编排、物料搬运、拣选和拆垛系统 | 无关企业软件和仓库外运输 | 3PL、零售商、物流运营商 | 物流用例的外围 TAM |
| 工业机器人装机基础 | 能为视觉创造附着点的机器人单元和自动化项目 | 没有机器人单元的纯手工流程 | OEM、Tier 1、制造商 | 装机基础需求面,不是完整收入池 |
| 现状替代支出 | 人工搬运、简单 2D 视觉、定制代码、固定式自动化 | 先进 3D AI 引导 | 运营和工程预算 | 主要替代空间 |
| 集成商与本地化服务 | 系统集成、现场调试、支持、培训 | 没有部署的纯硬件转售 | 集成商和终端客户 | 关键商业路径 |
边界逻辑把 Mech-Mind 视为工业自动化中的感知与部署层,而非完整机器人 OEM 或通用工厂软件厂商。
[CM001, CM002, CM003, CM004, CM005, CM006]2.2 规模测算视角与区域动态
公开市场数据方向上积极,但不能简单相加。IFR 显示全球工业机器人装机基数很大且仍在增长,但 Mech-Mind 无法捕获所有机器人硬件支出;可变现的楔子是骑在这组基数上的更窄视觉与引导层。公开源材料包里最接近的相邻市场是仓库自动化、仓库机器人和机器视觉。它们都足够大,能够支撑有意义的品类领导者,但范围、预测窗口和方法论不同;保留这些分布,比假装它们能收敛成一个共识 TAM 更诚实。中国在本章格外重要,因为它同时是最大工业机器人市场、领先的电子安装市场,也正变成一个越来越由本土供应商主导的环境。与此同时,公开来源对如何衡量机器人密度并不一致,这提醒我们区域基准必须严守分母。务实结论是,Mech-Mind 受益于中国深厚的自动化基础,也受益于全球机器视觉和仓库自动化需求扩张;但投资人仍需要受约束的可服务市场(SAM)视角,而不是宽泛的标题式 TAM。[CM007, CM008, CM009, CM010, CM011, CM012]
| 发布方 | 年份 | 地域 | 数值 / 指标 | 增长 | 方法论视角 | 置信度 | 局限 |
|---|---|---|---|---|---|---|---|
| IFR / World Robotics 2025 | 2024-2028 | 全球 | 2024 年安装 542k;2025 年预计 575k;2028 年 >700k | 长期正向 | 工业机器人装机基数与预测 | 高 | 机器人装机量不等于 Mech-Mind 能拿到的收入 |
| IFR / Yicai / SCIO | 2024-2028 | 中国 | 2024 年安装 295k 台;存量 >2m 台;占全球需求 54% | 据 IFR 摘要,到 2028 年平均潜力约 ~10% | 中国需求集中度与国产份额趋势 | 高 | 中国主导地位不会自动转化为海外份额 |
| People / Yicai / ChinaPower | 2023-2024 | 中国 | 公开引用的机器人密度为每 10k 名工人 166–470 台 | N/A | 自动化强度视角 | 中 | 公开密度指标的分母因来源而异 |
| Precedence Research | 2026-2035 | 全球 | 仓储自动化:2026 年 $29.30B,到 2035 年 $107.36B | 15.56% CAGR | 广义仓储自动化外层 TAM | 中 | 范围远宽于 Mech-Mind 可直接获取的层级 |
| Mordor Intelligence | 2026-2031 | 全球 | 仓储自动化:2026 年 $34.17B,到 2031 年 $65.74B | 13.98% CAGR | 按组件、终端用户和制约因素拆分充分的仓储模型 | 中 | 仓储口径混合了硬件、软件和服务 |
| Grand View Research | 2022-2030 | 全球 | 仓储机器人:2022 年 $4.31B,到 2030 年 $17.29B | 19.6% CAGR | 仓储自动化中的机器人子集 | 中 | 基准年较早;口径窄于整体仓储自动化 |
| Precedence Research / A3 | 2026-2029+ | 全球 | 机器视觉:2026 年 $26.07B;A3/Interact 预计到 2029 年 CAGR 约 ~7.7% | 长期 12.8%,中期 ~7.7% | 机器视觉市场视角,叠加 3D 软件、AI 和物流增长 | 中 | 仍宽于仅限机器人导引的 3D 视觉 |
| 分析师综合判断 | 2026 | 全球 | 工业 3D 机器人导引没有清晰独立的公开 TAM | N/A | 受证据约束的缺口判断 | 低 | 需要自下而上的工作流测算或管理层数据 |
本表刻意保留相邻但口径不同的市场视角,不强行拼成一个合成 TAM。最后一行把缺失的独立机器人导引 TAM 记为尽调发现,而不是市场规模估计。
[CM007, CM008, CM009, CM010, CM011, CM012]分层公开视角从更宽的相邻自动化市场,收窄到规模小得多的 Mech-Mind 式 3D 引导楔形市场;底层为分析估计,并明确未披露。
前三层是 Precedence Research 和 Grand View Research 发布的相邻市场数字,不是一套嵌套数据。1.2 是针对汽车、电子和物流中硬感知工作站的示例性分析楔形,只用于说明 Mech-Mind 真正可捕获区域远窄于公开的外层 TAM 数字。
[CM015, CM017, CM018, CM021, CM041]以 USD 十亿计的公开市场规模跨度很大,取决于视角是仓储自动化、机器视觉、仓储机器人,还是工业机器人安装市场价值。
第一行保留 Precedence 与 Mordor 对 2026 仓储自动化的口径差。其他行是同一币种下的单一来源相邻市场视角,并不是可直接比较的 2026 SAM。目的在于显示离散度和类别错配,而不是暗示这些数字可以相加。
[CM015, CM016, CM017, CM018, CM021]2.3 买方分群、垂直行业与采用路径
买方地图由流程驱动,而不是由客户名称驱动。在汽车和电子行业,类似 Mech-Mind 的系统通常落在工厂自动化、质量或制造工程预算里,因为回报来自周期时间、废品率降低、精度提升,以及在困难部件上的劳动力杠杆。物流场景中,自然买方是配送中心运营团队、3PL,以及部署拆垛、包裹导入、料箱拣选和混合 SKU 单件拣选单元的集成商。用户是产线操作员、仓库主管、自动化工程师和集成商团队;付款方是工厂或站点运营职能,而不是泛化的「AI」预算。这一结构很重要,因为在反光、透明、深色或随机堆叠物体上,传统视觉失效时,Mech-Mind 的公开适配度最强。这些条件反复出现在汽车零部件、电子装配、电池和物流包裹中。因此,系统集成商是核心商业层,而不是事后补丁,因为他们吸收部署复杂度,并为终端客户把工作单元本地化。[CM022, CM023, CM024, CM025, CM026, CM027]
| 细分市场 | 买方 | 用户 | 付费方 | 工作流 | 预算负责人 | 采用触发因素 |
|---|---|---|---|---|---|---|
| 汽车 OEM / Tier 1 | 工厂自动化或制造工程团队 | 单元操作员、质量工程师、机器人工程师 | 工厂资本开支 / 运营预算 | 机床上下料、在线测量、装配、检测 | 制造副总裁 / 工厂经理 / 自动化负责人 | 反光零件、精度、正常运行时间、降低报废率 |
| 电子 / 半导体 / 电池 | 质量、工艺或设备工程团队 | 产线主管、QA 团队、自动化工程师 | 工厂或产品线自动化预算 | OCR、胶路检测、透明物体抓取、精密装配 | 制造工程总监 / QA 负责人 | 多品种精密作业与可追溯性 |
| 物流 / 3PL / 电商配送中心 | 运营、创新或自动化项目负责人 | 仓库主管、一线员工、集成商 | 运营或站点自动化预算 | 拆垛、包裹导入、料箱拣选、单件拣选 | COO / 履约副总裁 / 配送中心总经理 | 用工短缺与吞吐压力 |
| 食品 / 医药 / CPG | 运营或包装自动化团队 | 包装负责人、质量团队、机器人工程师 | 运营资本开支和质量预算 | 码垛、拆垛、检测、难处理包装的搬运 | 工厂经理 / 质量负责人 | 安全性、一致性和用工杠杆 |
| 系统集成商和机器人工作站搭建商 | 集成商客户和解决方案团队 | 应用工程师和调试人员 | 集成商项目预算,最终转嫁给终端客户 | 现场设计、集成、部署、本地支持 | 集成商业务负责人 | 需要预集成视觉栈和更快部署 |
| 先在中国落地的出口制造基地 | 国内工厂先落地,再复制到海外 | 本地工程团队加海外支持团队 | 国内资本开支与出口项目预算混合 | 已验证的国内工作站复制到海外 | 区域运营负责人 | 将已验证工作流低风险输出海外 |
预算负责人字段基于工作流经济性,以及公开材料对部署、渠道和本地化的描述推断。Mech-Mind 未按垂直行业公布正式价目表。
[CM022, CM023, CM024, CM025, CM026, CM027]将买方细分映射到用户画像、付款方模型、预算权和最可能把 Mech-Mind 式部署拉进生产的采用触发点。
[CM023, CM024, CM025, CM026, CM027, CM028]2.4 增长向量与采用阻碍
需求驱动真实且多因共振。仓库运营商仍把劳动力短缺和人工成本列为最清晰的近期采用触发因素;IFR 和 A3 则把 AI 自主性、IT/OT 融合和物流数字化描述为机器人和视觉系统可处理流程范围扩大的背景。这个背景有利于 Mech-Mind,因为当客户希望从现有机器人单元里获得更高柔性时,其无代码和 3D 感知叙事最有力。但源材料包同样清楚地说明,需求不会无摩擦地转化为生产收入。资金审批率仍落后于口头兴趣;许多制造商难以选择正确技术、接入遗留系统,或提供维持项目推进所需的内部专业能力。当机器人更靠近人,安全、认证和人工监督也仍是重要部署关口。因此,市场不是被需求不足卡住,而是被实际成本、集成和准备度门槛卡住;这些门槛把试点需求和规模化铺开隔开。[CM030, CM031, CM032, CM033, CM034, CM035]
| 驱动因素 / 约束 | 方向 | 时间窗口 | 含义 | 尽调问题 |
|---|---|---|---|---|
| 劳动力可得性与用工成本压力 | 正向 | 当前 | 让仓库和工厂继续把自动化放在预算议程上 | 目标汽车、电子和物流客户的人员流失有多严重? |
| AI 自主性与 IT/OT 融合 | 正向 | 当前至中期 | 让 3D 视觉导引系统更灵活,也更容易拿到立项理由 | Mech-Mind 的 ROI 有多少来自软件灵活性,而不是硬件替换? |
| 物流机器视觉扩张 | 正向 | 中期 | 在传统工厂检测之外,形成增长更快的需求池 | 管线中物流相对制造的占比是多少? |
| 中国规模与政策支持 | 正向 | 当前至长期 | 带来密集需求、可背书案例和国产供应商动能 | 在中国验证的工作流向海外客户迁移的可行性有多强? |
| 资金审批缺口 | 负向 | 当前 | 兴趣不能顺畅转化为已签项目 | 试点中有多少因为预算而非技术原因停滞? |
| 高 AI 成本、数据基础设施与技能缺口 | 负向 | 当前 | 拖慢从早期采用者向外扩张 | 每次部署需要多少客户赋能或伙伴培训? |
| 存量系统集成与固定系统资本开支 | 负向 | 当前至中期 | 回本周期长,WMS/ERP 摩擦会推迟采用 | 各工作流的部署周期中位数和集成负担是多少? |
| 安全、验证与监督 | 负向 | 持续 | 认证和人机安全工作仍是关口 | Mech-Mind、集成商和终端客户之间,谁承担安全验证成本? |
本表综合了调研证据、行业趋势评论和监管指引。目的在于说明需求真实存在,但仍按工作流逐一转化,而不是像自上而下 TAM 幻灯片暗示的那样自动兑现。
[CM030, CM031, CM032, CM033, CM034, CM035]示例性指数漏斗,展示广义自动化需求如何经过资金、集成和工作流匹配关口收窄,最终变成与 Mech-Mind 相关的收入。
92 来自 Vention / Industry Week 调查,称自动化必不可少;80 合并了 Peerless 调查中的 48% 当前使用和 32% 计划使用;32 是同一调查中的已获批资金数字。最后阶段是分析估计,代表同时匹配 Mech-Mind 硬感知适配度的项目占比更窄。
[CM031, CM036, CM037, CM039, CM043]2.5 Mech-Mind 的全球位置与仍需尽调的问题
Mech-Mind 最适合作为具身工业 AI 的「铲子和镐」供应商来理解:它销售感知、规划和部署层,可以横跨汽车、电子和物流单元,而不要求绑定某一种机器人形态或某一个终端市场。这在战略上有吸引力,因为当客户购买的是柔性而不只是更多金属时,机器视觉、3D 软件和仓库自动化层可以跑赢底层机器人基数。公司在中国、日本、韩国、德国和美国的本地化布局也强化了这一点:它想通过全球制造和集成商生态销售,而不只依靠中国国内需求。承销上的谨慎点在于,这里可用的公开市场数字大多来自相邻市场。投资人仍缺少按地区和垂直行业拆分的收入结构、每个机器人单元的附着率,以及按应用拆分的流程级 ROI。因此,可投资问题不是市场是否足够大——答案是足够大——而是 Mech-Mind 能否反复把这套庞大装机基数转化为软件含量高、全球可支持、且部署摩擦可控的收入。[CM040, CM041, CM042, CM043]
2.6 展示材料
03竞争格局
3.1 格局与竞争类别
Mech-Mind 不属于一个整齐的同业组。最接近的替代方案可拆成四类。第一类是 Cognex、Keyence、Omron、ISRA Vision 等老牌机器视觉厂商,它们已经把受信任的自动化、检测和传感产品卖进大型工厂。第二类是 Photoneo、Roboception、Pickit 等聚焦 3D 引导的专业厂商,围绕料箱拣选、拆垛和相关搬运流程,用更窄的栈切入。第三类是 Intrinsic 等平台层替代者,它们打包技能、仿真和机器人控制抽象,而不是 Mech-Mind 式的单供应商栈。第四类是 Universal Robots 等机器人平台生态,硬件、市场合作伙伴和集成商会吸收部分相同买方预算。Mech-Mind 公开材料显示,它在这群对手中仍有差异化:公司把相机、机器人引导、深度学习、检测和预制工作单元包装成一个 Eye-Brain-Hand 方案。竞争问题因此不是替代者是否存在——它们显然存在——而是 Mech-Mind 的集成式流程包装,是否足以抵消周围老牌渠道、生态和定价优势。[CP001, CP002, CP006, CP011, CP014, CP016]
| 竞争对手 | 类别 | 公开产品范围 | 主要部署模式 | 垂直行业 / 地域信号 | 相比 Mech-Mind 的关键短板 |
|---|---|---|---|---|---|
| Mech-Mind Robotics | 直接全栈同类 | 3D 相机、规划、深度学习、检测、具身工作站 | 不绑定机器人品牌的工作流栈,通过集成商和直销部署销售 | 50+ 个国家;汽车、物流、电子、食品、零售 | 公开定价和精确赢率数据仍稀缺 |
| Cognex | 机器视觉龙头存量厂商 | 机器视觉、智能相机、视觉传感器、条码读取 | 工厂自动化存量厂商,装机信任基础广 | 全球工业客户基础;存量厂商声誉强 | 公开材料较少将其定位为一体化「眼-脑-手」工作站栈 |
| Keyence | 机器视觉 / 机器人导引存量厂商 | 2D、线扫、3D 相机,AI 与规则视觉,3D VGR | 报价驱动的自动化销售,配校准和 CAD 辅助设置 | 制造业装机基础大 | 公开来源更强调视觉深度,而非单一供应商工作流栈 |
| Photoneo | 聚焦 3D 机器人视觉的对手 | Locator Studio、Bin Picking Studio、拆垛、MotionCam-3D、扫描仪 | 围绕搬运和运动物体感知部署 3D 机器人视觉 | 工厂、仓库、物流 | 公开控制 / 检测层比 Mech-Mind 窄 |
| ISRA Vision | 偏检测的存量厂商 | 面向汽车、电池、玻璃、金属、纸张、塑料的机器视觉解决方案 | 工业检测与图像处理销售 | 在重检测行业覆盖广 | 公开材料较少强调不绑定机器人品牌的拣选 / 编程栈 |
| Intrinsic | 平台型替代者 | Flowstate、技能架构、视觉模型、数字孪生、AI 工具 | 跨第三方机器人和开放工具运行的软件平台 | 通过 FANUC、ROS、Gazebo、Open-RMF 做生态打法 | 本身不像打包式 3D 相机厂商 |
| Roboception | 聚焦 3D 感知的对手 | rc_visard 传感器,加 CADMatch、ItemPickAI、BoxPick、SLAM、URCap | 面向机器人导引的模块化传感器 + 软件模块 | 模块化感知买方和集成商 | 在检测和工作站上的公开广度窄于 Mech-Mind |
| Omron | 自动化存量厂商 / 相邻对手 | 机器人技术加机器人视觉系统,用于装配、检测、包装、运输 | 广泛的自动化和机器人销售路径 | 汽车、数字、食品、医疗、物流 | 机器人视觉只是更大自动化组合的一部分 |
| Universal Robots / Teradyne | 机器人平台生态替代者 | 协作机械臂、市场伙伴、部署与预算内容 | 机器人硬件加伙伴 / 集成商生态 | 协作机器人装机基础广;伙伴渠道 | 需要伙伴插件才能匹配 Mech-Mind 式完整视觉栈 |
| Pickit | 聚焦 3D 导引的对手 | 料箱拣选、拆垛、装配、在线测量 | 应用优先的 3D 视觉套件 | 适合 SMB 和协作机器人的搬运任务 | 工作流范围比 Mech-Mind 更聚焦 |
各行概括每个竞争对手最可见的公开产品和部署姿态,并非穷尽内部能力清单。
[CP001, CP011, CP012, CP013, CP014, CP016]Mech-Mind 在技术栈宽度上高于聚焦型 3D 引导厂商,但在渠道和平台控制上低于生态更重或已有装机基础的对手。
轴向评分是基于公开产品范围、生态站位和销售渠道信号得出的有证据支撑的顺序估计;相对位置比绝对分值更重要。
[CP011, CP012, CP014, CP016, CP019, CP022]3.2 产品广度与部署模式
公开证据显示,Mech-Mind 比多数专业 3D 引导对手更宽,但在渠道触达上窄于最大自动化老牌厂商。其中英文公开材料覆盖工业 3D 相机、激光轮廓仪、Mech-Vision、Mech-Viz、Mech-DLK、Mech-MSR 和具身机器人工作站,文档足迹则覆盖引导、检测、测量和机器人通信。这套公开栈比 Pickit 或 Roboception 更宽,后两者看起来更集中在拣选和感知模块;也不同于 Photoneo 对机器人视觉、料箱拣选、拆垛和高速 3D 相机的强调。Keyence 和 Cognex 仍是基准老牌厂商,因为它们把宽机器视觉产品线与长期自动化信任配在一起。Intrinsic 又是结构性不同的对手:它向上移动到技能、仿真和数字孪生流程,可跨第三方机器人运行。结果是,Mech-Mind 的主要优势不是拥有机器人手臂,而是减少在现有机器人单元上搭建困难 3D 视觉任务所需的独立供应商数量。[CP002, CP003, CP010, CP012, CP013, CP014]
| 采购标准 | Mech-Mind | Cognex / Keyence | Photoneo | Intrinsic | Roboception / Pickit | UR 生态 |
|---|---|---|---|---|---|---|
| 从感知到规划的单一供应商栈 | 高 | 中 | 中 | 低 | 低 | 低 |
| 专用 3D 机器人导引深度 | 高 | 中 | 高 | 低 | 高 | 低 |
| 深度学习 / AI 应用层 | 高 | 中 | 中 | 高 | 中 | 低 |
| 公开栈中的检测 / 测量层 | 高 | 高 | 中 | 低 | 低 | 低 |
| 数字孪生 / 仿真 / 开发者抽象 | 中 | 低 | 低 | 高 | 低 | 中 |
| 机器人硬件 / 市场杠杆 | 低 | 低 | 低 | 低 | 低 | 高 |
能力标签是基于公开材料的定性综合判断;比较的是公开栈宽度,而非基准测试性能数字。
[CP002, CP003, CP012, CP013, CP014, CP016]Mech-Mind 在单一供应商技术栈覆盖上领先,Intrinsic 在仿真和技能抽象上领先,UR 在机器人市场平台杠杆上领先。
单元格概括公开技术栈可见度,而非保密性能数据;低评级可能意味着公开证据有限,并不等于能力为零。
[CP002, CP003, CP012, CP013, CP014, CP016]3.3 商业形态、地理与生态
商业透明度远薄于产品营销。在已审阅的 Mech-Mind、Keyence、Photoneo、Intrinsic、Roboception、Pickit 和 ISRA 页面中,公开材料强调演示、用例和联系入口,而不是目录价格。Universal Robots 是这组源材料中的例外,因为其自有预算指南讨论工具、软件、培训、支持和集成商成本,一份独立价格指南也发布了手臂级别的大致区间。Mech-Mind 的第三方资料称其定价有竞争力、部署简单、对集成商友好,但公开标价条款仍未出现。地理布局是 Mech-Mind 的相对亮点:官方和独立来源都把公司置于 50 多个国家,Kr-Asia 还记录了东京实验室开设,强化了公司有意推进海外扩张的判断。即便如此,生态压力真实存在。Intrinsic 正把 Flowstate 接入 FANUC 和开源机器人工具,Universal Robots 则在硬件周围叠加市场和伙伴渠道。这些模式会让买方更愿意混搭组件,而不是选择一套紧密集成的视觉栈。[CP004, CP005, CP006, CP007, CP028, CP029]
| 供应商 | 公开定价信号 | 包装 / 销售路径 | 部署摩擦信号 | 含义 |
|---|---|---|---|---|
| Mech-Mind | 已审阅材料中无标价 | 竞争性价格定位,加对集成商友好的全栈 | 强调快速部署,但商务条款定制 | 商业验证比标价更重要 |
| Cognex | 已审阅材料中无公开定价 | 存量自动化销售 | 买方依赖装机信任和产品组合宽度 | 竞争靠信任和宽度,而不是公开价格 |
| Keyence | 已审阅材料中无公开定价 | 直接自动化销售,配设置工具和 CAD 辅助工作流 | 部署看起来结构化,但由报价牵引 | 可凭自动化可信度销售,而非透明价格 |
| Photoneo | 已审阅材料中无公开定价 | 应用聚焦的 3D 视觉套件 | 可能是项目制视觉销售 | 买方更看匹配度,而非标价成本 |
| Intrinsic | 已审阅材料中无公开定价 | 平台 / 技能 / 生态路径 | 软件集成导向更强 | 可作为控制层预算竞争,而不是相机 SKU |
| Universal Robots | 官方预算指南,加上独立机械臂价格区间 | 机器人硬件,加上合作伙伴和集成商生态 | 硬件价格可见,完整工作站成本取决于工装、软件和支持 | 在这组厂商里给出最清晰的公开成本锚 |
| Pickit | 已审材料未见公开定价 | 聚焦 3D 视觉应用销售 | 先卖应用,面向柔性自动化买家 | 相比定制集成,更像产品化方案 |
| ISRA / Omron | 已审材料未见公开定价 | 更广的自动化或检测销售动作 | 品类可信度可能压过价格透明度 | 既有厂商可用打包方案压制单点方案 |
本表比较价格透明度和打包方式,而非实际合同价值;大多数已审厂商不披露公开标价。
[CP028, CP029, CP030, CP031, CP032]3.4 护城河耐久性与竞争风险
Mech-Mind 最强的公开护城河主张是连贯性:它把 3D 传感、机器人编程、深度学习、检测和具身 Eye-Brain-Hand 工作单元包装成一套不绑定机器人品牌的方案。当客户处理透明物体、反光金属、拆垛和混合 SKU 拣选时,这一点尤其有价值;产品页面显示 Mech-Mind 正重压真实世界部署痛点。风险在于,如果大型老牌厂商持续扩展产品组合和伙伴关系,或软件平台让视觉和操作能力更容易跨机器人品牌互换,单靠连贯性未必耐久。Intel Market Research 指出,老牌厂商产品组合扩展和伙伴关系仍是持续力量;QYResearch 则把 Mech-Mind 放在拥挤的 3D 视觉供应商阵列之中,而不是阵列之外。公开规模信号方向上也很强,但并非完全精确:一个来源引用 10,000+ 台相机,另一个引用 15,000+ 个安装案例,官方首页则称 24,000+ 台相机。这不推翻其全球足迹,但意味着在称护城河耐久之前,尽调仍需要精确客户重叠、按机器人品牌拆分的附着率,以及对 Cognex、Keyence、Photoneo 和 UR 生态替代方案的赢单/输单数据。[CP006, CP009, CP023, CP026, CP027, CP035]
| 护城河或风险 | 方向 | 重要性 | 当前公开证据 | 尽调问题 |
|---|---|---|---|---|
| Eye-Brain-Hand 技术栈宽度 | 优势 | 单一厂商可覆盖感知、规划、AI 和检测 | Mech-Mind 公开技术栈覆盖相机、编程、深度学习、测量和工作站 | 客户支出中有多少落到 Mech-Mind,而不是合作伙伴? |
| 集成商友好型部署 | 优势 | 集成负担更低,可加快试点和规模化 | A3 档案强调产品易用、价格有竞争力,并聚焦集成商 | 从试点到稳定生产的中位用时是多少? |
| 既有厂商的产品组合与渠道力量 | 风险 | Cognex、Keyence、Omron 和 ISRA 带着既有信任和更广自动化版图入场 | Qviro 和 IMR 凸显既有厂商深度及伙伴扩张 | Mech-Mind 对阵具名既有厂商的胜率是多少? |
| 平台抽象与生态替代 | 风险 | Intrinsic、FANUC 及 UR 式生态可降低客户对单一集成视觉厂商的依赖 | Flowstate 加上 FANUC/ROS 和 UR 市场,拓宽了替代选项 | 哪些机器人品牌贡献了 Mech-Mind 大部分收入和附加率? |
| 拥挤的 3D 视觉赛道 | 风险 | 买家可评估多家可信的机器人引导专精厂商 | QYResearch 将 Mech-Mind 列入多家全球 3D 视觉厂商,并提到前五集中度 | Mech-Mind 在吞吐量或正常运行时间上,哪里已明确优于 Photoneo、Pickit 和 Roboception? |
| 规模指标不一致 | 风险 | 规模证据方向上强,但仍缺一个清晰的公开分母 | 公开来源分别提到 10,000+ 台相机、15,000+ 次安装和 24,000+ 台相机 | 当前经审计安装基数、活跃客户数和分地区收入是多少? |
风险清单把有支持的方向性证据与尚未解决的投资判断问题分开;多个关键商业指标仍未公开。
[CP005, CP006, CP028, CP032, CP037, CP038]公开证据最强的是 Mech-Mind 技术栈广度和全球覆盖,最弱的是透明定价和精确竞争胜率披露。
KPI 标签只压缩公开证据;非公开客户重叠、续约和输赢数据可能会实质改变护城河持续性判断。
[CP006, CP028, CP031, CP037, CP038, CP039]3.5 展示材料
04财务情况
4.1 收入模式与定价信号
公开证据指向解决方案驱动的收入模式,而不是纯软件订阅业务。Mech-Mind 销售一套「Eye-Brain-Hand」栈,涵盖工业 3D 相机、机器视觉软件、机器人编程软件和相关具身智能工作站,并通过咨询、方案设计、培训、部署和维护支持伙伴。这在财务上很重要,因为公司并不只是授权代码:部分收入几乎肯定来自硬件,部分来自软件许可,部分来自跟随集成商主导项目的部署或生命周期服务。官方材料反复强调伙伴和价值提供商,说明渠道经济性和实施支持是变现核心。缺失的是实际合同结构。已审阅来源没有发布标价、折扣阶梯、软件维护费率,或一次性部署收入与经常性支持之间的拆分。我们找到的唯一明确价格代理是历史性的:Jiemian 在 2021 年报道称,Mech-Mind 产品定价大约是外国主流竞争对手的一半。这在方向上有用,但对当前承销来说太旧。正确结论是,Mech-Mind 可能通过捆绑自动化单元或解决方案栈变现;如果软件复用和重复部署持续上升,长期经济性会更好,但公开数据还无法拆出已实现 ASP 或经常性收入质量。[CI001, CI002, CI003, CI004, CI005, CI006]
| 收入来源 | 公开证据 | 可能机制 | 当前状态 | 质量 | 尽调问题 |
|---|---|---|---|---|---|
| 工业 3D 相机 / 轮廓仪 | 官方 Mech-Eye 页面和产品栈 | 通过集成商或自动化项目销售硬件 | 明确在售 | 除非有重复补货,主要是一次性收入 | 提供各产品系列的 ASP、BOM 和硬件毛利率 |
| 视觉软件许可 | 官方 Mech-Vision 和 Mech-Viz 页面 | 按工作站、站点或项目的软件许可,加上更新 | 明确在售 | 可能经常性,但结构未披露 | 披露许可期限、维护附加率和续约率 |
| 部署 / 调试服务 | 官方关于和合作页面提到咨询、设计、培训、部署、维护 | 服务收入随安装和上线发生 | 明确暗示 | 有助于落地收入,但毛利率可能更低 | 拆分可计费服务占比和利用率 |
| 伙伴 / 渠道经济性 | 官方材料反复面向集成商和价值提供商 | 渠道折扣、转售毛利,或解决方案交付收入分成 | 已暗示 | 可扩大触达,但会压低实现价格 | 展示伙伴折扣区间和渠道结构 |
| 具身 AI / 机器人工作站系统 | 官方机器人工作站产品页和料箱拣选方案页 | 销售高 ASP 的打包方案,包含硬件、软件和部署 | 在售 | 可抬高单项目收入,但也增加交付复杂度 | 提供订单数、平均系统价值和服务附加率 |
公开证据支持硬件-软件-服务打包模式,但无法看清一次性收入与经常性收入的实际结构。
[CI001, CI004, CI006, CI009, CI010]| 问题 | 公开答案 | 最佳代理指标 | 投资判断风险 | 尽调问题 |
|---|---|---|---|---|
| 是否公布标价? | 未找到公开标价 | 官方页面仅提到价格有竞争力 | 无法对实现 ASP 或折扣做基准比较 | 收集当前报价和经销商价目表 |
| 相对海外同行的价格? | 仅有历史信息 | 界面称 2021 年定价约为海外主流竞品的一半 | 该信号可能过时,或只适用于促销 | 将当前报价与 Cognex / Keyence / 本土同行比较 |
| 哪些产品一起销售? | 可能打包销售 | 2021 年报道和官方产品栈指向相机 + 视觉 + 编程 + 服务打包 | 打包会遮蔽纯软件经济性 | 提供 SKU 层面的硬件、软件、服务收入拆分 |
| 是否披露支持节奏? | 部分披露 | 2021 年报道称,成熟项目按周计,新应用按月计 | 部署强度会挤压毛利率 | 展示当前平均实施和支持工时 |
| 经常性元素可见吗? | 不清楚 | 维护、培训和软件更新可能重复发生,但条款未披露 | 难以评估收入耐久度或净留存 | 披露维护续约和软件更新定价 |
定价证据大多是相对或历史口径;已审来源均未披露当前价目表、折扣梯度或维护条款。
[CI005, CI006, CI007, CI008, CI009]Mech-Mind 似乎靠产品与合作伙伴交付,把能力转成客户收入。
该桥接图是定性的,因为已审阅来源没有披露合同结构、平均售价(ASP)或收入确认政策。
[CI001, CI004, CI006, CI009, CI010]4.2 牵引力代理与单位经济逻辑
由于 Mech-Mind 是私营公司,最有信息量的财务代理是部署、市场份额和交付信号。官方材料称在近 50 个国家部署了 24,000+ 个单元,并拥有 100 多家 Fortune Global 500 客户;第三方报道则称公司连续五年领先中国 3D 视觉引导工业机器人细分市场,并在 2024 年达到 38% 份额。更早报道还显示,订单曾以每年超过三倍增长,2021 年团队已超过 300 人,并有可观研发投入。合在一起,这些指标支撑真实商业化规模,但还不能转化为干净的收入质量指标。单位经济可能是混合的。相机、轮廓仪和机器人工作站带来硬件采购,以及可能的库存或营运资本暴露;方案设计、部署和维护带来人工成本;但如果重复部署增长快于现场服务强度,软件复用、无代码编程和跨多个机器人品牌的标准化适配会创造杠杆。最有用的正向代理是标准化:Mech-Mind 有意避开制造机器人本体,转而试图把可复用的感知和控制层卖进多个行业。最有用的负向代理是集中度风险:近期分析认为,长尾客户仍更难拿下,小客户对成本敏感,高端汽车部署仍面临国际全栈竞争对手压力。[CI011, CI012, CI013, CI014, CI015, CI016]
| 驱动因素 | 公开证据 | 可能影响 | 置信度 | 尽调问题 |
|---|---|---|---|---|
| 硬件含量 | 3D 相机、激光轮廓仪和机器人工作站是核心产品 | 相比纯软件,毛利率更低,也可能需要库存和供应商管理 | 中 | 展示产品 BOM、供应商集中度和库存周转 |
| 软件复用 | 无代码 / 可视化编程和可跨多种用例复用的视觉栈 | 应用重复后,可抬高边际贡献率 | 中 | 披露软件附加率及各模块毛利率 |
| 部署和维护人力 | 官方生命周期支持,加上界面对服务周期的披露 | 带来人力密集型 COGS 和实施开销 | 中 | 提供实施工时、现场服务人数和支持比率 |
| 跨机器人品牌标准化 | 雪球称标准产品可适配 20+ 个机器人品牌 | 提升可扩展性,减少每个项目的定制工程量 | 低 | 按部署批次展示平均定制工时 |
| 国际支持网络 | 50 国部署说法,以及海外收入占比评论 | 增加本地支持、差旅和营运资金负担 | 低 | 披露分地区毛利率和海外服务成本 |
公开记录足以识别成本驱动因素,但不足以计算真实的单工作站毛利率、CAC 或回本周期。
[CI009, CI018, CI038, CI039, CI041]如果复用持续累积,成本重的首批部署可逐步走向更好利润率。
公司披露了业务构成,但没有披露单项目成本;节点展示方向,不是审计值。
[CI003, CI008, CI018, CI038, CI039, CI041]4.3 融资形成、估值路径与充足性
融资是本文件中记录最完整的部分。官方层面,Mech-Mind 称已完成 Series C+ 轮并累计融资 $300 million。CB Insights 显示的累计金额较低,为 14 轮 $222.36 million,但它也列出用户要求我们验证的关键节点:2022 年 8 月 Series D 融资 $38 million,估值 $858 million。后续来源对轮次规模的证据强于估值证据。Marketscreener 和多家中文媒体报道,2025 年 8 月 Series E-II 约 RMB500 million,投资团包括 Zhongshan Broad-Ocean Motor、CICC Porsche、Haihe、Hebei Structural Reform、Shanghe、Nanxiang、Tianjin Venture Capital、China Growth Capital,以及 Xiong'an 相关基金。到 2025 年末,多家媒体把累计融资描述为大约或超过 RMB2 billion。这足以说明公司带着相当充足的资产负债表支持进入 IPO 前窗口,但不足以计算现金跑道。我们没有找到关于账上现金、月度烧钱速度、债务或项目融资义务的已审阅披露。也没有找到权威公开来源能清楚确认 2022 年后估值超过 $1 billion;低声誉 IPO 前评论指向约 RMB8 billion,但这仍是报道估计,而不是已验证的升值轮。因此,资本故事强,但 2022 年 Series D 之后的估值路径仍部分不透明。[CI021, CI022, CI023, CI024, CI025, CI026]
| 主题 | 公开事实 | 已知信息 | 财务含义 | 尽调问题 |
|---|---|---|---|---|
| 官方累计融资 | USD 300M | 官方关于页面称 C+ 轮已完成,累计融资 USD300M | 显示资本基础较厚 | 将官方数字与股权结构表和各轮募资款核对 |
| 数据库累计融资 | USD 222.36M | CB Insights 列出 14 轮,累计融资更低 | 数据供应商口径不一致,增加可比公司和稀释分析难度 | 提供逐轮核对表 |
| 最新融资 | CNY 500M | 多家 2025 年来源称,2025 年 8 月 E-II 轮 / 新一轮融资接近 RMB500M | 定性看,显著拉长现金跑道 | 提供交割日期、新股与老股交易拆分,以及账上现金 |
| 经验证的 2022 年估值点 | USD 858M 投后估值代理 | CB Insights 列示 2022 年 8 月 Series D 融资 $38M、估值 $858M | 可作为估值路径锚点 | 提供 2022 年以来董事会批准的估值历史 |
| 2025 年末媒体累计口径 | ≈RMB 2B+ | 多篇媒体文章称累计融资约 RMB2B 或以上 | 符合融资密集型 Pre-IPO 公司的特征 | 披露准确累计股权融资及任何债务融资 |
| 现金 / 烧钱速度 / 现金跑道 | 已审来源均未披露现金余额、烧钱速度、现金跑道月数、债务或项目融资义务 | 无法靠公开证据计算现金跑道 | 提供月度烧钱速度、净现金、债务明细和最低现金约束条款 |
资本显然充足,但数据供应商对累计总额口径不一致,已审来源也没有披露现金、烧钱速度、现金跑道或债务。
[CI021, CI022, CI024, CI025, CI026, CI028]公开来源给出的融资和估值输入边界,显示已验证事实与报道说法之间的缺口仍然很大。
低 / 高边界来自不同来源类型。官方口径和数据库总额不一致,2022 年后的估值评论也没有监管文件支撑。
[CI021, CI022, CI026, CI028, CI029, CI030]近期资本可能如何投入研发、产品扩张、全球交付和 IPO 准备,而关键现金事实仍未公开。
该图只看方向,因为已审阅来源没有披露现金余额、债务或预算分配比例。
[CI024, CI025, CI031, CI033, CI034, CI040]4.4 IPO 考量与披露缺口
香港 IPO 叙事作为市场报道可信,但还不是由申报文件支撑的事实集。Yahoo Finance、The Standard、Ifeng 和 HKCD 转载的 Bloomberg 源报道称,Mech-Mind 正考虑或已秘密递交香港上市,目标约募资 $200 million;同时也强调时间和规模尚未最终确定,公司也未公开确认计划。这让投资人面对一种少见错配:证据足以相信资本市场准备已经启动,但披露还不足以像真正 IPO 候选公司一样承销业务。已审阅官方或类文件来源均未披露年度收入、ARR、毛利率、EBITDA、净利润、现金、烧钱速度、债务、递延收入或客户集中度。一篇 Eastmoney 的 IPO 前评论文章流传了 2024 年收入超过 RMB800 million、净利率超过 20% 的数字,但我们没有找到官方文件、招股书或主流类申报证据来交叉印证这些数字。注册中介显示,中国备案和财务账目在底层系统中某处存在,但本次可审阅材料并未公开。直到招股书或审计报表出现,正确财务结论是:Mech-Mind 有可信的产品市场牵引力和充足近期融资,但仍有太多仅限私下披露的指标,无法精确承销估值。[CI035, CI036, CI037, CI040, CI042]
| 缺失指标 | 公开状态 | 重要性 | 最佳公开代理指标 | 尽调路径 |
|---|---|---|---|---|
| 收入 / ARR | 官方未披露 | 无法分析收入倍数或增长质量 | 只能用安装基数、订单增长和客户数代理 | 索取按产品和地区拆分、经审计的历史收入 |
| 毛利率 / BOM | 未披露 | 卡住单位经济性和规模毛利率分析 | 硬件-软件-服务组合,加上 2021 年研发和支持披露 | 索取按产品线拆分的毛利率和服务附加 |
| 现金余额 / 烧钱速度 / 现金跑道 | 未披露 | 无法分析资金充足性和下一轮触发点 | 只有大额累计融资 | 索取月度烧钱速度、账上现金和 24 个月预测 |
| 客户集中度 / 续约 | 未披露 | 是渠道和收入耐久度投资判断的关键 | 100+ 家 Fortune 500 客户,以及头部客户风险评论 | 索取前 10 大客户收入占比,以及续约 / 扩张率 |
| 实现价格 / 折扣 | 未披露 | 无法建模正常化 ASP 和利润率 | 只有 2021 年相对定价这一历史信号 | 索取当前报价、渠道折扣和维护条款 |
| 债务 / 或有负债 | 未披露 | 可能改变企业价值和 IPO 准备度 | 登记中介提到备案,但没有公开明细 | 索取债务、担保、诉讼和进口依赖披露 |
最大缺口不是 Mech-Mind 是否跑出牵引力,而是私营公司披露在投资者能搭建传统 IPO 模型之前就断了。
[CI034, CI035, CI036, CI037, CI039, CI042]05产品与技术
5.1 产品定义与模块地图
最适合把 Mech-Mind 的产品公开材料理解为一套分层工业自动化栈,围绕料箱拣选、机床上下料、定位、装配、拆垛、码垛和在线检测等具体流程销售。2026 年公开材料显示,公司产品组合很宽但仍连贯:Mech-Eye 工业 3D 相机、Mech-Eye 3D 激光轮廓仪、Mech-Vision、Mech-Viz、Mech-DLK、Mech-MSR、Mech-Station InstaDepal,以及更新的具身智能「Eye-Brain-Hand」工作站。核心商业逻辑是,相机和轮廓仪家族采集可用 3D 数据,Mech-Vision 把数据转成位姿或检测结果,Mech-Viz 规划并仿真机器人运动,Mech-DLK 在传统视觉步骤不够时提供可训练 AI 模型。这套栈设计是真实差异化,因为它让 Mech-Mind 销售集成流程,而不是只卖传感器。主要提示是产品边界清晰度:当前官方英文和中文页面没有把 Mech-Recon 作为 2026 年独立页面展示,因此尽调仍需管理层解释,它仍是遗留术语、功能家族,还是内部范围模块,而不是对外销售 SKU。[CE001, CE004, CE005, CE006, CE007, CE009]
| 模块 / 资产 | 主要用户 | 公开成熟度 / 状态 | 技术差异化 | 关键尽调缺口 |
|---|---|---|---|---|
| Mech-Eye 工业 3D 相机 | 机器人集成商、自动化工程师、视觉团队 | 成熟 / 旗舰硬件系列 | 工作距离范围宽,耐环境光,3D 点云细节丰富,工业封装更坚固 | 需要按型号和应用提供实际吞吐量、缺陷率改善的独立证明 |
| Mech-Eye 3D 激光轮廓仪 | 检测、计量和质量工程师 | 成熟 / 专用检测产品线 | 4K 级线轮廓、最高 15 kHz 扫描率、微米级重复性、单次 HDR、支持 GenICam | 需要供应商和伙伴目录材料之外更广的第三方验证 |
| Mech-Vision | 视觉 / 应用工程师和工厂操作员 | 成熟 / 核心感知层 | 无代码 UI 将 3D 处理、深度学习、机器人通信和生产部署放在同一环境里 | 1000+ 款机器人模型、1-2 天调试、>99.99% 识别率等说法,需要客户层面验证 |
| Mech-Viz | 机器人编程人员和系统集成商 | 成熟 / 控制规划层 | 流程图式机器人编程、一键仿真、运动规划、碰撞检测、多 TCP 和拣选策略 | 需要证据说明复杂工作站里还有多少逻辑会迁移到机器人原生代码 |
| Mech-DLK | AI / 视觉专家和检测团队 | 扩张中 / 商业 AI 层 | 打通数据管理到部署的工作流,支持级联模型、多语言 SDK 和快速标注工具 | 公开性能指标来自公司自称,尚无独立基准测试 |
| 具身智能 Eye-Brain-Hand 工作站 | 具身 AI 试点、零售 / 物流创新团队、先进自动化买家 | 新兴 / 由发布驱动的扩张层 | 将 Mech-Eye、Mech-GPT 和 Mech-Hand 组合成通用感知-推理-执行栈 | 公开证据在演示和发布节奏上较强,但在手册、生产参考和支持边界上更薄 |
| Mech-Recon 公开足迹 | 解决方案架构师,以及想梳理技术栈的买家 | 不清楚 / 未以独立 2026 SKU 形式出现 | 该词在检索上下文中出现,但在保留的 2026 年官方产品页中,并未以文档清晰的独立 SKU 形式出现 | 管理层应说明 Mech-Recon 是旧品牌、功能族,还是非公开模块 |
状态标签反映公开文档深度和部署证据,不代表未披露收入结构或内部模块附加。
[CE001, CE004, CE006, CE007, CE009, CE014]公开产品架构把 Mech-Mind 从传感硬件延展到感知、规划、学习和具身 AI 扩展。
该运营栈来自公开产品、文档和发布材料,不是内部工程图。
[CE001, CE007, CE009, CE014, CE021, CE023]5.2 架构与运行流程
其运行模式从感知到动作,比多数工业视觉供应商公开得更清楚。Mech-Eye 相机和激光轮廓仪采集深度数据或轮廓数据;Mech-Vision 提供无代码感知环境,处理 3D 处理、匹配、深度学习支持的识别、机器人通信和生产部署;Mech-Viz 再用碰撞检测、抓取策略、多 TCP 和机器人品牌抽象来仿真并规划运动;Mech-DLK 则为更困难的识别或缺陷检测场景处理数据集管理、标注、训练、验证、级联和部署。中文产品页面在这里尤其具体,声称支持 1000+ 款机器人型号、1-2 天通信调试,以及通过生产界面实现分钟级换型流程。Eye-Brain-Hand 是传统栈之上的扩展层:它把 Mech-Eye、Mech-GPT 和 Mech-Hand 组合起来,面向更泛化的具身流程;AW 2026 证据显示,这一概念正进入透明物体搬运、人形机器人货架拣选和标准化拆垛单元。结果是一套可信架构:相机硬件、AI 推理和机器人执行,被设计成一个部署包落地。[CE010, CE011, CE012, CE014, CE015, CE017]
| 用户任务 | 当前工作流 | Mech-Mind 流程 | 可衡量 / 公开证明 | 限制 |
|---|---|---|---|---|
| 随机料箱或单件拣选 | 在没有手工夹具设计的情况下,定位并抓取多样、反光或紧密堆叠的物品 | Mech-Eye 捕获深度 -> Mech-Vision 识别并输出位姿 -> Mech-Viz 规划无碰撞抓取 | UR 市场和官方页面将该栈定位于料箱拣选、单件拣选和未知物体场景,覆盖 300-3500 mm 相机范围 | 公开证明最强的是能力声称,独立周期时间分布较弱 |
| 机床上下料与定位 / 装配 | 在工厂现场条件下寻找位姿、对齐机器人并精准放置零件 | 相机或激光数据输入视觉匹配和路径规划;面向 ABB/KUKA/FANUC/UR/Yaskawa/Kawasaki 控制器的品牌适配器负责执行 | 官方文档公布已测试控制器版本,以及 KUKA 专属标定 / 编程参考 | 未支持的控制器版本仍可能需要供应商支持和定制调试 |
| 拆垛 / 码垛 | 无需硬编码工作站,即可处理混合托盘、不稳定堆叠和 SKU 变化 | 经典相机 + 软件栈,或较新的 Mech-Station InstaDepal 工作站,处理位姿检测和托盘工作流 | AW 2026 页面称 InstaDepal 支持 30 分钟部署、随机到货和新 SKU | 30 分钟数字是供应商发布声称,而非第三方调试证据 |
| 检测 / 测量 | 以产线速度捕捉小缺陷、接缝、胶线和尺寸偏差 | 激光轮廓仪加 Mech-MSR / Mech-Vision 管线,生成点云、CAD 对比、OCR 和缺陷测量 | 轮廓仪宣传每条轮廓 4,096 点、15 kHz 扫描率和微米级重复性 | 多数性能证明来自官方或伙伴,而非基准实验室审计 |
| 透明或反光物体处理 | 为通常会打断结构光或检测工作流的物体成像 | 新的透明物体成像,加上具身「Eye + Brain」演示,将技术栈延伸到包装商品、瓶子、易拉罐和试管 | AW 2026 发布材料强调 ULTRA M-GL、亚毫米级透明物体拣选和泛化透明物品处理 | 现有证据偏发布和演示,还没有配套的大范围 GA 规格表或客户参考集 |
| 复杂识别 / 深度学习缺陷检测 | 在规则视觉不够用的重叠物体、OCR、分割或异常场景中训练模型 | Mech-DLK 负责标注、训练、验证、部署和 SDK 导出;Mech-Vision 随后把训练好的模型落到生产流程 | 中文材料称平均推理约 10 ms,比同业快约 40%,过杀率和漏检率较低 | 这些质量和速度数字来自公司说法,需在客户数据上复测 |
收益项混合了官方和伙伴描述,只有少量第三方佐证;不是标准化基准测试结果。
[CE009, CE010, CE012, CE014, CE020, CE022]代表性部署流程:从 3D 采集到位姿生成、机器人规划、执行和生产侧反馈。
该流程把公开的 Mech-Eye + Mech-Vision + Mech-Viz + Mech-DLK 工作流压缩成一个通用运行序列。
[CE009, CE010, CE014, CE018, CE026, CE028]5.3 集成与开发者接口
集成深度是 Mech-Mind 最强的公开技术信号之一。官方标准接口文档不只是点名机器人品牌,还列出 ABB、FANUC、KUKA、UR、Yaskawa、Kawasaki 等品牌中已测试的控制器系列和版本;仅 KUKA 分支就包括自动标定、示例程序、命令和错误信息参考。伙伴侧,Universal Robots 通过其市场分发 Mech-Mind,并明确描述一个即插即用的 URCap,可把 Mech-Eye 加 Mech-Vision 加 Mech-Viz 套件转成可部署的 UR 流程。对工业 3D 视觉来说,开发者接口异常开放:Mech-Mind 发布 GitHub 组织、链接 C++、Python、ROS、ROS2 和 HALCON 示例的社区帖子,提供带 Ubuntu/OpenCV/PCL 前置条件的 ROS 1 和 ROS 2 接口,以及显示支持 GenICam、基于 GigE 采集和第三方机器视觉互操作的手册材料。话虽如此,同一套文档也暗示部署仍依赖控制器选项、SDK 版本、配方调试和现场集成能力;因此,开放叙事降低了锁定效应,但没有消除实施复杂度。[CE027, CE028, CE029, CE030, CE031, CE032]
| 层级 / 组件 | 技术栈内作用 | 关键依赖 | 主要风险 |
|---|---|---|---|
| Mech-Eye 面阵相机 | 采集大视场 3D 点云,用于引导、定位和装配 | 光学、结构光成像、环境光处理,以及按工作距离选型 | 硬件已成熟,但表现仍高度依赖材料属性和部署条件 |
| Mech-Eye 激光轮廓仪 | 生成线扫深度数据,用于在线计量和缺陷检测 | 稳定的编码器 / 控制集成、HDR 设置和 ROI 调参 | 检测部署仍可能受安装设置、轮廓质量和软件参数影响 |
| Mech-Vision 感知层 | 运行 3D 处理、匹配、深度学习步骤、机器人通信和生产部署 | 项目步骤配置、数据集、参数配方和相机输入质量 | ROI、阈值或数据流配置不佳,可能导致无结果、超时或无效输出 |
| Mech-DLK 学习层 | 管理高难识别任务的标注、训练、验证、部署和级联 | 高质量训练数据、版本控制,以及接入 Mech-Vision 或自定义软件 | 公开指标能强化营销信号,但没有经过跨工作负载的独立基准测试 |
| Mech-Viz 规划 / 控制层 | 把机器人程序抽象成流程图,并配合仿真、碰撞检查和抓取规划 | 机器人品牌适配器、控制器选项、TCP 设置和运动规划约束 | 奇异点、无效抓取点或碰撞仍可能需要调参和厂商支持 |
| 标准接口 + 品牌适配器 | 把 Mech-Mind 输出接到 ABB、FANUC、KUKA、UR、Yaskawa、Kawasaki 等机器人 | 受支持控制器版本、必需的控制器软件选项和品牌专用手册 | 覆盖面确实广,但版本不匹配或文档缺口会拖慢调试上线 |
| SDK / ROS / GenICam 接口面 | 让外部软件完成自定义集成、参数控制和点云访问 | SDK 兼容性、Ubuntu/ROS 包、OpenCV/PCL 依赖和网络配置 | 开放接口降低锁定效应,但也把部分实施负担转给集成商 |
| Eye-Brain-Hand 扩展 | 在经典技术栈之上加入具身推理、自然语言驱动交互和灵巧操作 | 发布阶段的硬件 / 软件成熟度、Mech-GPT 表现和真实部署参考 | 公开证据仍比老一代相机 + 视觉 + 规划层更浅 |
这张架构表不是内部系统图,而是根据产品页、手册、SDK 文档和伙伴集成证据拼出的外部可见运营模型。
[CE007, CE008, CE010, CE014, CE015, CE018]5.4 成熟度、质量控制与产品风险
产品成熟度不只由营销文案支撑。官方公开材料声称有 24,000+ 台相机、100+ 家 Fortune Global 500 客户,并在约 50 个国家和地区运营;KR-Asia 和 Eastmoney 也独立描述了 15,000+ 个安装案例、本地化实验室、区域工程团队和自有相机工厂。公开质量信号同样有意义:相机和轮廓仪页面发布 IP65/IP67 以及广泛 EMC/安全认证,手册披露 GenICam 和 SDK 发布说明,故障排查文档列举了真实运行故障模式,如执行超时、相机连接失败、运动奇异点、无效抓取点和机器人碰撞。这些都是平台部署足够多之后积累运营边缘案例的特征。风险在于公开证明仍不对称。Mech-Mind 对硬件耐用性、机器人适配器和展会发布的记录,强于对软件安全架构、SLA 边界,或经独立验证的吞吐量和精度基准的记录。Eye-Brain-Hand 显然在推进,但其公开证据基础仍更多来自新闻,而不是手册;Mech-Recon 在保留的公开语料中仍文档不足。[CE002, CE003, CE005, CE020, CE022, CE036]
| 控制项 / 信号 | 公开状态 | 范围 | 缺口 / 风险 |
|---|---|---|---|
| 硬件环境防护 | 工业 3D 相机公开宣传 IP65,激光轮廓仪公开宣传 IP67 | 工厂现场粉尘、湿度和严苛部署耐受性 | 硬件耐用性无法回答更大技术栈的软件安全或正常运行时间问题 |
| 安全 / EMC 认证 | 相机和轮廓仪页面披露 CE、FCC、VCCI 以及更广的区域认证组合 | 电气 / EMC 采购与跨境部署准备度 | 认证深度集中在硬件,软件证明细节没有同等公开 |
| 耐久性 / 可靠性 | 官方相机材料披露 MTBF >= 100,000 hours,伙伴目录提到长时间连续测试 | 长周期硬件部署 | 这些仍是厂商阵营来源,不是现场故障率数据集 |
| 控制器兼容性表 | 标准接口文档列出主要机器人品牌已测试的控制器版本和必需选项 | 调试上线的可预测性和可支持性 | 未列出的版本也可能可用,但文档明确提醒通信可能失败,且可能需要支持 |
| 运行时故障排查面 | 状态码文档列举执行超时、相机连接失败、运动奇异、抓取点无效和机器人碰撞等状态 | 运营调试和安全调参 | 公开文档显示成熟度,也暴露集成商仍要管理大量故障模式 |
| 软件 / 云信任披露 | 与硬件和集成文档相比,公开披露偏薄 | 软件栈的安全架构、SLA、租户模型、补丁节奏和证明材料 | 买方仍需要信任中心材料、问卷或直接尽调,才能排除企业软件风险 |
这张信任表聚焦公开硬件质量、集成安全和披露面;不能替代安全问卷、工厂验收测试或 SLA 审查。
[CE005, CE027, CE036, CE037, CE041, CE050]| 日期 / 阶段 | 发布 / 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| 2023-05 | 社区帖链接了 C++、Python、ROS、ROS2 和 HALCON 示例仓库 | 历史 / 公开开发者触点 | 说明 Mech-Mind 在 2025-2026 年具身 AI 推进前,已经发布多语言示例 | 社区示例程序页面 |
| 2024 | KR-Asia 称 Mech-MSR 已加入产品线 | 历史 / 产品线扩张 | 显示公司从机器人引导切入专用 3D 测量和检测软件 | KR-Asia 公司简介 |
| 2024(手册 v2.4.0) | 激光轮廓仪手册记录 SDK 2.4.0 发布说明,包括 ROI 和外部信号改进 | 历史 / 技术维护 | 说明硬件软件工具已有稳定维护节奏 | 激光轮廓仪手册 |
| 2025-10-15 | mecheye_ros_interface 示例已更新到 SDK 2.5.1 | 公开维护信号 | 2026 发布周期前不久,开发者触点仍在刷新 | GitHub 发布页面 |
| 2025-12 | 官方新闻索引出现 iREX 2025 和新智能 2D 相机发布标记 | 近期 / 发布节奏 | 说明 AW 2026 之前,Mech-Mind 已在老款相机和软件产品线之外扩展 | 官方新闻索引 |
| 2026-02-02 | AW 2026 邀请页面 | 近期 / 预发布 | 确认公司围绕 Eye-Brain-Hand 和新品策划了 2026 发布活动 | 官方新闻索引 |
| 2026-03-12 | AW 2026 回顾:ULTRA M-GL、AIC-Lite GL 和 Mech-Station InstaDepal | 近期 / 已宣布并预览 | 显示路线图在推进,但部分发布项仍被描述为即将推出或今年晚些时候推出 | AW 2026 官方新闻 |
这张路线图表只记录公开发布标记和技术维护信号;内部发布列车、GA 承诺和 SKU 退市决策仍未披露。
[CE024, CE025, CE026, CE034, CE043]Mech-Mind 部署能否真正成功,取决于硬件匹配、数据质量、机器人适配器兼容性和集成商执行。
该依赖图强调外部可见的实施依赖,而不是每一项内部软件服务。
[CE027, CE028, CE030, CE036, CE037, CE038]基于公开证据,将 Mech-Mind 主要产品面的成熟度按 0-10 编制指数。
评分反映公开文档深度、集成证据、第三方佐证和发布成熟度,而不是内部 KPI。
[CE003, CE020, CE027, CE034, CE042, CE043]5.5 产品与技术结论
对 Mech-Mind 最好的判断是,它已经搭出一套成熟的传统 3D 视觉引导机器人工业栈,如今正试图借这套装机基础向更广的具身 AI 版图爬升。成熟层证据充分:加固 3D 硬件、激光轮廓仪、无代码感知和机器人规划软件、按控制器拆分的文档、UR 市场包装、ROS/SDK/GenICam 接口,以及围绕开发者资产的公开维护信号。这个组合让公司不只是相机供应商,也给了它差异化的集成护城河。 尽调负担落在下一步,以及仍被隐藏的部分。公开材料在周期时间、检测精度和部署速度上大量依赖供应商说法;在软件安全、SLA 或按模块实现的附着上没有同等深度;Mech-Recon 状态也不清楚。结论:Mech-Mind 在相机加视觉加规划栈上看起来技术差异化且商业成熟;Eye-Brain-Hand 和较新的 2026 年发布有前景,但在管理层拿出正式可用规格、客户证明,以及超出发布新闻层的架构和安全细节之前,仍应把它们作为扩张期权承销。[CE001, CE023, CE036, CE043, CE044, CE045]
06客户情况
6.1 客户足迹与垂直行业组合
Mech-Mind 的公开客户图景在总量上很广、在流程上很具体,但具名账户披露偏薄。当前英文和中文官方材料都声称已部署 24,000+ 个单元、拥有 100+ 家 Fortune 500 客户,并覆盖约 50 个国家或地区。第三方资料在方向上印证了这一规模,尽管较旧第三方快照仍引用 10,000+ 台相机或 15,000+ 个安装案例;因此,解读装机基数的正确方式是把它看成移动时间序列,而不是单一审计分母。 垂直行业组合比客户名称更清楚。官方英文页面聚焦汽车、物流、金属和机加工、电子、EV 电池,以及食品饮料。中文材料用汽车制造、物流搬运、重工业、轻工业、新能源和工业质检描述了大致相同的足迹。两种语言中,汽车看起来最深,其后是物流、电池或电子。公开证据因此支持真实的跨垂直部署广度,但没有提供干净的逐客户名单,让投资人按客户名称、地区或队列映射收入。[CU001, CU002, CU003, CU004, CU005, CU006]
| 细分市场 | 购买方 / 用户 / 付款方 | 公开用例 | 公开证据强度 | 战略价值 | 关键缺口 |
|---|---|---|---|---|---|
| 汽车制造 | OEM、一级供应商、工厂工程师、自动化团队 | 机床上下料、车轮 / 轮胎装配、焊接辅助、在线测量、玻璃涂胶 | 工作流广度高;具名客户少 | 公开最深的垂直行业,生产单元证据最强 | 没有公开名单区分 OEM 与供应商,也无法区分直销和集成商主导交易 |
| 物流 / 仓储搬运 | 3PL、包裹枢纽、包装和履约运营商 | 箱 / 周转箱 / 袋装拆垛、包裹导入、电池和锭料搬运 | 中 | 证明能力可泛化到汽车之外,并支撑仓储自动化逻辑 | 抓取到的公开来源没有具名用户 |
| 电子 / 家电 | 家电厂商、电子制造商、QA 团队 | RJ45 检测、洗衣机装配、压缩机和空调部件搬运 | 中 | 显示其在轻型制造中的精度和检测价值 | 客户名称和扩张深度未披露 |
| EV 电池 / 新能源 | 电池厂商、EV 模组产线、回收团队 | 电芯拆垛、模组拆解、模组到电池包装配、EV 充电 | 中 | 贴近快速增长的 EV 供应链,具备战略邻近性 | 公开来源未点名 CATL、BYD 或其他电池龙头 |
| 金属、机加工和重工业 | 钢铁集团、机械工厂、机加工单元 | 钢板折弯、铁路车轮上下料、履带板装配、螺栓拧紧 | 中 | 验证严苛环境和反光金属用例 | 客户名称藏在匿名案例描述后面 |
| 食品饮料和轻工业 | 包装消费品生产商和物料搬运方 | 官方材料出现箱体搬运和拆垛,但详细案例较少 | 低于汽车 / 物流 | 显示重制造之外的多垂直行业潜力 | 公开深度更薄,也缺少具名客户证据 |
结合英文和中文官方垂直标签;证据强度指公开证据质量,不代表收入占比。
[CU001, CU002, CU003, CU007, CU008, CU009]| 指标 | 数值 | 日期 / 口径批次 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 已部署设备 | 24,000+ | 当前官方页面 | 官方英文和中文关于页 / 首页 | 高 | 大规模装机基数支撑真实部署广度 | 未区分活跃设备和累计设备 |
| Fortune 500 客户覆盖 | 100+ | 当前官方页面 | 官方英文和中文页面 | 高 | 显示已切入大型企业制造账户 | 没有公开名单、直接客户数或站点数 |
| 地理覆盖 | 接近 50 个国家 / 地区 | 当前官方页面 | 官方英文和中文页面 | 高 | 全球支持和部署足迹可信 | 未披露按地区拆分的收入或装机基数 |
| 早期 A3 基线 | 10,000+ 台相机;1,500+ 家客户;50+ 个国家 | A3 简介口径 | A3 公司简介 | 中 | 展示当前官方数字上调前的早期公开快照 | 简介日期和方法论不明确 |
| 中间外部快照 | 15,000+ 次安装 | 2025 | KrASIA | 中 | 连接较早目录数字和较新官方数字 | 没有经审计的换算关系,把安装量同设备数或客户数打通 |
| 支持足迹 | 中国、美国、德国、日本、韩国 | 当前 | 官方页面和展会材料 | 高 | 支撑核心市场的售后服务和伙伴赋能 | 未披露各地区客户数 |
各行混合了公司声称和第三方快照;公开指标没有按统一定义审计。
[CU001, CU002, CU003, CU004, CU005, CU006]6.2 地理分布与渠道模式
当前官方材料明确显示 Mech-Mind 的客户覆盖全球,并列出中国、美国、德国、日本和韩国的办公室或设施。这一点重要,因为工业自动化里的客户证明不只看装机基数,还要看上线后部署、服务、培训和故障排查能否在本地处理。Mech-Mind 的中文和英文公开材料,加上 KrASIA,都指向一种模式:中心化验证产品,叠加本地化商业和支持存在。 商业化模式看起来更偏伙伴驱动,而不是纯直营。KrASIA 称区域系统集成商和代理负责部署与推广,Mech-Mind 专注产品开发和售后支持。中文官方关于页面又补充了面向集成商伙伴的培训、参考方案设计、展会支持和项目支持。UR+ 会员身份,以及在 Automate 和 iREX 的反复亮相,也符合这一模式:Mech-Mind 正搭建一个生态,终端客户触达很大程度上由机器人平台、本地集成商、展会或目录发现渠道中介,而不只是直接标杆客户营销。[CU025, CU026, CU027, CU028, CU029, CU030]
| 合作伙伴 / 渠道 | 角色 | 公开证据 | 客户影响 | 关键限制 |
|---|---|---|---|---|
| 区域系统集成商和代理商 | 部署、推广、本地实施 | KrASIA 本地化模式 | 上线更快,并提供本地语言支持 | 未披露合作伙伴质量和集中度 |
| Universal Robots / UR+ | 机器人平台集成 | UR+ 会员身份和即插即用 URCap 表述 | 降低协作机器人场景的集成摩擦 | 公开伙伴证据不包含客户数量 |
| 文档与最佳实践门户 | 培训和实施支持 | docs.mech-mind.net 教程和案例实践手册 | 提高合作伙伴和客户的复现能力 | 缺少公开使用量或完成率数据 |
| 全球办公室和东京设施 | 销售、工程、培训、展陈支持 | 官方英文 / 中文页面和 KrASIA | 改善核心出口市场的售后支持 | 未披露区域收入结构 |
| A3 / Automate / HowToRobot 目录 | 发现与线索获取 | 独立目录和展会名录 | 帮助中国以外买家和合作伙伴找到 Mech-Mind | 进入目录不等于已成交需求 |
| 展会发布(AW 2026 / iREX 2025) | 销售管线创建和生态信号释放 | 官方展会报道 | 支撑北美和日本的新客户、新伙伴获取 | 展会线索转为活跃客户的转化率未公开 |
合作伙伴生态行聚焦部署和支持机制,而不是各渠道对财务的贡献。
[CU026, CU027, CU028, CU029, CU030, CU031]公开证据显示,从市场发现到工作站部署,路径由合作伙伴主导;最大断点在具名客户证明和续约可见度。
序列是分析性梳理,不是数字漏斗;它概括公开来源中看到的商业旅程。
[CU026, CU027, CU028, CU029, CU032, CU038]6.3 具名客户验证与证明质量
客户侧最尖锐的问题,不是 Mech-Mind 是否有工业部署,而是标杆客户名称能否被公开验证。正面来看,官方案例库规模大且运营细节具体。负面来看,公司自有页面压倒性地描述匿名工厂、零部件或流程,而不是点名买方。这意味着,具名客户证明远弱于装机基数或垂直广度证明。 用户特别要求验证 BMW、SAIC、Geely 等名称,而不是默认假设。已审阅公开证据没有把 BMW 验证为 Mech-Mind 的具名客户:Mech-Mind 本次审阅的自有页面没有点名 BMW,抓取到的 BMW 相邻来源指向的是 UR 和 Hexagon,而不是 Mech-Mind。SAIC 和 SAIC-GM 情况类似:2026 年公开电池线部署报道点名 Zhiyuan/Nengzai,而不是 Mech-Mind。Geely、CATL、BYD、Foxconn 和 JD Logistics 在抓取集合中仍未获公开验证。结果是,行业适配和客户名称证明之间必须清晰区分。[CU015, CU016, CU017, CU018, CU019, CU020]
| 客户 / 队列 | 细分市场 | 公开部署 / 用例 | 量产 / 试点 | 结果 / 信号 | 局限 |
|---|---|---|---|---|---|
| BMW Group | 汽车 OEM | 已审阅的 BMW 公开机器人案例涉及 UR SortBot 和 Hexagon 人形机器人试点;抓取来源未点名 Mech-Mind | BMW 确有生产自动化;与 Mech-Mind 的关系未验证 | 说明 BMW 是该品类的现实买方类型 | 相邻的客户侧证据无法确认 Mech-Mind |
| SAIC / SAIC-GM | 汽车 OEM / EV 电池线 | 2026 年 Buick 电池线报道点名 Zhiyuan/Nengzai,而非 Mech-Mind | 确有生产自动化;与 Mech-Mind 的关系未验证 | 说明 SAIC 对具身和视觉自动化有需求 | 抓取来源指向另一家供应商,不是 Mech-Mind |
| Geely / 吉利 | 汽车 OEM | 抓取到的官方或客户侧来源没有把 Mech-Mind 与 Geely 同时点名 | 未验证 | 除了行业契合外,没有更多信号 | 缺少公开证据并不证明没有商业关系 |
| CATL / BYD / Foxconn / JD Logistics 队列 | 电池 / 电子 / 物流龙头 | 抓取到的 Mech-Mind 官方页面未点名这些账户 | 未验证 | EV 电池、电子和物流用例显示行业契合 | 需要在 NDA 下做客户访谈核验 |
| 100+ Fortune 500 队列 | 全球工业企业 | 英文和中文关于页 / 首页给出的官方汇总说法 | 汇总层面可能已进入生产 | 是跨地区、跨垂直行业的有用规模信号 | 没有公开名单或活跃账户分母 |
| 匿名汽车和重工业工厂 | 一级供应商 / 重工业 | 官方案例页面反复展示机床上下料、焊接、检测和搬运工作流,但隐去客户名称 | 生产单元证据 | 证明真实部署存在的最强公开证据 | 客户身份、合同规模和续约路径仍不透明 |
这张表有意按证据质量拆分:各行区分具名账户核验、相邻证据和汇总证据。
[CU019, CU021, CU022, CU023, CU024]矩阵把真实工业适配度和实际具名客户验证拆开;后者是本章最大的公开证明短板。
矩阵单元格概括证明质量,不代表收入或部署量。
[CU015, CU019, CU021, CU022, CU023, CU024]6.4 部署深度、流程广度与扩张信号
公开证明在应用层面最强。仅汽车页面就覆盖机床上下料、车轮装配、轮胎搬运、制动和车轴搬运、冲压件上架、玻璃涂胶、在线测量和焊接支持。物流页面覆盖纸箱、周转箱、袋装物和铅酸电池拆垛,以及包裹导入。EV 电池、电子和金属页面又显示模块拆解、module-to-pack 装配、连接器检测、家电装配、钢板折弯和红热车轮上下料等额外广度。 这种广度重要,因为它暗示 Mech-Mind 在客户工厂内部按流程扩张,而不是靠公开客户名称营销扩张。官方 2025 年汽车在线检测新闻尤其重要:它说明 Mech-Mind 正从机器人引导任务进入汽车产线相邻的质量和测量流程。即便如此,公开的工厂间铺开或账户扩张证据仍大多是间接的。公司证明了其产品正在被用于许多流程类别,但没有公开展示许多具名账户如何随时间从试点走到生产,再走向多站点扩张。[CU009, CU010, CU011, CU012, CU013, CU014]
漏斗在公开工作流证明和具名客户扩张或续约的公开证据之间收窄最明显。
流程阶段是定性的,来自已发布的渠道、案例研究和支持证据,而非已披露 CRM 漏斗数据。
[CU025, CU026, CU027, CU030, CU031, CU033]6.5 耐久性、留存与集中度提示
最大的剩余客户风险问题是耐久性和集中度。本次抓取到的官方或第三方资料来源都没有披露 NRR、GRR、流失率、续约率、最大客户占比或前 10 大客户集中度。这并不意味着业务缺少留存;它意味着当前公开记录无法让投资人衡量留存。同样的问题也适用于集中度。汽车看起来是公开最深的垂直行业,中国仍是公司的验证基础,但任何单一客户、垂直行业或地区的收入暴露都没有公开量化。 含义是,公开尽调可以支持一个中等信心判断:Mech-Mind 拥有真实跨垂直部署和可运作的伙伴主导支持模式。没有私下证据,它无法支持对客户质量、续约耐久性或集中度风险的高信心判断。本章的正确结论因此是二分的:采用证明真实,但具名账户承销仍不完整。任何投资观点都应要求客户访谈、头部客户明细、续约数据,以及成功多站点或多产品扩张案例,然后再把总量客户说法视为完全可承销。[CU024, CU036, CU037, CU038, CU042]
| 指标 | 数值 | 细分市场 | 置信度 | 尽调问题 |
|---|---|---|---|---|
| 净留存率(NRR) | 公司整体 | 低 | 索取按年份和前 20 大客户队列拆分的 NRR | |
| 总留存率(GRR) | 公司整体 | 低 | 索取按队列拆分的续约和收缩情况 | |
| 客户流失 | 具名公开参考客户 | 低 | 询问是否有主要公开参考客户在试点或首次部署后流失 | |
| 合同期限 / 续约时长 | 企业项目和集成商主导项目 | 低 | 索取 MSA 样本 / 支持条款和续约节奏 | |
| 多站点扩张证据 | 只能通过装机基数和地理覆盖说法间接推断 | 跨垂直行业 | 中 | 提供 3 个跨工厂或跨国家铺开的具名案例 |
| 公开满意度信号 | 未审阅到覆盖终端客户满意度的大规模评论语料 | 公开网页 | 低 | 用客户访谈替代基于抓取的满意度代理指标 |
空值表示抓取到的来源集中未公开披露,不代表数值为零。
[CU024, CU036, CU038]| 扩张驱动因素 / 风险 | 当前公开信号 | 影响 | 尽调路径 |
|---|---|---|---|
| 工作流切入后扩张 | 案例库覆盖同一工业垂直内的多类任务 | 如果单个工厂增加更多应用,增购潜力为正 | 索取单一账户内多应用扩张的具名案例 |
| 汽车行业集中度 | 汽车行业拥有最深的公开工作流案例库 | 收入可能集中在周期性强、资本开支重的买方 | 索取按头部垂直行业和头部 OEM / 供应商队列拆分的 ARR |
| 中国优先验证模式 | 产品先在中国验证,再向海外铺开 | 商业化效率高,但也可能意味着早期区域集中 | 索取中国与海外收入拆分 |
| 具名客户不透明 | 100+ Fortune 500 说法缺少公开名单 | 即使规模叙事成立,客户质量研判仍然偏弱 | 要求提供头部客户清单和可背书客户账户 |
| 集成商依赖 | 区域集成商和代理商负责部署与推广 | 交付质量和续约可能取决于合作伙伴能力 | 要求提供合作伙伴集中度、附加销售率和 SLA 归属 |
| 留存指标不透明 | 未公开 NRR、GRR、流失率或集中度数据 | 阻碍高置信度判断收入耐久性 | 要求提供队列续约、流失原因和未续约案例 |
每一行都把公开信号与支撑判断所需的具体私有证据对应起来。
[CU025, CU026, CU027, CU033, CU036, CU037]6.6 展示材料
07风险
7.1 风险优先级与传导
Mech-Mind 的风险画像不是由单一致命缺陷主导,而是由竞争压力、工业需求周期性和披露缺口相互作用推高。公开证据确认公司已有真实规模——部署 24,000+ 台设备、服务 100+ 家 Fortune Global 500 客户、足迹覆盖近 50 个国家——但可靠性和部署成熟度的最强证明仍主要来自公司自有材料。与此同时,独立市场资料把 Cognex、Keyence 等老牌厂商放在机器视觉采购决策的核心位置;中国 PMI 到 2026 年仍偏弱,外部分析也警示,中国放缓和贸易碎片化已经在重塑制造业需求。因此最高的剩余风险是商业执行被挤压:工业项目周期长、集成重,终端客户转谨慎时,项目最容易滑期;老牌厂商又有存量基础和渠道优势,而管理层还希望投资人承销一个没有公开审计收入、客户集中度或赢单 / 输单披露的 IPO。图 R001 和 R002 对风险栈排序,并展示外部冲击如何传导到收入、利润率和融资结果。[CR001, CR016, CR028, CR038, CR040, CR042]
综合风险最高象限包含既有厂商 / 渠道压力、出口管制暴露和客户集中度不透明;宏观和 IPO 不确定性略低一档,因为它们如何传导仍取决于尚未披露的信息。
[CR016, CR028, CR034, CR038, CR042, CR045]外部冲击不会孤立打到 Mech-Mind:中国产业需求偏弱、出口管制摩擦和既有厂商降价,会先传导为更长试点、更低转化和更高融资压力,最后才表现为估值风险。
[CR018, CR028, CR034, CR038, CR039, CR040]7.2 监管、法律与资本市场风险
Mech-Mind 依赖自己无法左右的政策时,监管风险最实质。2026 年 1 月 BIS 规则没有消除出口管制风险;该规则只是把一刀切假定改为某些先进半导体可逐案审查的狭窄路径,同时保留繁重的认证、了解客户(KYC)、测试和反转移义务。公开资料没有披露 Mech-Mind 当前计算硬件物料清单(BOM),但任何对美国来源先进 AI 芯片或相关模块的依赖,都会落在一个可能迅速再次收紧的政策框架下。资本市场风险同样只是悬而未决,并未解决。HKEX Chapter 18C 明确给机器人发行人留出通道,但该路径仍取决于经审计收入或未商业化门槛;公开传闻包只说 Mech-Mind 已秘密递表香港 IPO,希望募资约 US$200 million,最终条款仍未确定。Aiqicha 还给出另一项法律悬挂信号:Xiong’an 主体显示多起商业纠纷和诉讼相关记录。上述事项都不能证明投资逻辑今天断裂,但意味着 IPO 叙事应被视为有条件的期权,而不是已去风险的流动性。[CR028, CR029, CR030, CR031, CR032, CR033]
| 风险 | 司法辖区 | 状态 / 证据 | 可能性 | 严重性 | 缓解成熟度 | 残余敞口 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| 先进 AI 芯片出口管制和地缘政治收紧 | 美国 / 中国 | 2026 年 1 月,BIS 将部分芯片许可改为逐案审查,但仍保留严格的认证、测试和反转移条件;Mech-Mind 在 BOM 上的直接敞口未披露 | 中高 | 高 | 中低 | 高 | 获取当前算力 BOM、供应商地图和出口许可应急方案 |
| 香港第 18C 章 IPO 资格和时间不确定性 | 香港 | 机器人属于合资格特专科技行业,但公开资料看不出 Mech-Mind 是否达到经审计收入或替代上市门槛;IPO 报道称条款仍未确定 | 高 | 高 | 低 | 高 | 要求提供 FY2025/FY2026 经审计收入、投行备忘录和上市时间表 |
| 中国司法和商业纠纷压力 | 中国 | Aiqicha 列出雄安主体的多起商业纠纷、立案记录、开庭公告和诉讼关系,但公开资料里的案件层面背景很少 | 中 | 中 | 中低 | 中 | 收集完整案件进度、准备金影响、相对方和和解历史 |
| 海外扩张中的贸易、关税和本地化压力 | 美国 / 中国 / 出口市场 | 独立宏观资料指向贸易碎片化、关税压力和国家推动的科技自主;这些因素都可能改变采购、本地化和资本市场准入 | 中高 | 中高 | 中 | 中高 | 审查各区域销售敞口、本地化计划和关税转嫁假设 |
严重性同时看反向证据和披露缺口;政策依赖来自外部、且公司特定缓解措施未被公开记录时,残余敞口仍然高。
[CR028, CR029, CR030, CR031, CR032, CR033]7.3 运营可靠性、集成与客户不透明
运营问题不在于 Mech-Mind 能否演示先进感知——其自有材料显示,它在反光金属、透明容器、薄壁零件和在线检测上能力较强——而在于这些结果有多大概率在存量现场的规模化波动中站得住。公司可靠性博客明确写到,漂移、振动、灰尘、环境光和连续运行会降低相机精度,或在一段时间后导致故障。其他官方页面也承认,反光表面、折射、遮挡和几何形状变化会让识别和抓取失稳,把感知问题变成停机、返工和节拍丢失。Mech-Mind 的缓释措施有实质意义:IP65/IP67 硬件、最高 100,000 小时 MTBF 声称、漂移自动校正、仿真工具和无代码部署流程。但这些缓释措施大多由公司自己发布,公开资料中没有按行业展示正常运行时间、故障率或试点转量产转化率的现场队列数据。这个缺口重要,因为客户暴露仍不透明。官方页面列出汽车、物流、动力电池、重工业、电子和食品等核心部署领域,但没有公开来源披露头部客户依赖、行业收入拆分、续约率或按机器人品牌计算的附着率。[CR003, CR017, CR018, CR020, CR021, CR022]
| 失效模式 | 证据 | 可能性 | 严重性 | 缓解成熟度 | 残余敞口 | 未解决缺口 |
|---|---|---|---|---|---|---|
| 反光 / 透明 / 多品种零件场景的既有产线集成失败 | 官方博客反复把反射、折射、遮挡和环境光变化列为漏检或抓取不稳定的原因 | 高 | 高 | 中 | 高 | 按工作流拆分的上线后成功率以及报废 / 返工数据未公开 |
| 标定漂移和严苛工厂环境退化 | 公司材料明确把温度、振动、粉尘和连续运行时长列为漂移、精度下降或相机故障的来源 | 中 | 高 | 中高 | 中 | 未找到客户侧正常运行时间、质保或现场故障队列数据 |
| 概念验证和调试周期长 | 公开材料强调咨询、仿真、模板、培训和支持——这些有助于缓解风险,也说明部署工程量不小 | 高 | 高 | 中 | 高 | PoC 中位周期、试点转生产转化率和实施人力未披露 |
| 广泛客户标识营销掩盖客户集中度 | 官方页面列出多个行业和 Fortune 500 客户,但未披露头部账户、头部垂直行业或续约集中度指标 | 中高 | 高 | 低 | 高 | 需要前 10 大客户、前 5 大垂直行业、区域收入拆分和留存数据 |
| 全球支持能力相对装机规模承压 | 公司称已部署 24,000+ 台、拥有 600+ 名员工和全球办公室,但现场服务密度和 SLA 表现未公开 | 中 | 中高 | 中 | 中 | 需要区域工程师人数、响应时间 SLA 和升级处理统计 |
各行区分公司声称能缓解的事项和外部投资人能独立验证的事项;公开队列数据为空本身就是核心风险。
[CR003, CR017, CR018, CR020, CR021, CR022]7.4 老牌厂商、渠道依赖与周期性需求
竞争风险在结构上偏不利,因为 Mech-Mind 攻入的工作流本来就在老牌厂商的采购地图里。Cognex 的 10-K 显示其对物流、汽车、消费电子和大中华区有实质暴露;Keyence 的 3D VGR 产品页显示,自动标定、CAD 导入、路径规划和仿真并非 Mech-Mind 独有。独立排名和市场报告同样把 Cognex、Keyence、Basler、Omron、机器人 OEM 和系统集成商生态放在机器人视觉选型的中心。重要的是,Mech-Mind 的分销模式也依赖合作伙伴。KrASIA 称公司通过本地系统集成商和代理商合作;公开的价值提供商材料和联系页面显示,其支持网络分散在欧洲、美国、日本、韩国、北京和上海。资本开支强劲时,该模式能放大触达;资本开支转弱时,它可能拉低服务质量,并拉长本就顾问式的销售周期。宏观证据并不温和。中国 PMI 在 2025 年末和 2026 年 1 月再次跌破 50;NBR 认为,中国放缓和贸易关系恶化会冲击制造业出口商,并迫使它们多元化。对 Mech-Mind 而言,汽车、电子或物流客户的资本开支暂停,可能与老牌厂商降价、合作伙伴执行滑坡合并成一条下行闭环。[CR006, CR007, CR009, CR010, CR011, CR012]
| 依赖项 | 交易对手 / 市场 | 角色 | 集中度 | 失效场景 | 严重性 | 缓解措施 | 残余敞口 |
|---|---|---|---|---|---|---|---|
| 系统集成商和代理商 | 区域合作伙伴 | 部署、推广、本地化和一线执行 | 高但不透明 | 合作伙伴执行弱会拖慢试点、损伤可背书案例,并加重 Mech-Mind 售后负担 | 高 | 培训、价值提供商计划和公司支持体系 | 中高 |
| 大型工业终端市场 | 汽车、物流、电子、EV 电池、重工业 | 需求基础和标杆客户 | 行业集中度高;客户集中度未披露 | 资本开支收缩或审批延迟会压低订单转化和扩张 | 高 | 跨行业产品线和全球布局 | 高 |
| 先进算力和半导体投入 | GPU / 芯片 / 模组供应商 | 训练、推理和感知栈性能 | 不透明 | 许可延迟或投入不可得会拖慢路线图并推高成本 | 高 | 除一般产品标准化外,未见公开缓解措施 | 高 |
| 机器视觉现有厂商和相邻生态 | Cognex、Keyence、Basler、Omron、机器人 OEM / 集成商栈 | 具备既有信任和渠道深度的替代方案 | 高 | 现有厂商在核心工作流中捆绑销售、打折,或用更强支持压过 Mech-Mind | 高 | 一体化 Eye-Brain-Hand 栈和有竞争力定价主张 | 高 |
这张登记表把直接交易对手和生态依赖放在一起,因为主要下行传导来自商业层面,不只是合同披露。
[CR006, CR007, CR010, CR011, CR012, CR013]Mech-Mind 的运营模型依赖四个外部节点:系统集成商、大型工业买家、先进算力输入和既有视觉生态。任何一个都可能独立失效,叠加后会放大影响。
[CR007, CR008, CR010, CR013, CR016, CR038]7.5 团队、缓释成熟度与终止条件
执行风险高于产品层。公开材料仍高度围绕创始人兼 CEO Shao Tianlan 和少数可见管理者展开;董事会构成、审计准备度、内控成熟度和接班计划大多仍是私域信息。公司确有可见缓释项:600+ 名员工、自有相机工厂、全球子公司、伙伴培训,以及先在中国验证产品、再推向国际的策略。这些条件强于纯演示阶段故事,也解释了 Mech-Mind 为何能国际化扩张。但它们不能替代硬监控。关键的可投资问题包括:核心工作流中赢单率能否顶住 Cognex 和 Keyence;全球部署扩张时支持质量能否跟上;中国需求走弱是否拉长 PoC 到量产的周期;出口管制或贸易冲击是否拖慢路线图执行;传闻中的 IPO 能否转化为带可审计指标的公开文件。表 R005 把这些不确定性转成明确触发器。如果管理层在尽调中无法提供客户集中度、赢单 / 输单、BOM 和正常运行时间数据,正确姿态不是默认缓释有效,而是保留高剩余风险评级。[CR002, CR006, CR008, CR036, CR047, CR048]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓解措施 | 尽调路径 |
|---|---|---|---|---|---|
| 创始人 / CEO 和可见技术领导层 | 公开叙事仍高度集中在 Shao Tianlan,以及公开露出的有限梯队 | 中 | 高 | 全球化和产品化战略可见;公开资料中也能看到部分联合创始人和 VP 梯队 | 要求提供继任计划、授权运营角色和组织架构图 |
| 全球支持和现场执行 | 全球办公室和合作伙伴网络增加了跨时区、跨语言协调负担 | 中 | 高 | 600+ 名员工、本地子公司和培训体系是公开可见的缓解因素 | 要求提供区域服务人员人数、积压事项和客户升级处理指标 |
| IPO 前财务 / 治理准备度 | IPO 传闻材料没有配套公开的经审计收入、委员会结构或内控细节 | 高 | 高 | 强投资人阵容和特专科技上市路径已经存在 | 获取董事会材料、审计历史、财务控制团队架构和上市准备清单 |
| 机器人 + AI + 集成人才的招聘和留存 | 产品、支持和海外本地化同时扩张,需要稀缺工程人才 | 中 | 中高 | 公司称具备较深研发能力和本地化团队 | 要求提供离职率、薪酬区间和关键空缺岗位清单 |
执行风险取决于哪些能力必须同步扩张:研发宽度、支持质量和资本市场准备度。
[CR002, CR006, CR008, CR010, CR036, CR047]| 风险 | 可监控触发项 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 现有厂商渠道和定价压力 | 相对 Cognex / Keyence 的具名输赢单和折扣趋势 | 核心工作流连续三次战略性输单,或新交易毛利率压缩 >5pp | 重新评估护城河,下调增长假设,并要求提供竞标文件 |
| 出口管制 / 半导体获取 | 关键算力或相机组件需要许可或替代件 | 出货被拒、延迟 >90 天,或核心 BOM 被迫重设计 | 暂停路线图假设,重建产品成本 / 时间风险模型 |
| 中国宏观放缓和制造业周期性 | PMI 和订单转化趋势 | 官方或私营制造业 PMI 连续三个月低于 50,同时内部销售管线转化走弱 | 下调中国驱动的增长,并拉长销售周期假设 |
| IPO / 资本市场不确定性 | 从传闻走向可公开支撑上市流程的进展 | 没有公开申报、没有经审计收入桥,或较管理层预期延后 12+ 个月 | 移除 IPO 作为近期催化剂,并提高融资风险权重 |
| 客户集中度不透明 | 队列和集中度披露可得性 | 管理层在尽调中无法提供前 10 大客户、前 5 大垂直行业和续约数据 | 维持高风险评级,不把客户标识覆盖面当作集中度分散来支撑判断 |
| 现场可靠性和集成质量 | 可背书的正常运行时间和上线表现证据 | 没有可独立背书的正常运行时间数据,或旗舰应用中反复出现低于 95% 的表现 | 在确定敞口规模前,要求现场走访、客户访谈和实施款项保留 |
这些阈值是尽调工具,不是上市公司指引;每一项都把软叙事转成可观察的投资判断测试。
[CR028, CR034, CR035, CR038, CR039, CR049]7.6 图表
08估值
8.1 建议边界:公开数据究竟能支撑什么
Mech-Mind 显然已不再是投机性的种子阶段机器人公司。公开证据现在能支撑更扎实的底线:公司在 2025 年 8 月又完成一笔规模不小的融资,多项公开来源称其为独角兽,市场也把它作为秘密递表香港 IPO 的候选公司讨论。上述组合足以支撑后期阶段、持续融资能力,以及投资人对故事的认真兴趣;但还不足以支撑精确公允价值。公开来源仍未披露经审计收入、毛利率、烧钱速度、留存、客户集中度或股权结构条款。缺少这些项目,估值工作应定位为划边界,而不是给点估值。因此当前正确建议是继续研究。投资逻辑足够可投,值得继续推进,但文件证据还不足以证明只凭叙事就应付出溢价。公开证据可以支撑宽泛的独角兽级别区间和上市地可选性;还不能支撑高确信买入或清晰的 IPO 目标价。[CV034, CV035, CV036, CV039, CV040, CV041]
| 维度 | 评估 | 归因 | 置信度 | 决策含义 |
|---|---|---|---|---|
| 投资建议 | 继续研究 | 证据足以继续尽调,但不足以有信心支付溢价 | 中 | 入场前继续尽调 |
| 置信度 | 中 | 融资和独角兽身份已有交叉印证;财务披露没有 | 中 | 避免虚假精确 |
| 风险评级 | 高 | 私有市场不透明仍主导估值风险 | 中 | 要求硬披露 |
| 估值立场 | 偏高 | 独角兽叙事比经济性更清楚 | 中 | 不要假设 IPO 溢价合理 |
| 公开数据能够支撑什么 | 独角兽底线和融资可选项 | 2025 年融资、Hurun / Xinhua 认可和 IPO 报道彼此印证 | 高 | 故事是真的 |
| 公开数据不能支撑什么 | 精确公允价值 | 收入质量、利润率、烧钱速度和优先股堆叠仍被隐藏 | 高 | 还不能给出买入判断 |
这张表把本章转成投资人姿态,而不是假装公开数据足以支撑完整估值模型。
[CV034, CV035, CV036, CV039, CV040, CV041]| 立场 | 陈述 | 公开来源证据 | 重要性 | 置信度 |
|---|---|---|---|---|
| 投资逻辑 | 新资金叠加独角兽确认,说明投资人支持真实存在 | 2025 年 RMB500m 融资和 Hurun / Xinhua 独角兽框架 | 支撑继续尽调 | 高 |
| 投资逻辑 | 香港 IPO 可选项表明公司相信自己很快能达到公开市场标准 | 多家媒体报道其已秘密递交香港上市申请,目标募资约 $200m | 保持退出路径开放 | 中 |
| 投资逻辑 | 机器视觉和机器人市场足够大;若执行够强,仍有上行空间 | 2026 年市场规模测算和全球部署足迹 | 支撑上行可选项 | 中 |
| 反向逻辑 | 公开财务披露太薄,撑不起高信念尽调 | 未公开经审计收入、毛利率、烧钱速度或留存包 | 削弱公允价值置信度 | 高 |
| 反向逻辑 | 可比公司组噪音大,主要只能指方向 | 上市同行业务组合不同,非上市同行披露很少 | 拉大误差区间 | 中 |
| 反向逻辑 | 在软件式经济模型得到证明前,当前估值可能已经提前计入乐观预期 | 独角兽叙事很清楚,但商业化证据仍不完整 | 压缩安全边际 | 中 |
反向逻辑不是为了唱空而唱空;它抓住的是以独角兽级别价格入场时,恰恰缺失的证据。
[CV003, CV006, CV008, CV032, CV035, CV041]本章的建议逻辑很简单:新资金和独角兽身份先得到确认,但披露缺口仍在,所以结论是继续研究,而不是买入。
这条流程是本章的投资判断逻辑,不是公司发布的框架。
[CV001, CV006, CV035, CV039, CV040]公开 KPI 在融资和战略可选性上强,经营披露偏弱。
KPI 面板有意排除未披露的收入和利润率数字,而不是替它们做猜测。
[CV001, CV006, CV008, CV039, CV040, CV041]8.2 融资抬升、独角兽确认与 2022 基线的局限
最强的可获取融资事实是 2025 年 8 月轮次:多家英文和中文报道独立将其规模放在约 CNY500 million,并把资金用途指向继续投入 Mech-Mind 的具身智能眼脑手栈和全球商业化。可获取的公开报道也确认了 2022 年 CNY500 million 的 Series D 轮,为比较提供了可见的后期起点。更重要的估值变化发生在这两个点之间。到 2025-2026 年,China Daily 的 Hurun 报道明确把 Mech-Mind 放在 $1 billion 门槛之上;Xinhua 和 State Council 报道也分别称公司为中国独角兽。这些证据足以验证一个方向性抬升:公司从 2022 年融资时代进入了明确的独角兽区间。公开页面没有干净解决的,是订阅数据库常引用的 2022 年投后估值确切数字。做估值时,谨慎读法是公司显然已经进入独角兽框架,但抬升斜率的一部分仍被私募市场不透明遮住。[CV001, CV002, CV003, CV004, CV005, CV006]
8.3 可比框架与 IPO 路径选项
可比分析只有被当作框架才成立,不能把它当成一张能神奇解开 Mech-Mind 私募市场价格的表。Symbotic 是最接近的高增长仓储自动化参照,但它比 Mech-Mind 更偏安装和系统。Cognex 披露更清晰,也更接近机器视觉,但范围更窄、也更偏检测。Zebra 展示了多元化自动化硬件在投资人给出较低销售倍数时的交易状态;Teradyne 和 Universal Robots 则说明,从大型集团结构中抽出干净机器人倍数有多难。Mind Robotics 这类私募轮显示,投资人仍愿意为工业 AI 机器人支付溢价,但这些轮次同样披露稀疏,不应被当作持久公开市场支持的证明。退出路径上,HKEX Chapter 18C 是最可信的已披露通道,因为它明确容纳特专科技公司,并把机器人纳入先进硬件和软件。即便如此,公开来源仍未说明 Mech-Mind 会被筛为商业化还是未商业化,因此上市地可行性不等于承销确定性。[CV011, CV012, CV013, CV014, CV015, CV018]
| 参照对象 | 状态 | 公开指标 | 估值信号 | 相关性 | 关键局限 |
|---|---|---|---|---|---|
| Symbotic | 上市 | 2025 年收入 $2.247B | 2026 年 5 月市值 $32.61B(~14.5x) | 最接近的高增长仓储自动化参照 | 系统安装属性比 Mech-Mind 更重 |
| Cognex | 上市 | 2024 年收入 $914.5M | 2026 年 5 月市值 $10.99B(~12.0x) | 披露更清晰的机器视觉参照 | 工作流范围更窄,更偏检测 |
| Zebra | 上市 | 2024 年净销售额 $4.981B | 2026 年 5 月市值 $12.17B(~2.4x) | 低倍数自动化硬件锚点 | 业务过于多元,难以直接映射到具身 AI |
| Teradyne / Universal Robots | 上市集团 | 2024 年机器人业务收入 $364.8M | 集团市值 $56.11B;没有干净的分部 EV | 展示规模化机器人资产在申报文件里的样子 | 集团结构削弱直接倍数分析 |
| Mind Robotics | 非上市融资轮 | Series A 融资 $500M | 2026 年 3 月估值约 $2B | 可作为工业 AI 机器人私募轮胃口参照 | 披露稀疏,估值可能受热度影响 |
可比框架只具方向性。上市可比公司提供估值区间,非上市融资轮主要显示资金胃口,不等于完整承销过的经济模型。
[CV018, CV021, CV024, CV028, CV029, CV030]估值争议最受披露质量和业务结构证据影响,不是单靠又一轮高调融资就能定价。
敏感性条形只是定性权重辅助,不是统计因子模型。
[CV013, CV035, CV036, CV041, CV043, CV046]8.4 乐观 / 基准 / 悲观区间与估值立场
公司没有公开披露传统 IPO 模型所需的运营输入,用情景分析比给单一目标数更诚实。乐观情景下,如果披露的收入规模、持久海外部署和类软件利润率结构成立,投资人就可以穿透当前不透明,支持显著更高的公开市场或 pre-IPO 估值标记。基准情景下,Mech-Mind 仍是真实独角兽,技术和融资支持可信,但在披露包改善前,倍数扩张会受限。悲观情景下,IPO 或下一轮融资披露出的业务组合更像低倍数自动化硬件或服务,而不是软件。即使公司仍有战略价值,承销区间也会被大幅压缩。关键在于,当前公开证据支撑的是围绕独角兽底线和 18C 规模门槛的宽区间,而不是精确上限。因此本报告给出“偏高而非断裂”的判断:故事仍可能成立,但公开证明包还不够厚,不能为上行叙事激进付费。[CV032, CV033, CV041, CV042, CV044, CV045]
| 情景 | 成立前提 | 估值框架 | 概率信号 | 改变判断的证据 |
|---|---|---|---|---|
| 乐观 | 经审计收入规模浮现,全球部署持续叠加,并且看得见软件式利润率 | 示例可支撑区间:$1.8B-$2.5B | 需要多项新增披露,而不是再融一轮 | 收入、毛利率和客户队列证据 |
| 基准 | 公司仍是真独角兽,但披露只得到部分改善 | 示例可支撑区间:$1.1B-$1.6B | 与当前公开证据最一致 | 招股书级别披露,但经济模型仍不完整 |
| 悲观 | IPO 或下一轮融资暴露更重的硬件服务组合,或增长更弱 | 示例可支撑区间:$0.7B-$1.0B | 如果披露不及预期,估值压缩风险上升 | 更低利润率、更慢增长,或 IPO 撤回 |
| 乐观 / 悲观摆动因素 | 香港上市推进,并给出可信披露包 | 缩小当前不确定性折价 | 会显著收窄估值区间 | 申报稿和经审计报表 |
| 乐观 / 悲观摆动因素 | 客户质量和可复制性披露清楚 | 决定公开市场投资者看到的是平台,还是项目制生意 | 最重要的缺失运营变量 | 订单簿和队列数据 |
情景区间只是示例性承销区间,不是管理层指引,也不是二级市场目标价。
[CV041, CV042, CV044, CV045, CV047]现有公开证据只能支撑围绕独角兽估值底线的宽区间,而不是一个窄目标价。
这些区间是投资判断辅助,依据独角兽身份确认、18C 门槛和公开可比公司框架拼出;不是市场出清报价。
[CV041, CV042, CV044, CV045]8.5 最终尽调要求与投资逻辑终止触发器
剩余工作异常具体。投资人不需要另一场大而全的产品演示来提升估值信心;他们需要一套运营和治理材料,把独角兽叙事转化为可承销价格。四项要求最关键:经审计的 2024-2025 年财务、订单簿和客户队列数据、包含优先权条款的完整股权结构表,以及说明上市地、分类和募资用途的上市披露草案。没有这些材料,建议置信度应停留在中等。终止触发器也遵循同一纪律。如果香港 IPO 被撤回,如果下一轮价格低于当前独角兽框架,或者经审计披露显示其业务更接近低倍数自动化同行的低利润硬件服务组合,投资逻辑就需要迅速重置。因此估值问题不是“Mech-Mind 有意思吗?”,而是“哪些新披露能让当前标记具备可投资性?”在答案出现之前,正确姿态是主动尽调,而不是被动相信。[CV035, CV036, CV043, CV046]
| 触发项 | 阈值或事件 | 对投资逻辑的传导 | 行动含义 | 需要的核心证据 |
|---|---|---|---|---|
| IPO 撤回或无限期推迟 | 保密筹备后香港申报仍未推进 | 表明市场准备度或披露舒适度更弱 | 重设估值底线并等待 | 正式申报状态和顾问评论 |
| 下调估值融资或结构化救助融资 | 下一轮定价低于当前独角兽框架,或加入惩罚性条款 | 摧毁当前后期定价信号 | 不追价;按新股权结构表重新承销 | 条款清单和优先股堆叠 |
| 披露低毛利经济模型 | 经审计数字显示硬件服务组合占比高,毛利率弱 | 使 Mech-Mind 更接近低倍数自动化同行 | 大幅压缩估值区间 | 经审计利润表和分部组合 |
| 客户集中度极高 | 一两个客户主导商业化证明 | 削弱收入和平台叙事的持久性 | 即使收入增长,也降低信心 | 队列和集中度披露 |
| 国际部署质量不及预期 | 全球足迹存在,但重复付费部署很薄 | 平台故事变成试点故事 | 转向悲观情景 | 装机基础和续约数据 |
这些触发项的设计目的,是在新披露削弱经济模型时,阻止投资者继续为叙事买单。
[CV035, CV036, CV045, CV046]| 主题 | 缺失证据 | 重要性 | 责任方或尽调路径 | 优先级 |
|---|---|---|---|---|
| 经审计的 2024-2025 年财务数据 | 收入、毛利率、运营费用、烧钱速度、现金 | 任何公允价值测算都缺这个核心输入 | CFO、审计师或上市草案 | 关键 |
| 客户和订单簿队列 | 客户数量、集中度、续约 / 扩张、积压订单质量 | 告诉投资者部署能否复制 | 收入运营和财务 | 关键 |
| 股权结构表和优先权条款 | 优先股堆叠、反稀释、清算权、董事会控制权 | 为稀释和下行保护定价 | 法律顾问和财务团队 | 关键 |
| IPO 分类和路径 | 商业化 vs 未商业化 18C 处理、上市地逻辑、募资用途计划 | 决定上市门槛和风险框架 | 上市顾问和招股书草案 | 高 |
| 硬件与软件组合 | 装机基础、软件附着率、服务占比、按产品族拆分的利润率 | 区分平台经济模型和集成商经济模型 | 管理账和定价包 | 高 |
| 全球部署质量 | 按地区和垂直行业拆分的付费部署,而不只是展示活动 | 检验国际足迹能否转化为耐久收入 | 销售负责人和渠道报告 | 高 |
这些问题是最低配置;只有拿到它们,才能从叙事丰富的独角兽框架,推进到价格纪律约束下的投资决策。
[CV035, CV036, CV043]免责声明
本报告是基于公开证据的尽调快照,不构成投资建议。关键财务、法律、技术和合同事实仍未公开;投资前,应直接通过管理层和原始文件核验。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Mech-Mind was founded in 2016, with public records preserving a specific formation date of 2016-09-12. | 高 | SO009, SO013, SO025 |
| CO002 | Official Chinese company materials describe Mech-Mind as a company founded by a Tsinghua-returnee team. | 高 | SO009, SO025 |
| CO003 | Shao Tianlan is the founder and current CEO of Mech-Mind. | 高 | SO002, SO013 |
| CO004 | Public profiles say Shao Tianlan earned a bachelor’s degree from Tsinghua University’s School of Software and a master’s degree in robotics from the Technical University of Munich. | 高 | SO002, SO013, SO015 |
| CO005 | 36Kr’s project profile identifies Fu Ao as co-founder and business VP. | 中 | SO013 |
| CO006 | 36Kr’s project profile identifies Ding Youshuang as co-founder and R&D management VP. | 中 | SO013 |
| CO007 | Mech-Mind’s current official materials define the “Eye-Brain-Hand” stack as industrial 3D cameras, AI software or Mech-GPT, and dexterous hands that jointly deliver perception, reasoning, and manipulation. | 高 | SO001, SO006 |
| CO008 | Independent profiles frame Mech-Mind primarily as a supplier of standardized robot vision, intelligence, and manipulation components rather than a maker of full robot bodies. | 中 | SO006, SO015, SO017 |
| CO009 | Official Chinese company materials anchor Beijing as the R&D center and Shanghai as the sales and delivery base. | 中 | SO009 |
| CO010 | English and Chinese contact pages show a wider office network spanning Beijing, Shanghai, Japan, South Korea, Germany, the United States, and multiple other Chinese cities. | 高 | SO003, SO010 |
| CO011 | Current company pages say Mech-Mind’s business covers nearly 50 countries and regions. | 高 | SO001, SO009, SO010 |
| CO012 | Current official materials say Mech-Mind has deployed more than 24,000 units globally and serves more than 100 Fortune Global 500 clients. | 高 | SO001, SO009, SO023 |
| CO013 | The current Chinese company profile says Mech-Mind has more than 600 employees globally. | 中 | SO009 |
| CO014 | The official English team page names Professor Jianwei Zhang as founding technical advisor and chief scientist. | 中 | SO002 |
| CO015 | KrASIA identifies Xu Tingting as vice president of business and marketing and quotes her on Mech-Mind’s controlled overseas expansion strategy. | 中 | SO015 |
| CO016 | The fetched official pages do not disclose a full board roster or a complete executive bench beyond a small number of named leaders. | 中 | SO002, SO005 |
| CO017 | Mech-Mind says it has established an Integrity Compliance Committee and an Integrity Compliance Inspectorate Team for anti-corruption matters. | 中 | SO005 |
| CO018 | Mech-Mind publishes compliance@mech-mind.net as a reporting channel for suspected anti-corruption violations. | 中 | SO005 |
| CO019 | Official Chinese company materials say Mech-Mind has raised more than RMB 2 billion cumulatively and name IDG Capital, Meituan, Sequoia China, Source Code Capital, Intel Capital, and Qiming Venture Partners as backers. | 中 | SO009 |
| CO020 | The current official English about page still says Mech-Mind had closed its Series C+ round with total funding of USD 300 million. | 中 | SO001 |
| CO021 | 36Kr reported in September 2021 that Mech-Mind completed a near-RMB 1 billion C-series financing led by Meituan and IDG Capital, with Sequoia China and Source Code Capital following. | 高 | SO011, SO013 |
| CO022 | 36Kr’s project page labels Mech-Mind as an E-round company and records a 2025-03 E round with Nanxiang Venture Capital and the Hebei SOE Reform Fund. | 中 | SO013 |
| CO023 | The same 36Kr project page records a D round in 2023-08 and a D++ round in 2024-12 tied to Galileo Capital and China Xiong’an Group respectively. | 中 | SO013 |
| CO024 | MarketScreener reports that Mech-Mind raised CNY 500 million on 2025-08-25 from a syndicate including Broad-Ocean Motor, CICC-affiliated funds, Shanghai Nanxiang Venture Capital, Haihe Industry Fund, Hebei Structural Reform Fund, China Growth Capital, and China Xiong’an Group Fund Management. | 中 | SO018 |
| CO025 | Hong Kong financial-media coverage says China Xiong’an Group and the Hebei SOE Reform Fund were also shareholders by late 2025. | 高 | SO020, SO021 |
| CO026 | AASTOCKS, HKCD, and HKET reported in September 2025 that Mech-Mind had filed confidentially for a Hong Kong IPO targeting roughly USD 200 million, but the retained pack does not include an official filing text or prospectus. | 高 | SO019, SO020, SO021 |
| CO027 | Because current sources disagree between Series C, Series C+, and E-round labels, the safest synthesis is that Mech-Mind is a late-stage private company whose public stage metadata is stale across directories and older marketing pages. | 中 | SO001, SO013, SO024 |
| CO028 | Automate’s association profile says Mech-Mind has delivered more than 10,000 industrial 3D cameras in more than 50 countries and served more than 1,500 clients worldwide. | 中 | SO022 |
| CO029 | KrASIA reported that Mech-Mind’s technology was operating in over 50 countries with more than 15,000 installations worldwide by 2024. | 中 | SO015 |
| CO030 | The latest current official company pages update the scale marker to 24,000+ units across nearly 50 countries and regions, suggesting growth and a broader denominator than older camera-only or installation counts. | 高 | SO001, SO009, SO023 |
| CO031 | 36Kr’s founder interview says Mech-Mind had more than 15,000 global deployments and five consecutive years of market-share leadership around 2025. | 中 | SO012 |
| CO032 | The current Chinese company profile says Mech-Mind has ranked number one in market share for five consecutive years. | 中 | SO009 |
| CO033 | KrASIA reported that Mech-Mind opened a Tokyo robotics lab in March 2025 with a 1,000-square-meter facility including a 400-square-meter exhibition and training area. | 中 | SO015 |
| CO034 | QbitAI’s Davos coverage says Mech-Mind has subsidiaries in Germany, Japan, the United States, and South Korea and is pursuing productization, ecosystem building, and globalization. | 中 | SO014 |
| CO035 | Official AW 2026 coverage says Mech-Mind showed more than 10 demonstration units and globally debuted products including the Mech-Eye ULTRA M-GL and Mech-Station InstaDepal. | 高 | SO004, SO007 |
| CO036 | Official iREX 2025 coverage says Mech-Mind used a 360-square-meter booth and nearly 20 application units to showcase humanoid-retail, clothes-folding, transparent-object, and industrial scenarios. | 中 | SO008 |
| CO037 | Aiqicha’s company-risk snapshot lists 18 business disputes, 3 filing records, 6 hearing announcements, and 4 litigation relationships for Mech-Mind’s Xiong’an entity. | 中 | SO025 |
| CO038 | The same Aiqicha profile says the Xiong’an entity remains active and records 954 trademarks, 384 patents, and 47 software copyrights. | 中 | SO025 |
| CO039 | Public sources disagree on headquarters labelling: official company pages emphasize Beijing and Shanghai operating roles, while Automate and Hong Kong IPO coverage foreground Xiong’an or Hebei addresses tied to legal or association entities. | 中 | SO009, SO020, SO022 |
| CO040 | No retained public source in this pack confirms a current valuation, revenue figure, or ARR number for Mech-Mind. | 低 | SO001, SO009, SO024, SO017 |
| CO041 | AsiaICT describes Mech-Mind as an embodied-AI unicorn and says cumulative funding is nearly RMB 2 billion, but that unicorn framing is secondary media language rather than a disclosed company valuation. | 中 | SO017 |
| CO042 | Tracxn’s public summary still shows Stage: Series C and total funding of $200 million over 6 rounds, which lags later financing signals in 36Kr and 2025 market coverage. | 中 | SO024, SO013 |
| CO043 | Aiqicha preserves a 2017-05 Pre-A round led by Huachuang Capital / China Creation Ventures. | 中 | SO025 |
| CO044 | Aiqicha preserves an A and A+ financing step in 2019-04 at roughly hundred-million-RMB scale. | 中 | SO025 |
| CO045 | Aiqicha preserves an August 2019 Intel investment in Mech-Mind. | 中 | SO025 |
| CO046 | Aiqicha preserves a February 2020 B-round led by Sequoia China and a November 2020 B+ round backed by Source Code Capital and Sequoia China. | 中 | SO025 |
| CO047 | Aiqicha and 36Kr project materials indicate an earlier 2021-04 C round led by Meituan with Sequoia China and Source Code following before the larger September 2021 C-series financing. | 高 | SO013, SO025 |
| CM001 | Mech-Mind publicly positions itself as an industrial 3D-vision and AI stack for robotic automation rather than as a robot-arm OEM. | 中 | SM002, SM004 |
| CM002 | Typical Mech-Mind applications include bin picking, depalletizing and palletizing, machine tending, pick and place, and assembly. | 中 | SM001, SM004 |
| CM003 | Public 2026 show coverage presents Mech-Mind as solving hard-perception tasks involving transparent, reflective, randomly stacked, and mixed-SKU objects. | 中 | SM002, SM003 |
| CM004 | The relevant market boundary is the perception, planning, and deployment layer on top of robot cells, not all industrial-robot or factory-automation spend. | 中 | SM003, SM004 |
| CM005 | Included spend for Mech-Mind-like demand covers 3D sensors, vision and AI software, workflow configuration, integration, and support; most robot-arm hardware and broad facility automation should be excluded. | 中 | SM003, SM004 |
| CM006 | Status-quo substitutes include manual handling, simpler 2D vision, fixed automation, and custom-coded robot cells delivered by integrators. | 中 | SM003, SM008, SM018 |
| CM007 | IFR says 542,000 industrial robots were installed worldwide in 2024 and total operational stock reached 4.664 million units. | 高 | SM006, SM024 |
| CM008 | Asia accounted for 74% of global industrial-robot deployments in 2024. | 高 | SM006, SM024 |
| CM009 | China installed 295,000 industrial robots in 2024, representing about 54% of global deployments. | 高 | SM006, SM019, SM020, SM024 |
| CM010 | IFR forecasts global robot installations to rise to 575,000 in 2025 and to surpass 700,000 units by 2028. | 高 | SM006, SM024 |
| CM011 | China’s operational robot stock exceeded 2 million units in 2024 and domestic suppliers reached roughly 57% share of the home market. | 高 | SM006, SM020, SM024, SM027 |
| CM012 | People's Daily and Yicai both report IFR data showing China's robot density at about 470 units per 10,000 employees. | 高 | SM019, SM021 |
| CM013 | ChinaPower reports a lower China automation-intensity figure of 166 robots per 10,000 workers for 2024. | 中 | SM022 |
| CM014 | Public China robot-density figures are not fully apples-to-apples, because sources use different workforce denominators and measurement scopes. | 中 | SM019, SM021, SM022 |
| CM015 | Precedence Research places the global warehouse-automation market at $29.30 billion in 2026 and $107.36 billion by 2035. | 中 | SM015 |
| CM016 | Mordor Intelligence places the warehouse-automation market at $34.17 billion in 2026 and $65.74 billion by 2031. | 中 | SM014 |
| CM017 | Grand View Research sizes warehouse robotics at $4.31 billion in 2022 and $17.29 billion by 2030, with Asia Pacific the largest market in the base year. | 中 | SM013 |
| CM018 | Precedence Research sizes the machine-vision market at $26.07 billion in 2026 and says Asia Pacific held the largest regional share while automotive was the largest end-use vertical. | 中 | SM016 |
| CM019 | A3’s 2026 forum summary says machine-vision markets are expected to grow about 7.7% CAGR through 2029 and that 3D vision software, bin picking, and AI applications are the biggest growth areas. | 中 | SM017 |
| CM020 | A3’s 2026 forum summary says automotive remains the largest sector for vision integration while logistics and warehousing is the fastest-growing machine-vision segment at about 14.2% annual growth. | 中 | SM017 |
| CM021 | The local public source pack does not isolate a clean standalone 2026 TAM for industrial 3D robot guidance or AI vision, so a layered adjacent-market approach is required. | 中 | SM012, SM013, SM014, SM015, SM016, SM017 |
| CM022 | ChinaPower and 36Kr both show that electronics and automotive are the two largest robot-demand verticals in China. | 高 | SM020, SM022, SM026 |
| CM023 | Automotive demand for Mech-Mind-like systems centers on reflective metal handling, precision assembly, in-line measurement, and defect inspection. | 中 | SM002, SM003 |
| CM024 | Electronics demand for Mech-Mind-like systems centers on OCR, glue-bead inspection, transparent or reflective handling, and high-mix precision cells. | 中 | SM002, SM005 |
| CM025 | Logistics demand for Mech-Mind-like systems centers on depalletizing, palletizing, parcel induction, bin picking, and mixed-SKU piece picking. | 中 | SM002, SM003, SM004 |
| CM026 | Budget ownership for these deployments usually sits with plant operations, quality, manufacturing engineering, or DC operations rather than a standalone enterprise-AI budget. | 中 | SM005, SM018, SM019 |
| CM027 | System integrators are a critical channel and deployment layer for this market because Mech-Mind sells cameras and software suites designed to be embedded into customer workcells. | 中 | SM004, SM005 |
| CM028 | Mech-Mind’s stated China-first validation model makes domestic Chinese demand strategically important because products are proven at home before overseas rollout. | 中 | SM005 |
| CM029 | Mech-Mind’s offices and teams in Germany, the United States, Japan, and South Korea indicate a localization strategy rather than pure cross-border export selling. | 中 | SM004, SM005 |
| CM030 | Warehouse robotics adoption is still led by labor availability and labor cost, cited by 55% and 42% of survey respondents respectively. | 中 | SM009 |
| CM031 | Warehouse-robotics adoption momentum is real but budget-ready demand is narrower: 48% already use robots, 32% plan adoption within three years, and only 32% had approved funding. | 中 | SM009 |
| CM032 | ProMat 2025 coverage shows labor pressure is pulling AMRs, AS/RS, robotic picking, depalletizing, and inventory drones toward mainstream deployment conversations. | 中 | SM010 |
| CM033 | IFR’s 2026 trends frame AI autonomy, IT/OT convergence, safety, and labor shortages as the main forces shaping robotics demand. | 中 | SM007 |
| CM034 | IFR says cobots are well suited to low-volume, high-mix industries including automotive, electronics, logistics, bin picking, and end-of-line palletizing. | 中 | SM008 |
| CM035 | OSHA and IFR both imply that safety validation, oversight, and certification remain real deployment gates as robots operate closer to people. | 高 | SM007, SM011 |
| CM036 | IoT Analytics identifies high AI cost, insufficient data infrastructure, and workforce skill gaps as the leading AI-scaling barriers in machine building. | 中 | SM018 |
| CM037 | The Vention / Industry Week survey says 92% of manufacturers view automation as essential, but only 37% report significant or full automation and 50% struggle to identify the right technology. | 中 | SM023 |
| CM038 | Mordor says hardware still led 55.12% of 2025 warehouse-automation spending even as software is forecast to grow faster at 14.87% CAGR through 2031. | 中 | SM014 |
| CM039 | Mordor also flags fixed-system capex and legacy WMS integration complexity as material restraints that can delay warehouse-automation projects. | 中 | SM014 |
| CM040 | Mech-Mind fits globally as a picks-and-shovels industrial AI vendor selling the perception and no-code deployment layer across automotive, electronics, and logistics rather than a single-task robot OEM. | 中 | SM003, SM004, SM005 |
| CM041 | Mech-Mind’s strongest near-term SAM is likely high-throughput hard-perception workcells that already use industrial robots in China and in export-heavy manufacturing and logistics hubs. | 低 | SM002, SM006, SM020, SM022 |
| CM042 | The upside case is that faster-growing machine-vision, 3D-vision-software, and warehouse-automation layers let perception vendors outgrow the underlying industrial-robot hardware base. | 中 | SM015, SM016, SM017 |
| CM043 | The main underwriting risk is that public market figures are adjacent TAMs while actual capture depends on workflow ROI, integrator economics, and deployment speed that Mech-Mind does not disclose publicly. | 中 | SM014, SM018, SM023 |
| CP001 | Mech-Mind publicly frames its offer as an Eye-Brain-Hand stack spanning retail automation, depalletizing/palletizing, machine tending, piece picking, and bin picking. | 高 | SP001, SP002 |
| CP002 | Mech-Mind’s Chinese and documentation surfaces show a broader stack than most point-solution peers, including Mech-Eye cameras, laser profiling, Mech-Vision, Mech-Viz, Mech-DLK, Mech-MSR, and embodied robot stations. | 高 | SP003, SP007 |
| CP003 | Mech-Mind documentation spans 3D robot guidance, measurement and inspection, AI-based quality inspection, and robot communication and integration, indicating deployment tooling beyond sensing hardware alone. | 中 | SP003 |
| CP004 | A3 describes Mech-Mind as an industry-leading industrial 3D camera and software-suite company for intelligent robotics. | 中 | SP004 |
| CP005 | A3 says Mech-Mind sells easy-to-use products at a competitive price and works closely with integrators across applications such as bin picking, depalletizing, and pick-and-place. | 中 | SP004 |
| CP006 | Official and independent sources both place Mech-Mind’s deployed footprint at global scale across more than 50 countries or regions. | 高 | SP001, SP004, SP005 |
| CP007 | Independent sources say Mech-Mind was founded in 2016 by Shao Tianlan, whose background includes Tsinghua University and the Technical University of Munich. | 中 | SP005, SP006 |
| CP008 | AsiaICT reports nearly RMB 2 billion of cumulative funding for Mech-Mind, while KrASIA reports more than USD 200 million, indicating unusually deep backing for an industrial-vision startup even if the currency framing differs. | 中 | SP005, SP006 |
| CP009 | Mech-Mind’s 2026 Automation World release extends the stack into standardized depalletizing cells, transparent-object handling, reflective-part bin picking, and AI-enabled 2D/3D inspection. | 中 | SP027 |
| CP010 | Official and independent Mech-Mind materials emphasize plug-and-play or fast-deployment positioning rather than long custom coding cycles. | 中 | SP005, SP007 |
| CP011 | Qviro describes Cognex as a leading global machine-vision and barcode-reading company known for accurate and reliable smart-camera and vision-sensor products. | 中 | SP022 |
| CP012 | Keyence’s 3D VGR package combines a four-camera and one-projector imaging unit with automatic robot-camera calibration, CAD upload, and path planning for assembly, depalletizing, and machine tending. | 中 | SP008 |
| CP013 | Keyence’s broader vision portfolio includes all-in-one smart cameras and modular high-speed controllers for 2D, line-scan, and 3D cameras using both AI and rule-based tools. | 中 | SP009 |
| CP014 | Photoneo’s public portfolio spans Locator Studio, Bin Picking Studio, depalletization, delayering, MotionCam-3D, and multiple 3D scanners, showing broad 3D-perception coverage for handling workflows. | 中 | SP010, SP011, SP012 |
| CP015 | Photoneo specifically markets MotionCam-3D for moving-object applications such as conveyors, assembly in motion, logistics, and inspection. | 中 | SP012 |
| CP016 | Intrinsic competes less as a camera vendor and more as a software-control layer with Flowstate, a skills architecture, an Intrinsic Vision Model, and a broader platform ecosystem. | 中 | SP013, SP014, SP015 |
| CP017 | Flowstate supports digital twins, simulation-to-real transfer, Python and C++ development, graphical UI, custom skills, pose estimation, and motion planning. | 中 | SP014 |
| CP018 | Intrinsic’s 2026 FANUC partnership extends that platform onto FANUC industrial and collaborative robots and into ROS, Gazebo, and Open-RMF workflows. | 中 | SP016 |
| CP019 | Roboception’s rc_visard family is surrounded by modular software blocks such as CADMatch, ItemPickAI, BoxPick, SLAM, and URCap integrations, making it a focused but modular 3D-perception rival. | 中 | SP017 |
| CP020 | Omron Robotics positions itself across automotive, digital, food, medical, and logistics workflows spanning inbound handling, assembly, inspection, packaging, and transport. | 中 | SP018 |
| CP021 | Qviro identifies Omron as a broad factory-automation vendor whose robot-vision systems help robots inspect, assemble, handle materials, and dispense liquids. | 中 | SP022 |
| CP022 | Universal Robots markets collaborative arms as safe, flexible systems that are easy to deploy, program, and scale, and it layers marketplace and partner channels around the hardware. | 中 | SP019 |
| CP023 | Universal Robots’ current lineup spans lighter e-Series-style collaborative arms and higher-payload next-generation arms such as UR20 and UR30. | 中 | SP019 |
| CP024 | Pickit positions itself around focused 3D-vision applications such as bin picking, depalletizing, assembly, and in-line measurement rather than a wider eye-brain-hand stack. | 中 | SP025 |
| CP025 | ISRA Vision is a machine-vision incumbent with public industry coverage across automotive, battery, glass, metals, paper, plastics, and other inspection-heavy sectors. | 中 | SP026 |
| CP026 | QYResearch lists Mech-Mind, Keyence, Pickit3D, Roboception, and OMRON among a broader global field of 3D-vision-for-robot suppliers. | 中 | SP024 |
| CP027 | Qviro’s 2026 ranking places Cognex, Keyence, Omron, and Pickit among leading robot-vision brands, supporting the case that buyers have several recognized alternatives to evaluate. | 中 | SP022 |
| CP028 | Public price transparency is low across the reviewed Mech-Mind, Keyence, Photoneo, Intrinsic, Roboception, Pickit, and ISRA surfaces, which emphasize demos, use cases, and contact motions instead of list pricing. | 中 | SP004, SP008, SP010, SP013, SP017, SP025, SP026 |
| CP029 | Universal Robots is the clearest public cost anchor in this source pack because its own budgeting guide discusses tooling, software, training, service, and integrator costs even though quotes remain tailored. | 中 | SP020 |
| CP030 | A third-party 2025 price guide places UR arms roughly from USD 23,000 for a UR3e to more than USD 85,000 for a UR20 and says integration and accessories can double those base prices. | 中 | SP021 |
| CP031 | Mech-Mind’s A3 profile claims competitive pricing, but no public Mech-Mind list price appears in the reviewed official or industry sources. | 中 | SP001, SP004 |
| CP032 | Mech-Mind’s partner and integrator orientation differs from Intrinsic’s developer-platform motion and from Universal Robots’ robot-plus-marketplace commercial model. | 中 | SP004, SP014, SP019 |
| CP033 | Mech-Mind’s public vertical mix centers on automotive, logistics, electronics, food, fulfillment, and retail-style piece picking. | 中 | SP001, SP004, SP005, SP027 |
| CP034 | Photoneo, Pickit, and Roboception look strongest in classic 3D robot-guidance workflows such as bin picking, depalletizing, and hard-part handling rather than in broad inspection stacks. | 中 | SP011, SP017, SP025 |
| CP035 | Keyence, Cognex, Omron, and ISRA bring broader incumbent inspection or automation reach than Mech-Mind in channels and installed trust. | 中 | SP009, SP022, SP023, SP026 |
| CP036 | Intrinsic and Universal Robots are more ecosystem-centric substitutes than like-for-like Mech-Mind peers because they extend skills, partners, and software abstraction across third-party robots. | 中 | SP014, SP016, SP019, SP020 |
| CP037 | Intel Market Research says incumbents such as Cognex and Keyence keep strengthening positions through expanded product portfolios and partnerships. | 中 | SP023 |
| CP038 | QYResearch says the global top five 3D-vision-for-robot players held an aggregated revenue share in 2024, implying a market where concentration and feature convergence can pressure smaller vendors. | 中 | SP024 |
| CP039 | Mech-Mind’s strongest public differentiation is a one-vendor Eye-Brain-Hand workflow stack that combines 3D sensing, robot programming, deep learning, inspection, and standardized workcells. | 高 | SP002, SP003, SP007, SP027 |
| CP040 | That differentiation is strongest when customers want one robot-agnostic layer for transparent objects, reflective metals, depalletizing, and bin-picking workflows rather than a standalone camera or arm. | 中 | SP001, SP014, SP027 |
| CP041 | Mech-Mind is more exposed where buyers already standardize on incumbent automation stacks or prefer platform ecosystems that let them mix robot brands, skills, and partner add-ons. | 中 | SP016, SP019, SP020, SP023 |
| CP042 | Public Mech-Mind deployment metrics vary by source between 10,000-plus cameras, 15,000-plus installations, and 24,000-plus cameras, so the overall footprint is supported but the precise denominator still needs diligence. | 中 | SP001, SP004, SP005 |
| CI001 | Official company materials describe Mech-Mind's monetized portfolio as an embodied-intelligence "Eye-Brain-Hand" stack spanning 3D cameras, AI software suites, and dexterous-hand / robot-station products. | 高 | SI001, SI002, SI003, SI004, SI005 |
| CI002 | The Mech-Vision product page says the software is used to build automation applications including bin picking, machine tending, piece picking, depalletizing, palletizing, and assembly. | 中 | SI004 |
| CI003 | The Mech-Viz page says its code-free interface supports one-click simulation and motion planning for demanding automation tasks. | 中 | SI005 |
| CI004 | Official about and cooperation pages say Mech-Mind supports partners through consulting, solution design, training, deployment, and maintenance. | 高 | SI001, SI008 |
| CI005 | Official materials mention competitive pricing but do not publish a price list or pricing formula. | 中 | SI001 |
| CI006 | Jiemian reported that Mech-Eye, Mech-Vision, and Mech-Viz were typically sold together as an eye / visual-nerve / brain stack. | 中 | SI016 |
| CI007 | Jiemian reported in 2021 that Mech-Mind's product pricing could be roughly half that of mainstream foreign competitors. | 中 | SI016 |
| CI008 | Jiemian reported that mature support cycles could be measured in weeks while new applications could take months. | 中 | SI016 |
| CI009 | Across official product, solution, and case-study pages, the commercial offer looks like a deployment-specific bundle of hardware, software, and implementation support rather than a pure stand-alone SaaS product. | 中 | SI003, SI004, SI005, SI006, SI007 |
| CI010 | Because official materials are aimed at integrators and value providers, channel economics and partner-led delivery likely matter materially to realized revenue. | 中 | SI001, SI008 |
| CI011 | Official company materials say Mech-Mind has deployed more than 24,000 units across nearly 50 countries and regions. | 高 | SI001, SI002 |
| CI012 | Official company materials say Mech-Mind serves more than 100 Fortune Global 500 clients. | 高 | SI001, SI002 |
| CI013 | Official materials say Mech-Mind deployments span automotive, food and beverage, logistics, home appliances, EV batteries, metal and machining, and electronics. | 中 | SI001 |
| CI014 | Jiemian reported that Mech-Mind's team had already exceeded 300 people by 2021. | 中 | SI016 |
| CI015 | Jiemian reported that orders were growing more than threefold per year and that one quarter's orders matched a prior full year. | 中 | SI016 |
| CI016 | Pedaily and the 科创板日报 article reposted on Xueqiu both said Mech-Mind ranked first in China's 3D-vision-guided industrial robot market for five consecutive years through 2024. | 中 | SI009, SI019 |
| CI017 | The 科创板日报 article reposted on Xueqiu said Mech-Mind held 38% share of China's 3D-vision-guided industrial robot market in 2024. | 中 | SI019 |
| CI018 | The 科创板日报 article reposted on Xueqiu said overseas revenue share had risen to about 50% after the company's overseas expansion push. | 低 | SI019 |
| CI019 | Craft's profile still showed only 200 enterprise customers across 10 customer countries as of December 2020, underscoring how third-party datasets can lag the company's current official deployment claims. | 中 | SI022 |
| CI020 | Official company messaging emphasizes measurable ROI, but no reviewed source disclosed the ROI formula, realized payback period, or contract value behind that claim. | 中 | SI001 |
| CI021 | The official about page says Mech-Mind has closed a Series C+ round and raised USD300 million in total. | 中 | SI001 |
| CI022 | CB Insights lists Mech-Mind as having raised $222.36 million over 14 rounds. | 中 | SI021 |
| CI023 | CB Insights identifies Mech-Mind's latest round as Series E-II dated August 26, 2025. | 中 | SI021 |
| CI024 | Marketscreener, Pedaily, and 163 all support that the August 2025 round was about CNY500 million and included state-linked and industrial investors such as Broad-Ocean, CICC Porsche, Haihe, Hebei Structural Reform, Nanxiang, Tianjin Venture Capital, China Growth Capital, and Xiong'an-linked capital. | 高 | SI009, SI011, SI020 |
| CI025 | Multiple August 2025 sources say proceeds from the latest round were earmarked for Eye-Brain-Hand R&D, product-line expansion, broader scenario coverage, and stronger global commercialization and customer service. | 中 | SI009, SI018, SI020 |
| CI026 | CB Insights says Mech-Mind's August 1, 2022 Series D raised $38 million at a reported $858 million valuation. | 中 | SI021 |
| CI027 | Yahoo Finance's Bloomberg syndication and Ifeng both say Mech-Mind's 2021 C round was led by Meituan, with Sequoia China among participants. | 中 | SI012, SI014 |
| CI028 | The Standard, HKCD, Sohu, and 163 each describe Mech-Mind's cumulative funding by late 2025 as roughly or above RMB2 billion. | 中 | SI013, SI015, SI017, SI020 |
| CI029 | Low-reputation pre-IPO commentary on Eastmoney placed Mech-Mind's September 2025 valuation around RMB8 billion, which would imply unicorn status. | 低 | SI025 |
| CI030 | Because the only post-2022 valuation step-up figure we found was low-reputation market commentary, the company's claimed move from an $858 million 2022 valuation to confirmed unicorn status by 2025 remains unverified. | 中 | SI021, SI025, SI026 |
| CI031 | Yahoo Finance, The Standard, Ifeng, and HKCD all reported that Mech-Mind was considering or had confidentially submitted a Hong Kong IPO targeting roughly $200 million, while also saying details were not final. | 高 | SI012, SI013, SI014, SI015 |
| CI032 | Yahoo Finance and Ifeng both said Mech-Mind had not publicly confirmed the reported Hong Kong IPO plan when contacted or before publication. | 中 | SI012, SI014 |
| CI033 | The combination of a reported RMB500 million 2025 round, cumulative funding around or above RMB2 billion, and the company's own USD300 million total-funding claim suggests financing dependency has been reduced but not eliminated. | 中 | SI001, SI011, SI013 |
| CI034 | No reviewed source disclosed cash on hand, monthly burn, runway months, debt balances, or project-finance obligations, so capital adequacy cannot be modeled beyond qualitative balance-sheet strength. | 高 | SI001, SI011, SI012, SI023 |
| CI035 | No reviewed official or filing-type source disclosed annual revenue, ARR, gross margin, EBITDA, or net income. | 高 | SI001, SI023 |
| CI036 | Eastmoney pre-IPO commentary claimed 2024 revenue above RMB800 million, net margin above 20%, and order backlog into 2026. | 低 | SI025 |
| CI037 | That Eastmoney revenue-and-margin estimate is not independently verified by the reviewed official, filing-type, or mainstream independent sources, so it should not be treated as confirmed operating performance. | 中 | SI001, SI023, SI025 |
| CI038 | Jiemian reported roughly RMB100 million of R&D spend and a roughly even split between R&D and service staff, implying a cost base heavier than pure software. | 中 | SI016 |
| CI039 | The 科创板日报 article reposted on Xueqiu argues that smaller customers remain cost-sensitive and harder to onboard, leaving Mech-Mind still relatively dependent on head customers. | 中 | SI019 |
| CI040 | Eastmoney pre-IPO commentary flagged imported core-component dependence and possible EU export-control risk as threats to cost and overseas expansion. | 低 | SI026 |
| CI041 | Jiemian and the 科创板日报 article both frame 3D-vision robotics as an increasingly crowded market that requires continued product and ecosystem investment to defend share. | 中 | SI016, SI019 |
| CI042 | Info-clipper says Chinese-registry reports and financial statements exist for the company, but those materials were not publicly accessible in the reviewed set, leaving a filing-access gap ahead of any IPO diligence. | 中 | SI023 |
| CE001 | Mech-Mind's public 2026 product surface includes Mech-Eye industrial 3D cameras, Mech-Eye 3D laser profilers, Mech-Vision, Mech-Viz, Mech-DLK, Mech-MSR, Mech-Station InstaDepal, and the Eye-Brain-Hand station. | 高 | SE001, SE029 |
| CE002 | Mech-Mind's homepage claims 100+ Fortune Global 500 clients, 24,000+ cameras installed worldwide, and coverage of roughly 50 countries and regions. | 中 | SE001 |
| CE003 | KR-Asia and Eastmoney both describe Mech-Mind as operating at material deployment scale, citing 15,000+ cumulative installations or shipments and broad international reach. | 中 | SE022, SE027 |
| CE004 | The Mech-Eye industrial camera lineup covers working distances from roughly 300 mm to 3500 mm across short-, mid-, and long-range models. | 高 | SE002, SE009, SE024 |
| CE005 | Mech-Eye industrial cameras are publicly marketed as IP65-rated and certified across CE, FCC, VCCI, UKCA, KC, ISED, NRTL, and RoHS, with MTBF >= 100,000 hours. | 高 | SE002, SE009 |
| CE006 | The public Mech-Eye family spans UHP, NANO, PRO, LSR, and DEEP variants rather than a single sensor chassis. | 高 | SE002, SE009, SE024 |
| CE007 | Mech-Eye 3D laser profilers are positioned for measurement and inspection with 4,096 data points per profile, scan rates up to 15 kHz, and micron-level repeatability. | 高 | SE003, SE023, SE025 |
| CE008 | The laser-profiler line supports single-shot HDR plus C++/C#/Python APIs, GenICam compatibility, and GigE-based acquisition. | 高 | SE003, SE023 |
| CE009 | Mech-Vision is publicly positioned as a no-code graphical machine-vision environment for bin picking, machine tending, palletizing, depalletizing, and assembly workflows. | 高 | SE004, SE010 |
| CE010 | Official Mech-Vision materials attach 3D processing, model creation and matching, 2D/3D deep learning, and robot communication to one deployment environment. | 高 | SE004, SE010 |
| CE011 | Chinese Mech-Vision materials say the product integrates robot communication, 3D workpiece recognition, path planning, and production deployment in a single software surface. | 中 | SE010 |
| CE012 | Chinese Mech-Vision materials claim the software includes 1000+ robot models and can complete robot-communication tuning in 1-2 days. | 中 | SE010 |
| CE013 | Chinese Mech-Vision materials claim object-recognition accuracy above 99.99% and fastest recognition speed of 10 ms in real production settings. | 中 | SE010 |
| CE014 | Mech-Viz is publicly described as a code-free robot-programming environment with one-click 3D motion simulation. | 高 | SE005, SE011 |
| CE015 | Mech-Viz materials consistently emphasize motion planning, collision detection, and picking-strategy planning as core built-in functions. | 高 | SE005, SE011 |
| CE016 | Mech-Viz is positioned as robot-brand-agnostic through standardized communication and a unified workflow instead of robot-native programming languages. | 高 | SE005, SE011, SE021 |
| CE017 | Mech-DLK publicly covers advanced AI tasks such as object detection, segmentation, OCR, anomaly detection, and complex recognition rather than only simple inference. | 高 | SE006, SE012 |
| CE018 | Mech-DLK is marketed as an end-to-end training lifecycle spanning dataset management, labeling, training, validation, deployment, and model cascading. | 高 | SE006, SE012 |
| CE019 | Mech-DLK exposes SDKs in multiple languages including C, C++, C#, and Python for secondary development. | 高 | SE006, SE012 |
| CE020 | Chinese Mech-DLK materials claim average inference around 10 ms, roughly 40% faster than comparable products, with low overkill and miss rates. | 中 | SE012 |
| CE021 | The Eye-Brain-Hand station is explicitly described as combining Mech-Eye, Mech-GPT, and Mech-Hand into one embodied-intelligence stack. | 高 | SE007, SE027 |
| CE022 | Public Eye-Brain-Hand materials say the concept can run across single-arm, dual-arm, humanoid, retail, logistics, and industrial scenarios rather than one fixed robot form. | 高 | SE007, SE027 |
| CE023 | AW 2026 evidence shows Mech-Mind using Eye-Brain-Hand in transparent-object picking, humanoid shelf picking, and standardized depalletizing demonstrations. | 高 | SE028, SE029 |
| CE024 | AW 2026 materials say the Mech-Eye ULTRA M-GL was newly launched for transparent and translucent objects and would soon be officially launched. | 高 | SE028, SE029 |
| CE025 | AW 2026 materials also position the Mech-Eye AIC-Lite GL as a newly launched 2D camera series expected to launch later in 2026 for 2D+3D collaborative scenarios. | 高 | SE028, SE029 |
| CE026 | AW 2026 materials market Mech-Station InstaDepal as a standardized palletizing and depalletizing station with plug-and-play deployment in 30 minutes. | 中 | SE028 |
| CE027 | Official standard-interface documentation lists tested controller families and versions for ABB, FANUC, KUKA, UR, Yaskawa, Kawasaki, and other robot brands. | 高 | SE013, SE015 |
| CE028 | The Universal Robots marketplace page says Mech-Mind 3D Vision bundles Mech-Eye plus vision and robot-programming software and uses a plug-and-play URCap for UR integration. | 高 | SE021, SE008 |
| CE029 | Mech-Mind's UR ecosystem news post says the solution was extensively tested for compatibility and integration capabilities with Universal Robots cobots. | 中 | SE008 |
| CE030 | The KUKA documentation branch includes automatic calibration, example programs, command references, and error-message references, showing controller-specific adapter depth. | 高 | SE015, SE013 |
| CE031 | Mech-Mind's public engineering surface includes community-linked C++, Python, ROS, ROS2, and HALCON sample resources. | 中 | SE016, SE017 |
| CE032 | The ROS 1 interface documents Ubuntu 20.04, ROS Noetic, OpenCV, PCL, and point-cloud capture services for Mech-Eye cameras. | 中 | SE018 |
| CE033 | The ROS 2 interface documents Ubuntu 22.04, ROS Humble, OpenCV, PCL, and analogous point-cloud capture and parameter services. | 中 | SE020 |
| CE034 | The public mecheye_ros_interface release history shows a sample update to SDK 2.5.1 in October 2025. | 中 | SE019 |
| CE035 | Laser-profiler manual material says the device works through Mech-Eye SDK or third-party machine-vision software and supports a GenICam interface. | 中 | SE023, SE025 |
| CE036 | Mech-Mind's troubleshooting docs enumerate runtime failures including execution timeout, failed camera connection, motion singularity, invalid pick point, and robot collision detected. | 中 | SE014 |
| CE037 | The troubleshooting docs say the default timeout for getting Mech-Vision results is 10 seconds and explicitly recommend tuning timeout settings for longer-running projects. | 中 | SE014 |
| CE038 | KR-Asia reports that Mech-Mind internationalizes products only after validating them in China and then scales via local integrators and agents. | 中 | SE022 |
| CE039 | KR-Asia also says Mech-Mind built a Tokyo lab and maintains local teams in Japan, South Korea, Germany, and the United States for sales, engineering, and training. | 中 | SE022 |
| CE040 | Partner catalogs consistently describe Mech-Mind as adaptable to mainstream robot brands and supportive of secondary software development. | 中 | SE024, SE025, SE026 |
| CE041 | Partner catalogs market Mech-Mind on price advantage and reliability, including claims around pricing below comparable products and extended durability testing. | 中 | SE024, SE026 |
| CE042 | Eastmoney's WAIC 2025 coverage describes a first panoramic Eye-Brain-Hand showcase, a company workforce above 600, a self-owned camera factory, and cumulative shipments above 15,000 sets. | 中 | SE027 |
| CE043 | The official news index shows an active Feb-March 2026 product and event cadence around Eye-Brain-Hand and new launch announcements. | 高 | SE028, SE029 |
| CE044 | The retained official English and Chinese product and docs surfaces do not expose Mech-Recon as a standalone public 2026 product page, leaving its current scope under-documented. | 低 | SE001, SE009, SE010, SE011, SE012, SE013, SE029 |
| CE045 | The retained public materials are richer on capability claims than on independently audited deployment KPIs such as realized cycle time, attach rate, or error-rate distributions by module. | 中 | SE004, SE005, SE006, SE010, SE011, SE012 |
| CE046 | Across official pages, docs, manuals, and developer assets, Mech-Mind exposes SDK/API, ROS, GenICam, GigE Vision, and sample-code surfaces that are relatively open for an industrial-vision stack. | 高 | SE003, SE016, SE018, SE020, SE023 |
| CE047 | The public deployment flow links Mech-Eye capture, Mech-Vision recognition and communication, Mech-Viz planning and simulation, and robot execution without requiring custom code as the default story. | 高 | SE005, SE010, SE021 |
| CE048 | Mech-Vision's production interface and parameter-recipe tooling are designed for ongoing line monitoring and recipe switching rather than purely offline engineering. | 中 | SE010, SE014 |
| CE049 | Chinese Mech-Vision materials claim the production interface can shrink changeover and troubleshooting to minutes without direct engineer intervention. | 中 | SE010 |
| CE050 | The retained public corpus discloses hardware certifications and runtime controls much more clearly than software-security architecture, SLA boundaries, or attestation detail for the software stack. | 中 | SE001, SE013, SE014, SE029 |
| CE051 | The older camera + vision + planning stack is documented much more deeply than Eye-Brain-Hand, whose public evidence still leans heavily on launch and trade-show materials rather than manuals. | 中 | SE002, SE004, SE005, SE013, SE028, SE029 |
| CU001 | Current English and Chinese official pages each state that Mech-Mind has deployed 24,000+ units globally. | 高 | SU002, SU017 |
| CU002 | Current English and Chinese official pages each state that Mech-Mind serves 100+ Fortune 500 clients. | 高 | SU001, SU017 |
| CU003 | Current English and Chinese official pages each state that Mech-Mind covers roughly 50 countries or regions. | 高 | SU002, SU016 |
| CU004 | A3's member profile still uses an older baseline of 10,000+ cameras, 1,500+ clients, and 50+ countries. | 中 | SU013 |
| CU005 | KrASIA reported more than 15,000 installations in 50+ countries in 2025. | 中 | SU015 |
| CU006 | Public installed-base and client-count snapshots are time-stamped rather than directly comparable, because reviewed sources move from 10,000+ cameras and 1,500+ clients to 15,000+ installations and then 24,000+ deployed units. | 中 | SU013, SU015, SU002 |
| CU007 | Current English official surfaces name automotive, logistics, metal and machining, electronics, EV battery, and food and beverage as core served verticals. | 中 | SU001, SU002 |
| CU008 | Current Chinese official surfaces name automotive manufacturing, logistics handling, heavy industry, light industry, new energy, and industrial quality inspection as core served categories. | 中 | SU016, SU017 |
| CU009 | Official automotive pages document brake, axle, camshaft, wheel, tire, stamping, door-panel, glass-gluing, and spot-welding workflows. | 中 | SU003, SU008, SU009, SU010, SU011 |
| CU010 | The logistics page documents parcel induction, case, tote, and sack depalletizing, plus lead-acid-battery and magnesium-ingot handling. | 中 | SU004 |
| CU011 | The electronics page documents RJ45 inspection, washing-machine assembly, compressor-crankshaft loading, AC-feet bin picking, and counterweight handling. | 中 | SU005 |
| CU012 | The EV-battery page documents battery-cell depalletizing, module disassembly, module-to-pack assembly, and EV charging. | 中 | SU006 |
| CU013 | The metals page documents steel-plate bending, red-hot railway-wheel tending, track-shoe assembly, bolt tightening, and steel-bar machine tending. | 中 | SU007 |
| CU014 | Mech-Mind's 2025 inline-inspection news shows the company extending automotive proof from robot guidance into 100% inline inspection on production lines. | 中 | SU027 |
| CU015 | Reviewed public case pages emphasize application modules and workstation tasks rather than named enterprise accounts. | 中 | SU003, SU004, SU005, SU006, SU007 |
| CU016 | Across the fetched English and Chinese customer-facing pages reviewed for this chapter, Mech-Mind does not name BMW on its own public surfaces. | 中 | SU001, SU002, SU003, SU016, SU017 |
| CU017 | The BMW case study carried by the World Robot Conference attributes SortBot automation to UR hardware and BMW's own logistics program, not to Mech-Mind. | 中 | SU022 |
| CU018 | BMW Group's 2026 Leipzig humanoid announcement names Hexagon Robotics for that pilot. | 中 | SU023 |
| CU019 | The reviewed public evidence does not verify BMW as a named Mech-Mind customer. | 低 | SU016, SU022, SU023 |
| CU020 | Reviewed 2026 SAIC and SAIC-GM battery-line coverage attributes the named deployment to SAIC-GM and Zhiyuan/Nengzai rather than Mech-Mind. | 中 | SU024, SU025 |
| CU021 | The reviewed public evidence does not verify SAIC or SAIC-GM as named Mech-Mind customers. | 低 | SU024, SU025, SU016 |
| CU022 | Across the fetched customer-facing pages reviewed for this chapter, Mech-Mind does not name Geely or 吉利. | 低 | SU001, SU002, SU003, SU016, SU017 |
| CU023 | Across the fetched customer-facing pages reviewed for this chapter, Mech-Mind does not name CATL, BYD, Foxconn, or JD Logistics. | 低 | SU001, SU002, SU006, SU016, SU017 |
| CU024 | Public named-customer proof is materially weaker than public reach claims because official surfaces disclose sector breadth and use-case depth but not an auditable named customer list. | 中 | SU001, SU002, SU013, SU016, SU017 |
| CU025 | KrASIA says Mech-Mind enters new markets only after products are validated in China. | 中 | SU015 |
| CU026 | KrASIA says regional system integrators and agents handle deployment and promotion while Mech-Mind focuses on product development and after-sales support. | 中 | SU015 |
| CU027 | The Chinese official about page says Mech-Mind provides integrator training, reference-solution design, exhibition support, and a full delivery and after-sales system. | 中 | SU017 |
| CU028 | UR+ membership confirms plug-and-play integration with Universal Robots cobots for bin picking, machine tending, assembly, palletizing, and depalletizing. | 中 | SU012 |
| CU029 | Current English and Chinese official pages list offices or facilities in China, the US, Germany, Japan, and Korea. | 高 | SU001, SU017 |
| CU030 | AW 2026 and iREX 2025 official pages show ongoing international trade-show investment aimed at new customers and partners in North America and Japan. | 中 | SU026, SU028 |
| CU031 | Automate 2026 exhibitor materials position Mech-Mind around Chicago and emphasize logistics, food and beverage, and automotive use cases for US buyers. | 中 | SU014 |
| CU032 | The documentation site provides typical case practices, tutorials, and integration manuals for 3D robot guidance. | 中 | SU021 |
| CU033 | Public deployment depth appears strongest in automotive because the official library spans more distinct workflows there than in any other named vertical. | 中 | SU003, SU008, SU009, SU010, SU011, SU004, SU005, SU006, SU007 |
| CU034 | Logistics and EV-battery deployments are publicly evidenced mainly through use-case pages rather than named customer endorsements. | 中 | SU004, SU006, SU016 |
| CU035 | A3, HowToRobot, and Craft all describe Mech-Mind as a supplier or software company serving automation partners and industrial users. | 中 | SU013, SU019, SU020 |
| CU036 | No fetched official or third-party profile source discloses NRR, GRR, churn, or renewal rates. | 中 | SU001, SU002, SU013, SU015, SU017 |
| CU037 | No fetched official or third-party profile source discloses top-customer share, top-10 share, or customer concentration by ARR. | 中 | SU001, SU002, SU013, SU015, SU017 |
| CU038 | Public evidence supports production deployment at the cell or workflow level, but multi-site expansion is usually implied by aggregate installed-base claims rather than named account rollouts. | 中 | SU002, SU003, SU004, SU005, SU006, SU007, SU015 |
| CU039 | Xueqiu's 2025 profile repeats global 15,000+ shipped units and near-100 Fortune 500 factory usage in automotive, logistics, new energy, and 3C electronics, but it still does not provide an auditable named customer list. | 低 | SU018 |
| CU040 | HowToRobot's 500-999 employee band is directionally consistent with the Chinese official disclosure of 600+ employees supporting sales, delivery, training, and after-sales. | 中 | SU019, SU017 |
| CU041 | Current English and Chinese pages signal expansion beyond manufacturing into retail, service, or household scenarios, but the strongest public customer proof remains industrial. | 中 | SU002, SU016 |
| CU042 | The main remaining diligence bottleneck is not cross-vertical deployment existence but which named accounts renew, expand, and concentrate revenue. | 中 | SU002, SU013, SU017 |
| CR001 | Current English and Chinese company profile pages say Mech-Mind has deployed more than 24,000 units in nearly 50 countries and serves 100+ Fortune Global 500 customers. | 高 | SR001, SR002 |
| CR002 | The Chinese company profile says Mech-Mind has 600+ global employees, a self-owned camera factory, and a complete delivery, training, and after-sales system. | 中 | SR002 |
| CR003 | Mech-Mind says it supports integrators and partners across consulting, solution design, training, deployment, maintenance, and competitive pricing, which implies a services-heavy industrial-sales model rather than a self-serve software model. | 中 | SR001 |
| CR004 | A newer official reliability page cites 17,000+ installed cameras across 40+ countries, which is directionally strong but lower than the 24,000-unit figure on current about pages. | 中 | SR003 |
| CR005 | KrASIA reported that Mech-Mind technology was operating in 50+ countries with 15,000+ installations in 2024, adding a third public denominator for deployment scale. | 中 | SR030 |
| CR006 | KrASIA says Mech-Mind introduces products abroad only after they have been validated in China because overseas after-sales logistics are more complex. | 中 | SR030 |
| CR007 | KrASIA says Mech-Mind relies on local system integrators and agents in each region for deployment and promotion. | 中 | SR030 |
| CR008 | QbitAI says Mech-Mind has established subsidiaries in Germany, Japan, the United States, and Korea as part of a localized global operating system. | 中 | SR010 |
| CR009 | Mech-Mind’s Automate 2025 page says the company showcased 10+ solutions across automotive manufacturing, smart logistics, and AI quality inspection at an event with 870 exhibitors and 40,000+ visitors. | 中 | SR011 |
| CR010 | Keyence’s 3D VGR system offers automatic robot-camera calibration, CAD upload, collision-free path planning, and built-in picking simulation for assembly, depalletizing, and machine tending. | 中 | SR023 |
| CR011 | Cognex’s SEC filing says logistics, automotive, and consumer electronics together represented about 60% of 2024 revenue. | 中 | SR022 |
| CR012 | Cognex’s SEC filing says about 18% of 2024 revenue came from Greater China. | 中 | SR022 |
| CR013 | Cognex’s SEC filing says its primary contract manufacturers are located in Indonesia and Malaysia. | 中 | SR022 |
| CR014 | Cognex’s SEC filing lists export and import restrictions and trade tariffs among the international risks facing its machine-vision business. | 中 | SR022 |
| CR015 | Qviro’s 2026 ranking places Cognex, Keyence, and Basler among the top robot-vision brands. | 中 | SR024 |
| CR016 | Intel Market Research says the vision-guided robotics market is led by Cognex and Keyence, with Basler, Omron, robot OEMs, and Mech-Mind competing in the same field. | 中 | SR025 |
| CR017 | Intel Market Research says deploying vision-guided robotics can require upfront investment that often exceeds $100,000 per robotic cell. | 中 | SR025 |
| CR018 | Intel Market Research says integration with existing robotic arms and legacy factory infrastructure can prolong setup and disrupt operations during transition. | 中 | SR025 |
| CR019 | Intel Market Research says uncontrolled lighting, reflective surfaces, and scarcity of skilled technicians remain meaningful deployment restraints for machine-vision robotics. | 中 | SR025 |
| CR020 | Mech-Mind’s reliability blog says temperature fluctuations, vibration, dust, ambient-light interference, and continuous operation can cause 3D cameras to drift, lose accuracy, or fail over time. | 中 | SR003 |
| CR021 | The same reliability blog says Mech-Eye cameras are designed for harsh environments with IP65/IP67 protection and certified MTBF of up to 100,000 hours. | 中 | SR003 |
| CR022 | The reliability blog says Mech-Mind provides auto-correction tools that compensate for drift caused by temperature changes or component aging. | 中 | SR003 |
| CR023 | The reliability blog says manual calibration, custom scripting, and toolchain assembly can slow projects and waste engineering time. | 中 | SR003 |
| CR024 | Mech-Mind’s reflective-parts blog says high reflectivity, ambient light, clutter, stacking, and occlusion can compromise recognition and operational stability. | 中 | SR004 |
| CR025 | The reflective-parts blog says some automotive body-panel workflows still require about ±0.5 mm positioning accuracy plus thermal compensation. | 中 | SR004 |
| CR026 | The transparent-objects blog says transparent-object handling is a full-stack challenge where unstable perception becomes downtime, rework, or missed cycle times. | 中 | SR005 |
| CR027 | The 2026 automation-goals blog says manufacturers are pushing for higher mix, tighter tolerances, and zero-defect quality, raising the performance bar for industrial vision systems. | 中 | SR006 |
| CR028 | BIS announced on 2026-01-13 that certain advanced semiconductor exports to China, including Nvidia H200-class products, would move to case-by-case license review under strict conditions. | 高 | SR013, SR014 |
| CR029 | Finnegan says the 2026 BIS pathway still requires abundant U.S. supply, a 50% cap on China/Macau shipments, rigorous know-your-customer checks, and independent U.S. testing. | 中 | SR014 |
| CR030 | Finnegan says reexports and in-country transfers of the same items remain subject to a presumption of denial. | 中 | SR014 |
| CR031 | HKEX and Han Kun both show that robotics and automation fall within the Chapter 18C specialist-technology listing route. | 高 | SR015, SR016 |
| CR032 | Han Kun says a Chapter 18C commercial company needs at least HK$250 million of revenue in its most recent audited financial year, while a pre-commercial company is a separate category. | 中 | SR016 |
| CR033 | HKEX says temporary modifications to Chapter 18C minimum initial market-cap requirements run from 2024-09-01 to 2027-08-31. | 中 | SR015 |
| CR034 | HKCD, Tencent, HKET, and AASTOCKS all reported that Mech-Mind confidentially filed for a Hong Kong IPO targeting roughly US$200 million (about HK$1.56 billion). | 中 | SR017, SR018, SR019, SR020 |
| CR035 | The same IPO reports say final fundraising size and timing were still undecided. | 中 | SR017, SR019 |
| CR036 | Aiqicha says the Xiong’an entity has 18 business disputes, 3 filing records, 6 hearing announcements, and 4 litigation relationships. | 中 | SR021 |
| CR037 | Aiqicha also says the Xiong’an entity has 954 trademarks, 384 patents, and 47 software copyrights. | 中 | SR021 |
| CR038 | CommBank reports China’s official manufacturing PMI fell to 49.3 in January 2026, with new orders at 49.2 and new export orders at 47.8. | 中 | SR027 |
| CR039 | CNBC reports that the private China manufacturing PMI fell to 49.9 in November 2025 while the official PMI remained in contraction at 49.2 for an eighth straight month. | 中 | SR026 |
| CR040 | NBR says China’s slowdown will have adverse effects on manufacturing exporters and accelerate supply-chain rearrangement as trade relations worsen. | 中 | SR028 |
| CR041 | Deloitte’s 2026 manufacturing outlook references sub-50 PMI conditions, tariff tracking, and supply-chain resilience pressure, underscoring a cautious capex backdrop. | 中 | SR029 |
| CR042 | Official company materials emphasize automotive, logistics, EV batteries, heavy industry, electronics, and food & beverage, but none disclose revenue split, top-customer dependence, or renewal concentration. | 中 | SR001, SR002, SR003 |
| CR043 | Public materials emphasize no-code or low-code deployment, simulation, training, and support, which suggests adoption still runs through integrator-led engineering rather than low-friction self-serve software. | 中 | SR001, SR003, SR030 |
| CR044 | Public IPO sources disclose a confidential filing rumor but not audited revenue, profitability, or customer concentration data, leaving Chapter 18C eligibility unresolved in public evidence. | 中 | SR015, SR016, SR017, SR018, SR019 |
| CR045 | The mismatch between 15,000+ installations, 17,000 cameras, and 24,000+ units across public sources indicates disclosure drift that complicates underwriting of deployment momentum. | 中 | SR001, SR002, SR003, SR030 |
| CR046 | Mech-Mind’s Value Provider Cooperation page and KrASIA both indicate a formal partner and integrator cooperation model. | 中 | SR007, SR030 |
| CR047 | English and Chinese contact pages show offices in Germany, the U.S., Japan, Korea, Beijing, and Shanghai, confirming a geographically distributed support network. | 中 | SR008, SR009 |
| CR048 | QbitAI says deployment is accelerating most clearly in manufacturing and logistics tasks with clear boundaries and limited direct human interaction, implying concentration in bounded industrial use cases rather than broad humanoid generality. | 中 | SR010 |
| CR049 | KrASIA and QbitAI show a real mitigation strategy: validate products in China first, then scale through local partners, subsidiaries, and training capacity abroad. | 中 | SR010, SR030 |
| CR050 | The main remaining public-diligence blockers are audited revenue and Chapter 18C fit, customer concentration, independent uptime, competitive win-loss data, and export-control BOM exposure. | 中 | SR003, SR015, SR016, SR021, SR030 |
| CR051 | No public win-loss or pricing-discount dataset against Cognex or Keyence was found in the retained sources. | 中 | SR001, SR024, SR025, SR030 |
| CR052 | No public BOM disclosure in retained sources identifies which compute or semiconductor inputs would be directly exposed to BIS-sensitive licensing. | 中 | SR001, SR006, SR013, SR014 |
| CR053 | AAStocks independently repeats the confidential Hong Kong filing and roughly US$200 million target, but provides only a truncated wire summary, underscoring shallow public disclosure around the rumored listing. | 低 | SR020 |
| CR054 | Mech-Mind’s AW 2026 page shows the company is still broadening product and application scope in 2026, which creates commercial opportunity but also adds execution load. | 中 | SR012 |
| CV001 | Multiple August 2025 reports confirm that Mech-Mind completed a new financing round of about CNY500 million. | 高 | SV001, SV002, SV003 |
| CV002 | The August 2025 round added investors including Xiong'an-linked funds, Dayang Motor, Huachuang Capital, CICC Porsche Venture Capital, and regional funds. | 中 | SV001, SV002, SV003 |
| CV003 | Public reports say the August 2025 proceeds are earmarked for further development of Mech-Mind's embodied-intelligence “eye-brain-hand” stack, broader product lines, and global commercialization. | 中 | SV002, SV003, SV004 |
| CV004 | Accessible public coverage confirms that Mech-Mind also completed a 2022 Series D financing round worth CNY500 million. | 中 | SV007 |
| CV005 | Accessible public evidence shows a directional valuation step-up from a 2022 late-stage private round to explicit unicorn framing by 2025-2026, even though the exact 2022 post-money is not cleanly fetchable. | 中 | SV007, SV008, SV009 |
| CV006 | China Daily's Hurun gazelle coverage says Mech-Mind had already surpassed a $1 billion valuation and was set to join the Hurun Unicorn List in 2026. | 中 | SV008 |
| CV007 | Separate Xinhua and State Council reporting also describe Mech-Mind as a Chinese unicorn company. | 高 | SV009, SV010 |
| CV008 | English and Chinese IPO reports consistently say Mech-Mind is planning a Hong Kong IPO targeting about $200 million of proceeds. | 高 | SV005, SV006, SV011, SV025, SV026, SV027, SV030 |
| CV009 | The IPO reporting also says the company filed confidentially and that size and timing remain subject to change. | 高 | SV005, SV011, SV025, SV030 |
| CV010 | Recent IPO and fundraising coverage puts Mech-Mind's cumulative capital raised at roughly RMB2 billion. | 中 | SV006, SV026, SV029, SV030 |
| CV011 | HKEX Chapter 18C created a listing path for specialist technology companies that may not yet meet conventional revenue- or profit-based listing tests. | 高 | SV012, SV020 |
| CV012 | Robotics and automation sit inside Chapter 18C's advanced hardware and software industry bucket. | 高 | SV012, SV020 |
| CV013 | Under the temporary Chapter 18C thresholds described by HKEX, a commercial company needs at least HK$4 billion of expected market cap and a pre-commercial company needs at least HK$8 billion. | 中 | SV020 |
| CV014 | Public sources do not disclose whether Mech-Mind would qualify as a commercial or pre-commercial Chapter 18C issuer because latest audited revenue is not public. | 中 | SV020, SV005, SV006 |
| CV015 | The public record therefore supports Hong Kong as a plausible venue but not a fully underwritten listing route or classification. | 中 | SV012, SV020, SV005 |
| CV016 | Symbotic reported $2.247 billion of revenue for fiscal 2025. | 中 | SV022 |
| CV017 | Symbotic's market capitalization was about $32.61 billion in May 2026. | 中 | SV013 |
| CV018 | Using those public figures, Symbotic screens at roughly 14.5x market-cap-to-revenue, illustrating how high-growth automation platforms can sustain double-digit public multiples. | 中 | SV013, SV022 |
| CV019 | Cognex reported $914.5 million of 2024 revenue. | 中 | SV023 |
| CV020 | Cognex's market capitalization was about $10.99 billion in May 2026. | 中 | SV014 |
| CV021 | Using those public figures, Cognex screens at roughly 12.0x market-cap-to-revenue. | 中 | SV014, SV023 |
| CV022 | Zebra reported $4.981 billion of 2024 net sales. | 中 | SV024 |
| CV023 | Zebra's market capitalization was about $12.17 billion in May 2026. | 中 | SV015 |
| CV024 | Using those public figures, Zebra screens at roughly 2.4x market-cap-to-sales, anchoring a much lower multiple for diversified automation hardware. | 中 | SV015, SV024 |
| CV025 | Teradyne's Robotics segment is made up of Universal Robots and Mobile Industrial Robots. | 中 | SV021 |
| CV026 | Teradyne's Robotics revenue declined 2.8% to $364.8 million in 2024. | 中 | SV021 |
| CV027 | Teradyne's market capitalization was about $56.11 billion in May 2026. | 中 | SV017 |
| CV028 | Teradyne is not a clean direct comp because the robotics business sits inside a much larger test-equipment company with very different segment economics. | 中 | SV017, SV021 |
| CV029 | Mind Robotics raised $500 million in a March 2026 Series A that reportedly valued the company at around $2 billion. | 中 | SV019 |
| CV030 | Simply Wall St warned investors to watch ownership, revenue sharing, and future funding or IPO disclosures before value crystallizes in Mind Robotics. | 中 | SV018 |
| CV031 | Private industrial-robotics rounds show strong capital appetite, but disclosure often lags valuation marks. | 中 | SV018, SV019 |
| CV032 | The machine vision market was estimated at $23.48 billion in 2026 and projected to reach $35.43 billion by 2030. | 中 | SV028 |
| CV033 | Large market size supports upside optionality for Mech-Mind, but market breadth alone does not prove company-level unit economics. | 中 | SV028, SV023 |
| CV034 | Public evidence is enough to confirm that Mech-Mind is a late-stage robotics platform with unicorn-level scale and active financing optionality. | 中 | SV001, SV008, SV009, SV010 |
| CV035 | Public evidence is not enough to price revenue quality, gross margin, burn, retention, or customer concentration precisely. | 中 | SV005, SV006, SV018 |
| CV036 | Public evidence is also not enough to price dilution, liquidation preferences, or board-control overhang. | 中 | SV005, SV018, SV026 |
| CV037 | Among public names, Symbotic is the closest high-growth warehouse-automation reference even though it is more systems-installation heavy than Mech-Mind. | 中 | SV013, SV022 |
| CV038 | Cognex is the cleaner machine-vision disclosure reference, but it is narrower in workflow scope and more inspection-centric than Mech-Mind. | 中 | SV014, SV023 |
| CV039 | The most supportable current recommendation is research-more rather than buy. | 中 | SV001, SV005, SV008, SV018 |
| CV040 | Recommendation confidence should stay medium because funding and IPO optionality are visible while operating disclosure remains thin. | 中 | SV005, SV008, SV018 |
| CV041 | Valuation stance is stretched because unicorn framing arrived before public financial proof, even if sector comps show investors will pay up for automation growth. | 中 | SV013, SV014, SV018, SV019 |
| CV042 | A public-data fair-value band can only be broad and anchored near the unicorn floor plus 18C thresholds, not a precise IPO ceiling. | 中 | SV008, SV020 |
| CV043 | The most important diligence asks are audited 2024-2025 financials, customer/order-book cohorts, cap-table and preference terms, and draft listing disclosures. | 中 | SV005, SV018, SV020 |
| CV044 | Bull-case upside requires disclosed revenue scale, durable international deployments, and a software-like margin profile. | 中 | SV028, SV001, SV009 |
| CV045 | Bear-case compression would follow if an IPO or next round reveals a heavier hardware-services mix, slower growth, or weaker margins than unicorn pricing assumes. | 中 | SV018, SV021, SV024 |
| CV046 | Thesis-break triggers are a pulled IPO, a down-round, or audited disclosure that pushes Mech-Mind toward lower-multiple automation-hardware comparables. | 中 | SV018, SV020, SV024 |
| CV047 | The 2025 financing round proves investor appetite and commercialization ambition, but by itself it does not prove sustainable monetization. | 中 | SV001, SV003, SV018 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Mech-Mind Robotics | About Us | Mech-Mind Robotics | Founded in 2016, Mech-Mind Robotics is a global leader in embodied intelligence robotics... With 24,000+ units deployed across nearly 50 countries and regions. |
| SO002 | Mech-Mind Robotics | Our Team | Mech-Mind Robotics | Tianlan Shao — Founder and CEO — Bachelor's degree from School of Software, Tsinghua University. Master's degree from Technical University of Munich. |
| SO003 | Mech-Mind Robotics | Contact Us | Mech-Mind Robotics | |
| SO004 | Mech-Mind Robotics | News | Mech-Mind Robotics | 03/12/2026 Mech-Mind at AW 2026 ... 12/08/2025 Join Mech-Mind at iREX 2025 ... |
| SO005 | Mech-Mind Robotics | Statement on Anti-Corruption | Mech-Mind Robotics | Mech-Mind has established the Integrity Compliance Committee as our internal supreme authority to process Anti-Corruption related affairs. |
| SO006 | Mech-Mind Robotics | Embodied Intelligence "Eye-Brain-Hand" Robot Station | General-Purpose Embodied AI | The Embodied Intelligence “Eye-Brain-Hand” Robot Station integrates ... Mech-Eye, Mech-GPT, and Mech-Hand. |
| SO007 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | |
| SO008 | Mech-Mind Robotics | Mech-Mind at iREX 2025 | Winning Industry-Wide Acclaim for Full-Stack Robot "Eye-Brain-Hand" Showcase and Global Product Premieres | |
| SO009 | 梅卡曼德机器人 | 关于我们 - 梅卡曼德机器人 | 梅卡曼德机器人由清华海归团队于2016年创办... 600+ 全球员工 ... 20亿+ 累计融资 ... 24000+ 全球落地相机台数。 |
| SO010 | 梅卡曼德机器人 | 联系我们 - 梅卡曼德机器人 | 全球10+业务中心,业务覆盖近50国家及地区。 |
| SO011 | 36Kr | 36氪独家 | “AI+3D视觉+机器人”解决方案提供商「梅卡曼德机器人」再获近10亿元C系列融资 | 梅卡曼德机器人近期再次完成C系列融资近10亿元。本轮融资由美团、IDG资本领投,老股东红杉中国、源码资本跟投。 |
| SO012 | 36Kr | 创·问——梅卡曼德机器人邵天兰:具身智能没有“英雄主义”,只有“魔鬼细节” | 今天,梅卡曼德的产品服务了全球100+的《财富》500强客户,业务覆盖了五十多个国家和地区,连续五年市占率第一。 |
| SO013 | 36Kr PitchHub | 梅卡曼德机器人 | 项目信息-36氪 | E轮 北京市 2016年09月 ... 2025-03 E轮 ... 2024-12 D++轮 ... 2023-08 D轮。 |
| SO014 | 量子位 | 达沃斯聚焦技术新前沿,梅卡曼德创始人邵天兰受邀分享具身智能落地实践 | |
| SO015 | KrASIA | This Chinese robotics firm is making factory AI modular, global, and scalable | Fast forward to 2024, and Mech-Mind’s technology is reportedly operating in over 50 countries, with more than 15,000 installations worldwide. |
| SO016 | KrASIA | China’s AI firms look outward as WAIC 2025 takes on global flavor | |
| SO017 | Asia ICT / PencilNews / Great Wall Strategy Consulting | Tsinghua University Nurtures Embodied AI Unicorn: Mech-Mind Equips Robots with Vision and Intelligence, Securing Nearly RMB 2 Billion in Funding | Since its inception, Mech-Mind has garnered multiple rounds of funding ... amounting to nearly RMB 2 billion. |
| SO018 | MarketScreener / S&P Capital IQ | Mech-Mind Robotics Technologies Ltd. announced that it has received CNY 500 million in funding from a group of investors | Mech-Mind Robotics Technologies Ltd. announced that it has received CNY 500 million in a round of funding on August 25, 2025. |
| SO019 | AASTOCKS | <IPO>Meituan-Backed AI Firm Mech-Mind Files Confidentially for HK Listing: Wire | |
| SO020 | 香港商報 | 【新股最前線】梅卡曼德機器人據報秘密申港IPO 集資15.6億 | 梅卡曼德機器人已獲IDG、美團、紅杉中國等多輪投資,累計融資20億元人民幣。 |
| SO021 | 香港經濟日報 HKET | 新股IPO|梅卡曼德機器人據報秘密申港IPO集資15.6億 從事AI機械視覺軟件 | |
| SO022 | Automate | Mech-Mind Robotics-3D Vision and Automation Solutions | Since founded in 2016, we have delivered 10,000+ industrial 3D cameras in 50+ countries and regions and served 1,500+ clients worldwide. |
| SO023 | Automate Show | Mech-Mind Robotics Technologies Co. Ltd. | With 24,000+ units deployed across nearly 50 countries and regions, Mech-Mind’s “Eye + Brain” solutions power industries such as logistics, F&B, and automotive. |
| SO024 | Tracxn | Mech-Mind | Founded Year 2016 ... Stage Series C ... Mech-Mind has raised a total funding of $200M over 6 rounds. |
| SO025 | 爱企查 | 梅卡曼德(雄安)机器人科技股份有限公司 - 爱企查 | 该公司曾涉及18项经营纠纷、3件立案信息、6个开庭公告、4起涉诉关系。 |
| SM001 | Mech-Mind Robotics | Embodied AI & 3D Vision for Robots and More | Mech-Mind provides 3D cameras and software suites for robotic system integrators. Typical 3D vision applications include bin picking, depalletizing, machine tending, piece picking and item picking. We have delivered 3000+ applications for 1000+ clients worldwide. |
| SM002 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | The system can be widely applied across industries such as food, medicine, e-commerce, and logistics, and its high-precision 3D camera and AI software are also demonstrated on axle shafts, sheet metal parts, splines, motors, ECU glue beads, tires, and battery cells. |
| SM003 | Industry Asia Pacific | Mech-Mind at Automation World 2026 | This full-stack approach enables autonomous bin picking and piece picking of mixed SKUs, 100% inline inspection for automotive parts, and no-code deployment in hours instead of weeks. |
| SM004 | Automate | Mech-Mind Robotics-3D Vision and Automation Solutions | Mech-Mind is an industry-leading company focusing on industrial 3D camera and software suite for intelligent robotics. Since founded in 2016, we have delivered 10,000+ industrial 3D cameras in 50+ countries and regions and served 1,500+ clients worldwide. |
| SM005 | KrASIA | This Chinese robotics firm is making factory AI modular, global, and scalable | Fast forward to 2024, and Mech-Mind’s technology is reportedly operating in over 50 countries, with more than 15,000 installations worldwide. |
| SM006 | International Federation of Robotics | World Robotics 2025 report – INDUSTRIAL ROBOTS – released by IFR | The new World Robotics 2025 statistics on industrial robots showed 542,000 robots installed in 2024. China is by far the world’s largest market, representing 54% of global deployments. |
| SM007 | International Federation of Robotics | Top 5 Global Robotics Trends 2026 | The global market value of industrial robot installations has reached an all-time high of US$ 16.7 billion and future demand will be driven by AI autonomy, IT/OT convergence, safety, and labor gaps. |
| SM008 | International Federation of Robotics | Collaborative Robots - How Robots Work alongside Humans | Manufacturing industries have been early adopters of cobot technology. This includes automotive, electronics, aerospace, consumer goods, pharmaceuticals, logistics and warehousing. |
| SM009 | SupplyChain247 | Labor Shortages Fuel Robotics Growth in Warehouses, New Study Finds | 55% cited labor availability constraints as the #1 motivator, 42% cited labor costs, and only 32% have approved funding for new robotics initiatives. |
| SM010 | Logistics Viewpoints | ProMat 2025: Robotics Steps Up to Tackle the Warehouse Labor Crisis | ProMat 2025 felt particularly focused on solutions designed to alleviate the strain on human capital, with robotics taking center stage as a powerful and increasingly viable answer. |
| SM011 | OSHA | OSHA Technical Manual (OTM) - Section IV: Chapter 4 | OSHA’s technical manual describes robot-system hazards, safeguarding, operation, maintenance, and personnel protection as integral deployment requirements for industrial robotics. |
| SM012 | Interact Analysis | Warehouse Automation - Research Products | We have produced the report through extensive research, conducting more than 100 in-depth research interviews and analyzing more than 120 companies. |
| SM013 | Grand View Research | Warehouse Robotics Market Size & Trends Report, 2030 | The global warehouse robotics market size was estimated at USD 4.31 billion in 2022 and is projected to reach USD 17.29 billion by 2030, growing at a CAGR of 19.6% from 2023 to 2030. |
| SM014 | Mordor Intelligence | Warehouse Automation Market - Industry Size & Growth 2025 - 2031 | The Warehouse Automation Market size is expected to increase from USD 29.98 billion in 2025 to USD 34.17 billion in 2026 and reach USD 65.74 billion by 2031. |
| SM015 | Precedence Research | Warehouse Automation Market Size To Hit USD 107.36 Bn By 2035 | The global warehouse automation market size accounted for USD 25.27 billion in 2025 and is predicted to increase from USD 29.30 billion in 2026 to approximately USD 107.36 billion by 2035. |
| SM016 | Precedence Research | Machine Vision Market Size to Surpass USD 76.89 Billion by 2035 | The global machine vision market size accounted for USD 23.06 billion in 2025 and is predicted to increase from USD 26.07 billion in 2026 to approximately USD 76.89 billion by 2035. |
| SM017 | Vision Systems Design | Key Economic Insights from the 2026 A3 Business Forum | The biggest growth areas are 3D vision software, bin picking, and AI applications. Logistics and warehousing represent the fastest growing market segment for machine vision, with anticipated annual growth near 14.2%. |
| SM018 | IoT Analytics | AI in machine building 2026: Adoption, barriers, use cases, and leading sub-industries | 54% of machine builders cite high AI costs as a critical barrier, while insufficient data infrastructure and workforce skill gaps both stand at 43%. |
| SM019 | Yicai Global | China's Industrial Robot Installations Rose 5% Last Year Amid Global Decline | China accounted for 54 percent of the global market and robot density reached 470 units per 10,000 workers, ranking only behind South Korea and Singapore. |
| SM020 | State Council Information Office | IFR: China leads global industrial robot market with record installations | China's industrial robot stock reached a record 2,027,000 units in 2024. In China, the electrical and electronics sector continued to lead demand with 83,000 units installed in 2024, followed by the automotive industry with 57,200 units. |
| SM021 | People's Daily Online | China's industrial robot industry expands rapidly | China's robot density reached 470 units per 10,000 employees in 2023. This figure more than doubled compared to 2019, according to a report by the IFR. |
| SM022 | ChinaPower / CSIS | Is China Leading the Robotics Revolution? | In 2024, China installed 295,000 new industrial robots. In the electronics industry, robotics adoption has grown at an average rate of 16 percent annually since 2019, with a total of 83,000 units installed in 2024. |
| SM023 | RoboticsTomorrow | As 2026 Approaches, U.S. Manufacturers Call Automation Critical: Yet Most Still Lag in Adoption, New Report Finds | While 92% of manufacturers agree automation is essential for long-term competitiveness, only 37% report having significant or full automation in place. 50% struggle to identify the right technology. |
| SM024 | The Robot Report | IFR: industrial robot deployments have doubled in 10 years | The IFR today released its World Robotics 2025 Report that showed 542,000 industrial robots were installed worldwide in 2024 and China represented 54% of global deployments. |
| SM025 | Business Wire | Europe’s Auto Industry Installed 23,000 New Robots – IFR Reports | Car makers account for around a third of annual manufacturing installations in Europe and the combined number of 23,000 automotive robot installations was ahead of North America in 2024. |
| SM026 | 36Kr | "World Robotics Report 2025" Released: China Dominates Global Market Share, India Ranks 6th in Comeback, Japan, US, South Korea, Germany See Declines | In 2024, the electrical/electronics industry had an installation volume of 129,000 units while the automotive industry had an installation volume of 126,000 units; China's electronics installs reached 83,000 and automotive installs 57,200. |
| SM027 | International Federation of Robotics | China’s 15th Five-Year Plan (2026-2030) marks pivot to innovation | China’s huge domestic market offers enormous potential: the share of local suppliers in domestic industrial robot installations increased from 30% in 2020 to 57% in 2024, and 64% of industrial robots in the global electronics industry are installed in China. |
| SP001 | Mech-Mind Robotics | Embodied AI & 3D Vision for Robots and More | |
| SP002 | Mech-Mind Robotics | Embodied Intelligence "Eye-Brain-Hand" Robot Station | |
| SP003 | Mech-Mind Robotics | Mech-Mind Documentation | |
| SP004 | Association for Advancing Automation (A3) | Mech-Mind Robotics-3D Vision and Automation Solutions | Member of A3 | |
| SP005 | KrASIA | This Chinese robotics firm is making factory AI modular, global, and scalable | |
| SP006 | AsiaICT | Tsinghua University Nurtures Embodied AI Unicorn: Mech-Mind Equips Robots with Vision and Intelligence, Securing Nearly RMB 2 Billion in Funding | |
| SP007 | 梅卡曼德机器人 | 梅卡曼德机器人- 通用智能机器人的AI大脑和3D视觉 | |
| SP008 | KEYENCE | 3D Vision-Guided Robotics - 3D VGR series | |
| SP009 | KEYENCE | Vision Systems | |
| SP010 | Photoneo | Machine Vision and Automation Solutions | Photoneo Focused on 3D | |
| SP011 | Photoneo | Automated Bin Picking | Photoneo Focused on 3D | |
| SP012 | Photoneo | 3D Camera | MotionCam | Photoneo Focused on 3D | |
| SP013 | Intrinsic | Intrinsic | |
| SP014 | Intrinsic | Intrinsic Flowstate | |
| SP015 | Intrinsic | Intrinsic 3D vision system | |
| SP016 | Intrinsic | Accelerating Physical AI: FANUC Integrates with Intrinsic and Flowstate | |
| SP017 | Roboception | rc_visard 3D-Stereosensor | |
| SP018 | OMRON Robotics | OMRON Robotics | Transforming Manufacturing with Robotics | |
| SP019 | Universal Robots | Robotic Arm | Robot Arms for Industrial Automation | |
| SP020 | Universal Robots | Universal Robots Pricing Guide - Cost Factors and Budget Planning | |
| SP021 | Standard Bots | Universal Robots price guide: What to expect (new and used costs) | |
| SP022 | Qviro | Top 10 Robot Vision Manufacturers & Brands 2026 | |
| SP023 | Intel Market Research | Machine Vision Vision Guided Robotics Market Outlook 2026-2034 | |
| SP024 | QYResearch | Global 3D Vision for Robot Sales Market Report, Competitive Analysis and Regional Opportunities 2025-2031 | |
| SP025 | Pickit | Rethink production efficiency with flexible and smart 3D robot vision | |
| SP026 | ISRA VISION | About us | |
| SP027 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | |
| SI001 | Mech-Mind Robotics | About Us | Mech-Mind Robotics | Founded in 2016, Mech-Mind has closed its Series C+ round with total funding of USD 300 million. |
| SI002 | Mech-Mind Robotics | 梅卡曼德机器人- 通用智能机器人的AI大脑和3D视觉 | 100+《财富》500强客户 |
| SI003 | Mech-Mind Robotics | Mech-Eye Industrial 3D Cameras | Advanced 3D Sensor | Mech-Mind Robotics | |
| SI004 | Mech-Mind Robotics | Vision Software | Mech-Vision Machine Vision Software | Mech-Mind Robotics | |
| SI005 | Mech-Mind Robotics | Mech-Viz Robot Programming Software | Mech-Mind Robotics | |
| SI006 | Mech-Mind Robotics | Robotic Bin Picking with 3D Vision | Mech-Mind Robotics | |
| SI007 | Mech-Mind Robotics | Logistics | Mech-Mind Robotics | |
| SI008 | Mech-Mind Robotics | Value Provider Cooperation | Mech-Mind Robotics | |
| SI009 | 投资界 | 梅卡曼德完成新一轮近5亿元融资 | 梅卡曼德日前完成近5亿元新一轮融资。 |
| SI011 | S&P Capital IQ / MarketScreener | Mech-Mind Robotics Technologies Ltd. announced that it has received CNY 500 million in funding from a group of investors | Mech-Mind Robotics Technologies Ltd. announced that it has received CNY 500 million in a round of funding on August 25, 2025. |
| SI012 | Yahoo Finance (Bloomberg syndication) | 美團投資的人工智能機器人公司梅卡曼德據悉計畫在香港上市 | 根據知情人士,美團投資的梅卡曼德機器人計畫在香港首次公開募股(IPO),擬籌資約2億美元。 |
| SI013 | The Standard | Meituan-backed AI robotics firm plans a HK IPO to raise US$200m | The report said the artificial intelligence robotics firm is talking with advisers and has filed confidentially for a share sale. |
| SI014 | 凤凰网科技 | 美团投资的AI机器人公司梅卡曼德将在港IPO 融资2亿美元 | 知情人士称,美团投资的AI机器人公司梅卡曼德计划在中国香港进行首次公开招股(IPO),融资大约2亿美元。 |
| SI015 | 香港商報 | 【新股最前線】梅卡曼德機器人據報秘密申港IPO 集資15.6億 | 梅卡曼德機器人已獲IDG、美團(3690)、紅杉中國等多輪投資,累計融資20億元人民幣。 |
| SI016 | 界面新闻 | 订单数每年超3倍增长,这家公司要做智能机器人基础设施供应商 | 邵天兰告诉界面新闻,“梅卡曼德机器人的订单数每年都有3倍以上的增长。” |
| SI017 | 搜狐 | 梅卡曼德再获近5亿融资,雄安、 海河基金等国资扎堆投 | 累计融资额近20亿元。 |
| SI018 | 新浪财经 | 梅卡曼德完成新一轮近5亿元融资 | |
| SI019 | 雪球 / 科创板日报 | 梅卡曼德完成近5亿元融资,3D视觉机器人激烈竞争下优势如何 | 中小客户因技术能力有限、成本敏感度高,导致公司收入仍偏依赖头部客户,长尾市场开拓效率待提升。 |
| SI020 | 蓝鲸新闻 / 网易号 | 梅卡曼德(雄安)机器人完成近5亿元融资,此前已获美团、红杉中国等机构多轮投资 | |
| SI021 | CB Insights | Mech-Mind Stock Price, Funding, Valuation, Revenue & Financial Statements | Mech-Mind's valuation in August 2022 was $858M. |
| SI022 | Craft | Mech-Mind Company Profile - Office Locations, Competitors, Revenue, Financials, Employees, Key People, Subsidiaries | Craft.co | Enterprise Customers 200 |
| SI023 | Info-clipper | Mech-Mind Robotics Technologies Ltd. China, Beijing | Info-clipper.com | Info-clipper.com brings you a complete range of reports and documents featuring legal and financial data, facts, analysis and official information from Chinese Registry. |
| SI024 | Tracxn | Mech-Mind | Mech-Mind has raised a total funding of$200M over 6 rounds. |
| SI025 | 东方财富网财富号 | 梅卡曼德 计划于2026年香港和纳斯达克双重上市 | 2024年收入预计突破8亿元,净利率超20%,订单排期直达2026年。 |
| SI026 | 东方财富网财富号 | 梅卡曼德近期在上市计划、融资进展、业务发展等方面均有重要动态,具体如下:上市计划 | 供应链集中度:核心零部件依赖进口可能影响成本与交付稳定性。 |
| SE001 | Mech-Mind Robotics | Embodied AI & 3D Vision for Robots and More | Mech-Mind Robotics | |
| SE002 | Mech-Mind Robotics | Mech-Eye Industrial 3D Cameras | Advanced 3D Sensor | |
| SE003 | Mech-Mind Robotics | Mech-Eye 3D Laser Profilers | High-Speed 3D Sensors | |
| SE004 | Mech-Mind Robotics | Vision Software | Mech-Vision Machine Vision Software | |
| SE005 | Mech-Mind Robotics | Mech-Viz Robot Programming Software | Mech-Mind Robotics | |
| SE006 | Mech-Mind Robotics | Vision Software | Mech-DLK Deep Learning Software | Mech-Mind Robotics | |
| SE007 | Mech-Mind Robotics | Embodied Intelligence "Eye-Brain-Hand" Robot Station | General-Purpose Embodied AI | |
| SE008 | Mech-Mind Robotics | MECH-MIND ROBOTICS IS NOW PART OF THE UR+ ECOSYSTEM | |
| SE009 | 梅卡曼德机器人 | Mech-Eye 工业级3D相机 - 梅卡曼德机器人 | |
| SE010 | 梅卡曼德机器人 | Mech-Vision 机器视觉软件 - 梅卡曼德机器人 | |
| SE011 | 梅卡曼德机器人 | Mech-Viz 机器人编程软件 - 梅卡曼德机器人 | |
| SE012 | 梅卡曼德机器人 | Mech-DLK 深度学习软件 - 梅卡曼德机器人 | |
| SE013 | Mech-Mind Documentation | Standard Interface Adaptation | |
| SE014 | Mech-Mind Documentation | Status Codes and Troubleshooting | |
| SE015 | Mech-Mind Documentation | KUKA | |
| SE016 | Mech-Mind Online Community | GitHub addresses of Mech-Eye SDK sample programs | |
| SE017 | GitHub | Mech-Mind · GitHub | |
| SE018 | GitHub / MechMindRobotics | GitHub - MechMindRobotics/mecheye_ros_interface: Official ROS interface for Mech-Eye cameras. | |
| SE019 | GitHub / MechMindRobotics | Releases · MechMindRobotics/mecheye_ros_interface | |
| SE020 | GitHub / MechMindRobotics | mecheye_ros2_interface/README.md at main · MechMindRobotics/mecheye_ros2_interface | |
| SE021 | Universal Robots | Mech-Mind 3D Vision | |
| SE022 | KR-Asia | This Chinese robotics firm is making factory AI modular, global, and scalable | |
| SE023 | ManualsLib | MECH MIND MECH-EYE 3D LASER PROFILER USER MANUAL Pdf Download | |
| SE024 | JM Vistec | Mech-Mind Catalogue_JMVS | |
| SE025 | CL Electronics | 3D Vision & AI for Robots and More Mech-Mind Robotics Product Catalog | |
| SE026 | CeraThai | Mech-Mind Robotics Product Catalog_CRT | |
| SE027 | 东方财富网 | 梅卡曼德携自研通用机器人“眼脑手”全栈技术产品亮相WAIC 2025 _ 东方财富网 | |
| SE028 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | Mech-Mind Robotics | |
| SE029 | Mech-Mind Robotics | News | Mech-Mind Robotics | |
| SU001 | Mech-Mind Robotics | Embodied AI & 3D Vision for Robots and More | Mech-Mind Robotics | |
| SU002 | Mech-Mind Robotics | About Us | Mech-Mind Robotics | With 24,000+ units deployed across nearly 50 countries and regions, Mech-Mind's "Eye + Brain" solutions power industries such as logistics, F&B, and automotive. |
| SU003 | Mech-Mind Robotics | Automotive | Mech-Mind Robotics | |
| SU004 | Mech-Mind Robotics | Logistics | Mech-Mind Robotics | |
| SU005 | Mech-Mind Robotics | Electronics | Mech-Mind Robotics | |
| SU006 | Mech-Mind Robotics | EV Battery | Mech-Mind Robotics | |
| SU007 | Mech-Mind Robotics | Metal & Machining | Mech-Mind Robotics | |
| SU008 | Mech-Mind Robotics | Automotive case page 2 | Mech-Mind Robotics | |
| SU009 | Mech-Mind Robotics | Automotive case page 3 | Mech-Mind Robotics | |
| SU010 | Mech-Mind Robotics | Automotive case page 4 | Mech-Mind Robotics | |
| SU011 | Mech-Mind Robotics | Automotive case page 5 | Mech-Mind Robotics | |
| SU012 | Mech-Mind Robotics | MECH-MIND ROBOTICS IS NOW PART OF THE UR+ ECOSYSTEM | Mech-Mind Robotics | |
| SU013 | Association for Advancing Automation | Mech-Mind Robotics-3D Vision and Automation Solutions | Member of A3 | |
| SU014 | Automate Show | Mech-Mind Robotics Technologies Co. Ltd. | |
| SU015 | KrASIA | This Chinese robotics firm is making factory AI modular, global, and scalable | To support localization, Mech-Mind partners with system integrators and agents in each region. This approach leverages local expertise, minimizes distribution conflicts, and enables faster scaling. |
| SU016 | 梅卡曼德机器人 | 梅卡曼德机器人- 通用智能机器人的AI大脑和3D视觉 | |
| SU017 | 梅卡曼德机器人 | 关于我们 - 梅卡曼德机器人 | 梅卡曼德自研的机器人AI大脑+3D视觉产品已经在汽车、物流、重工等众多领域跨行业、规模化落地,服务于全球100+《财富》500强客户,业务覆盖近50国家和地区。 |
| SU018 | 雪球 | 未来产业Top50之梅卡曼德(Mech-Mind) | |
| SU019 | HowToRobot | Mech-Mind | HowToRobot | |
| SU020 | Craft | Mech-Mind Company Profile - Office Locations, Competitors, Revenue, Financials, Employees, Key People, Subsidiaries | Craft.co | |
| SU021 | Mech-Mind Robotics Documentation | Typical Case Practices | |
| SU022 | World Robot Conference | 【世界机器人大会·应用案例】看宝马集团如何玩转协作自动化? | 每天,SortBot可分拣超过1000个货箱。 |
| SU023 | BMW Group | BMW Group: First humanoid robot introduced in Plant Leipzig | |
| SU024 | AMTS | SAIC快讯 | 上汽集团率先实现人形机器人量产线应用 | |
| SU025 | CnEVPost | SAIC-GM deploys wheeled humanoid robots on Buick battery assembly line - CnEVPost | SAIC-GM has deployed humanoid robots on Buick's battery production line. |
| SU026 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | Mech-Mind Robotics | |
| SU027 | Mech-Mind Robotics | Precise, Efficient, Fast to Deploy—Mech-Mind "Eye + Brain" Enables 100% Inline Inspection on Automotive Production Lines | Mech-Mind Robotics | |
| SU028 | Mech-Mind Robotics | Mech-Mind at iREX 2025 | Winning Industry-Wide Acclaim for Full-Stack Robot "Eye-Brain-Hand" Showcase and Global Product Premieres | Mech-Mind Robotics | |
| SR001 | Mech-Mind Robotics | About Us | Mech-Mind Robotics | |
| SR002 | 梅卡曼德机器人 | 关于我们 - 梅卡曼德机器人 | |
| SR003 | Mech-Mind Robotics | Why Leading Integrators and End-Customers Choose Mech-Mind for Reliable 3D Vision - MechMind | |
| SR004 | Mech-Mind Robotics | From Challenge to Reliability: Handling Reflective Parts with Mech-Mind “Eye + Brain” - MechMind | |
| SR005 | Mech-Mind Robotics | Seeing the Invisible: Why Transparent Objects Are So Hard for Robots - MechMind | |
| SR006 | Mech-Mind Robotics | From Factory Floors to Global Lighthouses: 3D Vision's Role in 2026 Industrial Automation Goals - MechMind | |
| SR007 | Mech-Mind Robotics | Value Provider Cooperation | Mech-Mind Robotics | |
| SR008 | Mech-Mind Robotics | Contact Us | Mech-Mind Robotics | |
| SR009 | 梅卡曼德机器人 | 联系我们 - 梅卡曼德机器人 | |
| SR010 | 量子位 | 达沃斯聚焦技术新前沿,梅卡曼德创始人邵天兰受邀分享具身智能落地实践 | |
| SR011 | Mech-Mind Robotics | Mech-Mind at Automate 2025 | Winning Industry Acclaim for Advanced AI + 3D Vision Robotics Technologies | |
| SR012 | Mech-Mind Robotics | Mech-Mind at AW 2026 | Advancing Embodied Intelligence "Eye-Brain-Hand" with New Products and Applications | |
| SR013 | Bureau of Industry and Security | Department of Commerce Revises License Review Policy for Semiconductors Exported to China | |
| SR014 | Finnegan | BIS’s New 2026 License Review Process for AI Chips | |
| SR015 | Hong Kong Exchanges and Clearing | Listing of Specialist Technology Companies | |
| SR016 | Han Kun Law Offices | Listing Regime for Specialist Technology Companies in Hong Kong | |
| SR017 | 香港商報 | 【新股最前線】梅卡曼德機器人據報秘密申港IPO 集資15.6億 | |
| SR018 | 腾讯新闻 | 梅卡曼德机器人,拟赴香港上市,传已秘密递表,计划募资2亿美元 | |
| SR019 | 香港经济日报 | 新股IPO|梅卡曼德機器人據報秘密申港IPO集資15.6億 從事AI機械視覺軟件 | |
| SR020 | AAStocks | <IPO>Meituan-Backed AI Firm Mech-Mind Files Confidentially for HK Listing: Wire | |
| SR021 | 爱企查 | 梅卡曼德(雄安)机器人科技股份有限公司 - 爱企查 | |
| SR022 | Securities and Exchange Commission | cgnx-20241231 | |
| SR023 | KEYENCE | 3D Vision-Guided Robotics - 3D VGR series | |
| SR024 | Qviro | Top 10 Robot Vision Manufacturers & Brands 2026 - Qviro Blog | |
| SR025 | Intel Market Research | Machine Vision Vision Guided Robotics Market Outlook 2026-2034 | |
| SR026 | CNBC | China's factory activity unexpectedly contracts in November, missing estimates, private survey shows | |
| SR027 | Commonwealth Bank of Australia | China’s factory slowdown continues | |
| SR028 | National Bureau of Asian Research | The Trajectory and Implications of China’s Economic Slowdown | |
| SR029 | Deloitte | 2026 Manufacturing Industry Outlook | |
| SR030 | KrASIA | This Chinese robotics firm is making factory AI modular, global, and scalable | |
| SV001 | S&P Capital IQ / MarketScreener | Mech-Mind Robotics Technologies Ltd. announced that it has received CNY 500 million in funding from a group of investors | |
| SV002 | Sina Finance | 梅卡曼德完成新一轮近5亿元融资 | 梅卡曼德日前完成近5亿元新一轮融资。 |
| SV003 | Pedaily / Zero2IPO | 梅卡曼德完成新一轮近5亿元融资 | 梅卡曼德日前完成近5亿元新一轮融资。 |
| SV004 | China High-Tech Industry Herald | 梅卡曼德完成新一轮近5亿元融资,加速具身智能“眼脑手”全栈技术进化与全球规模化落地 | |
| SV005 | Bloomberg | Meituan-Backed AI Robotics Firm Mech-Mind Is Said to Plan HK IPO | Mech-Mind plans in Hong Kong an initial public offering aiming to raise about $200 million, according to people familiar with the matter. |
| SV006 | The Standard | Meituan-backed AI robotics firm plans a HK IPO to raise US$200m | |
| SV007 | 36Kr English | Mech-Mind Robotics has completed a Series D financing round worth 500 million yuan. | |
| SV008 | China Daily | China's fast-growing firms make up one-third of global gazelle ranking | Five Chinese companies, including robot makers Mech Mind and AgiBot, have each surpassed a valuation of $1 billion and are set to join the Hurun Unicorn List in 2026. |
| SV009 | Xinhua | Innovation thrives in north China's "city of the future" | Mech-Mind Robotics, a Chinese unicorn company... exporting products to over 50 countries and regions. |
| SV010 | The State Council of the People's Republic of China | China nurtures unicorn enterprises via sci-tech innovation | Mech-Mind, a Chinese unicorn firm focusing on industrial 3D cameras and AI-powered software for intelligent robotics... |
| SV011 | Yahoo Finance Hong Kong / Bloomberg | 美團投資的人工智能機器人公司梅卡曼德據悉計畫在香港上市 | |
| SV012 | Hong Kong Exchanges and Clearing | Listing of Specialist Technology Companies | A new chapter (Chapter 18C) has been added to the Main Board Listing Rules to provide a new listing pathway for Specialist Technology Companies. |
| SV013 | CompaniesMarketCap | Symbotic (SYM) - Market capitalization | |
| SV014 | CompaniesMarketCap | Cognex (CGNX) - Market capitalization | |
| SV015 | CompaniesMarketCap | Zebra Technologies (ZBRA) - Market capitalization | |
| SV016 | CompaniesMarketCap | ABB (ABBN.SW) - Market capitalization | |
| SV017 | CompaniesMarketCap | Teradyne (TER) - Market capitalization | |
| SV018 | Simply Wall St | Rivian’s Mind Robotics Unicorn Raises New Questions On Future Valuation | Watch for disclosures on Mind Robotics ownership, revenue sharing, and any future funding or IPO discussions that could crystallize value for Rivian. |
| SV019 | TechCrunch | Rivian spin-out Mind Robotics raises $500M for industrial AI-powered robots | |
| SV020 | HKEX Group | 18C, Explained | Commercial Companies are those that have met the prescribed commercialisation revenue threshold... Pre-Commercial Companies haven’t yet met this threshold... |
| SV021 | Teradyne | Teradyne, Inc. Form 10-K for fiscal year ended December 31, 2024 | Our Robotics segment is comprised of two business units: Universal Robots and Mobile Industrial Robots. |
| SV022 | Stocklight / Symbotic | Symbotic Annual Report 2025 (Form 10-K) | |
| SV023 | Stocklight / Cognex | Cognex Corporation Annual Report 2025 (Form 10-K) | |
| SV024 | Stocklight / Zebra Technologies | Zebra Technologies Corporation Annual Report 2025 (Form 10-K) | |
| SV025 | TradingView / GuruFocus | China's AI Robot Darling Eyes $200M Hong Kong IPO as Investor Frenzy Heats Up | |
| SV026 | Hong Kong Commercial Daily | 【新股最前線】梅卡曼德機器人據報秘密申港IPO 集資15.6億 | |
| SV027 | Tencent News / Ryanben Capital | 梅卡曼德机器人,拟赴香港上市,传已秘密递表,计划募资2亿美元 | |
| SV028 | The Business Research Company | Global Machine Vision Market Report 2026 | |
| SV029 | NetEase / Blue Whale News | 梅卡曼德(雄安)机器人完成近5亿元融资,此前已获美团、红杉中国等机构多轮投资 | |
| SV030 | Sohu / 独角兽早知道 | 传梅卡曼德机器人已秘密提交香港上市申请,预计募资2亿美元,IDG资本、美团、红杉中国等参投 |