Humanoid
欧洲首个纯人形机器人独角兽,完成 $152M Series A 轮后投后估值达 $1.35B
Humanoid 执行速度快,是欧洲人形机器人头部公司,手握顶级合作伙伴和大额预订订单;但 $1.35B 的尚未形成收入估值,仍押在尚未验证的商业转化上。
封面要素
公司概况
Humanoid 是一家总部位于 London 的工业人形机器人公司,Artem Sokolov 于 2024 年创立。公司设计并制造 HMND 01 Alpha 系列人形机器人(一个续航 4 小时的轮式版本和一个双足版本,均有 29 个自由度、15 kg 双臂载荷),并自研名为 KinetIQ 的视觉-语言-动作 AI 大脑。短短两年内,公司以 $1.35 billion 投后估值完成 $152 million Series A 轮,成为欧洲首个纯人形机器人独角兽;同时与 Schaeffler、Bosch、Siemens、SAP 和 NVIDIA 建立 Fortune 500 合作关系,并披露 34,000 台、约 ~$2.4 billion 的预订单簿。
- 成立时间
- 2024-05-01
- 创始人
- Artem Sokolov
- 创立地点
- London, United Kingdom
- 总部
- London, United Kingdom (engineering offices in Boston and Vancouver)
- 产品
- HMND 01 Alpha 工业人形机器人包括轮式版本(约 ~220 cm,最长 4 小时续航)和双足版本(约 ~179 cm),具备 29 个自由度、15 kg 双臂载荷、NVIDIA Jetson/Intel 计算栈和自研 KinetIQ AI 大脑,面向制造和物流工作流销售。
- 客户
- 制造、物流 / 仓储和零售领域的工业买方,以 Schaeffler、Bosch 等汽车和工业 OEM 为锚点。
- 商业模式
- 向工业客户销售人形机器人,并以机器人即服务方式部署;公司由垂直整合的硬件 + AI 栈和 Bosch 制造合作支撑,收入仍处于商业化前阶段。
- 阶段
- Series A
- 融资情况
- 以 $1.35B 投后估值完成 $152M Series A 轮(2026-07-21 宣布),由 Prime Movers Lab 领投,Schaeffler、Bosch、Fubon Financial Holding Venture Capital 和 Aglaé Ventures 参投;截至目前累计融资 $270M。
执行摘要
主要优势
- 欧洲首家纯人形机器人独角兽,执行速度突出:从创意到独角兽约两年,5-7 个月做出原型。
- 蓝筹工业客户给出验证:Schaeffler(公开披露最大的人形机器人商业协议)、Bosch(订单加制造伙伴)、Siemens、SAP 和 NVIDIA。
- 硬件加 AI 垂直整合,核心是 HMND 01 Alpha 和 KinetIQ;先做轮式方案,更贴近短期工业落地。
- 需求信号大:据报已有 34,000 台、约 $2.4 billion 的预订订单储备;$152M Series A 也让资金弹药充足。
主要风险
- $1.35B 估值对应的确认收入几乎为零;投资逻辑取决于把非约束性预订转成实际交付。
- 美国和中国对手资金更厚:Figure 约 $39B、Apptronik 约 $5B、Tesla Optimus、低成本 Unitree 都可能用资本或价格压住 Humanoid。
- 技术和安全交付仍有风险:自主能力成熟度、远程操控依赖、规模化可靠性、人形机器人安全认证都还要跑通。
- 客户集中在 Schaeffler / Bosch 一带,也依赖创始人 Artem Sokolov;地缘政治和出口管制观感会放大风险。
未决问题
- 34,000 台预订的合同状态仍不清楚:是否具约束力、是否只是 LOI、是否有定金和取消条款;20,000 到 34,000 台的口径差异也未核清。
- 收入、毛利率、月烧钱速度、现金跑道,以及股权结构表 / 清算优先权条款均未披露。
- Bosch 与 Schaeffler 的订单台数归属,以及 Series A 前几轮融资的准确规模(隐含约 $118M)仍不清楚。
- HMND 01 的部署可靠性、自主运行与远程操控占比、安全认证状态,都缺少独立验证。
目录
01公司概况
1.1 身份、产品与布局
Humanoid 是一家 2024 年成立的英国 AI 与机器人公司,面向工业工作销售 HMND 品牌人形劳动力,而不是把自己定位成实验室演示项目。这样一家年轻机器人创业公司,身份已经少见地具体:自有官网、产品页和融资报道都指向工业人形机器人、HMND 01 Alpha 系列,以及支撑机器人集群编排的自研 AI 框架 KinetIQ。公司的运营足迹也足够清楚,可以作为后续报告的事实底座。总部在 London,Boston 和 Vancouver 则被描述为工程卫星团队或办公室。产品叙事有两个近期形态:面向工业部署的 Alpha Wheeled,以及用于双足研发和未来应用的 Alpha Bipedal。公开参数支持 15 kg 载荷、29 个自由度和工业级移动速度,但一个产品规格细节仍有冲突:官方双足页面重复了部分共享 Alpha 指标,独立摘要则采用双足版本 179 cm、90 kg、1.5 m/s 和三小时续航的标准值。[CO001, CO002, CO003, CO004, CO005, CO006]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 注意事项 |
|---|---|---|---|---|
| 公司身份 | Humanoid / HMND;工业人形机器人 | 2026-07-21 | 高 | 品牌和产品表述已公开,但法律实体结构需要更深入审查 |
| 成立时间 | 2024;一份资料写明为 2024 年 5 月 | 2024 | 高 | 未审阅准确的公司注册文件 |
| 总部 / 办公室 | 伦敦总部;工程办公室位于 Boston 和 Vancouver | 2026-07-21 | 高 | Silicon Republic 另提到 San Diego;本报告采用的标准地点为 London / Boston / Vancouver |
| 最新轮次 | $152M Series A 轮 | 2026-07-21 | 高 | 除融资额外,头条披露的轮次条款仍未公开 |
| 投后估值 | $1.35B | 2026-07-21 | 高 | 未披露优先权结构或老股转让细节 |
| 累计融资 | 迄今累计融资 $270M | 2026-07-21 | 高 | 暗示此前资本约 $118M |
| 员工人数 | 封面采用约 160–200 人区间 | 2026-07-21 | 中 | 检索到的公开来源从 130 多人到 250 多人不等 |
| 客户 / 预订单 | 标准基线为 34,000 份预订单,价值约 $2.4B | 2026-07-21 | 中 | 本章检索到的 CNET 报道称约 25,000 份预订单 |
| 收入 / ARR | 未公开披露 | 2026-07-21 | 中 | 需要管理层数据室或客户发票 |
| 地点 | London, Boston, Vancouver | 2026-07-21 | 高 | 未审阅逐实体子公司图谱 |
快照结合了经交叉印证的公开事实和本报告标准输入,并标出检索到的公开证据存在冲突或缺失之处。
[CO001, CO003, CO004, CO013, CO015, CO016]概览投资逻辑把身份、产品、资本、合作伙伴和未解决依赖项串联起来。
流程是定性尽调地图,不是所有权或流程图。
[CO001, CO003, CO005, CO008, CO009, CO011]封面 KPI 在融资和估值上强,但私人商业转化数据偏弱。
由于公开来源数值不一,预订单和员工数 KPI 特意标注限制说明。
[CO003, CO013, CO015, CO016, CO017, CO018]1.2 领导层、团队与关键人风险
创始人叙事既是真资产,也是尽调风险。公开资料一致将 Artem Sokolov 列为创始人兼 CEO,并把他在 Humanoid 之前的经历与制造、投资和家族企业背景相连。机器人公司很吃制造纪律,这让 Humanoid 的战略和规模化故事更可信。但叙事风险也集中到一个人身上:Forbes 文章把 Sokolov 描述为前俄罗斯珠宝大亨,使创始人背景、地缘政治、制裁敞口和投资人舒适度都成为明确尽调议题,而不是脚注。Jarad Cannon 作为 CTO 提供了重要技术平衡,第三方报道指向他拥有超过 13 年机器人和 AI 执行经验。更广泛的团队厚度披露不够精确。公开来源从超过 130 名专业人员到超过 200 名工程师、250 多名员工不等,因此本章把标准封面区间设为约 160–200,同时标记该差异以便跟进。[CO009, CO010, CO011, CO012, CO013, CO014]
| 人物 / 群体 | 角色 | 背景 / 报道 | 创始人市场匹配或职能覆盖 | 关键人依赖 |
|---|---|---|---|---|
| Artem Sokolov | 创始人兼 CEO | 制造、投资和家族企业背景;公开资料显示他在 2024 年创立 Humanoid | 工业机器人商业化具备较强的战略 / 制造叙事 | 高:公司故事、融资和创始人负面尽调高度围绕创始人展开 |
| Jarad Cannon | 首席技术官 | 第三方报道称其于 2025 年 5 月出任 CTO,在机器人和 AI 领域有 13 年以上经验 | 为创始人主导的战略补上机器人执行和可规模化产品领导力 | 中:公开证据点名 CTO,但未披露完整工程领导梯队 |
| 工程 / 研究团队 | 技术梯队 | 公司 / 媒体来源称有 130 多至 200 多名工程师、研究人员和专业人员 | 支撑 London、Boston 和 Vancouver 三地快速原型与部署叙事 | 中:准确员工数和具名校友名单需要私下确认 |
列表覆盖公开具名领导层和已披露团队梯队;由于董事会、CFO / COO 和完整技术领导名单未公开,本表并不完整。
[CO009, CO010, CO011, CO012, CO013, CO014]1.3 融资、估值与投资人图谱
2026 年 7 月 21 日的 Series A 轮是公司概况中最关键的事件。多家独立媒体报道了相同的核心经济数字:$152 million Series A 轮、$1.35 billion 投后估值,以及截至目前 $270 million 累计融资。由此可推算,Series A 轮之前约有 $118 million 既有资本。Prime Movers Lab 领投;具名共同投资人重要,因为其中几家不只是财务标识。Schaeffler 和 Bosch 带来工业验证和潜在制造杠杆;Fubon Financial Holding Venture Capital 增加亚洲金融资本触达;Aglaé Ventures 通过 Silicon Republic 报道的 Arnault 相关载体提供奢侈品家族办公室信号。因此,估值标签不只是价格。它建立在一个主张上:Humanoid 已经足够快地把资本与战略部署、制造关系配对,成为欧洲首个纯人形机器人独角兽。待解问题在经济层面:已审阅的公开来源均未给出董事席位、清算优先权、债务、老股交易、创始人稀释或现金跑道。[CO015, CO016, CO017, CO018, CO019, CO020]
| 利益相关方 | 角色 | 控制权 / 经济重要性 | 尽调要求 |
|---|---|---|---|
| Prime Movers Lab | Series A 轮领投方 | 领投 $152M 轮次,并锚定科学突破型投资人信号 | 确认董事席位、按比例跟投权和投资载体经济安排 |
| Schaeffler | 投资人、客户 / 部署伙伴、执行器供应商 | 四位数机器人部署协议和优先执行器供应商角色,使其成为验证基石 | 审阅有约束力的订单条款、部署里程碑、验收标准和取消权 |
| Bosch | 投资人和制造伙伴 | 据报合同制造角色叠加硬件 / 供应链专长,可能降低规模化风险 | 将制造服务合同与任何客户订单或台数主张分开核验 |
| Fubon Financial Holding Venture Capital 创投 | Series A 轮共同投资人 | 带来亚洲金融资本和潜在生态触达 | 确认战略权利、区域排他性和后续跟投意愿 |
| Aglaé Ventures | Series A 轮共同投资人 | Silicon Republic 将 Aglaé 与 Bernard Arnault / LVMH 的资本信号相连 | 确认参与只是财务投资,还是面向零售 / 服务机器人的战略投资 |
| Artem Sokolov / 管理层 | 创始人控制中心 | 创始人主导战略和外部叙事,是融资与招聘的核心 | 执行背景、制裁、关联方和关键人连续性核查 |
| SAP、Siemens、NVIDIA 与 Martur Fompak | 工业概念验证生态 | POC 和工厂集成让商业可信度不只停留在投资人说法上 | 验证真实环境结果、安全签核和 POC 后扩展承诺 |
本图谱按角色而非股权比例映射具名利益相关方;没有公开文件披露控制权或持股比例。
[CO019, CO020, CO021, CO022, CO024, CO025]1.4 商业证明与里程碑时间线
里程碑记录显示,公司把成立、产品、合作和独角兽融资压缩进大约两年。可公开追溯事件从 2024 年成立开始,经过 2025 年 Jarad Cannon 出任 CTO,随后在 2026 年随 SAP/Martur Fompak、Siemens/NVIDIA、Schaeffler、Bosch 和 Series A 里程碑加速。Schaeffler 是公开文件最强的商业证明,因为官方公告描述了一项具有约束力的分阶段部署和供应协议,涉及四位数规模的轮式机器人,并从 2026 年底开始在德国生产现场部署。Bosch 在战略上也重要,但具体订单 / 部署数量不够干净:融资报道支持 Bosch 作为合同制造伙伴,而一个二级来源把 1,000 台机器人订单和 100,000 台产能目标归因于 Schaeffler/Bosch 生态。本章把 Schaeffler 视为已披露的最大部署协议,并把 Bosch 订单数量保留为缺口,而不是已定事实。[CO021, CO022, CO023, CO024, CO025, CO026]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2024 | Artem Sokolov 在 London 创立 Humanoid | 创立 | 2024 年创立 | Artem Sokolov | 为一家非常年轻的机器人公司建立身份基线 |
| 2025-05 | Jarad Cannon 出任 CTO | 治理 | CTO 任命 | Jarad Cannon; Humanoid | 为创始人主导的业务补上技术执行领导力 |
| 2025 | HMND 01 Alpha 原型推进 | 产品 | Alpha 原型 / 快速开发 | Humanoid 工程团队 | 支撑速度叙事,但仍早于规模化客户交付 |
| 2026-02-05 | Forbes 报道 $200M 融资洽谈和俄罗斯创始人叙事 | 负面 | Series A 前融资报道 | Forbes; Artem Sokolov | 让创始人背景和估值风险尽调变得明确 |
| 2026-03-30 | SAP 和 Martur Fompak 宣布汽车物流 POC | 合作 | 真实生产物流概念验证 | Humanoid; SAP; Martur Fompak | 显示早期工业工作流验证 |
| 2026-04-16 | Siemens / NVIDIA 物理 AI 工厂部署公布 | 合作 | 工厂试点;7 个月原型主张 | Humanoid; Siemens; NVIDIA | 将 HMND Alpha 接到仿真和工厂集成栈上 |
| 2026-05-13 | Schaeffler 分阶段部署和供应协议公布 | 合作 | 四位数机器人部署;2026 年底德国投产 | Humanoid; Schaeffler | 最强的公开商业部署证据 |
| 2026-06-25 | Bosch 合作出现在官方新闻室和融资报道中 | 合作 | 据报提供合同制造 / 技术支持 | Humanoid; Bosch | 潜在制造扩产杠杆;准确订单数量仍未解决 |
| 2026-07-21 | Series A 轮公布 | 融资 | $152M Series A 轮;$1.35B 投后;累计融资 $270M | Prime Movers Lab、Schaeffler、Bosch、Fubon 与 Aglaé | 确认欧洲专注人形机器人独角兽里程碑 |
| 2026-Q4 | 计划推出 Beta 机器人 | 规模化 | 计划将 Beta 版本机器人部署到客户工业现场 | Humanoid;工业客户 | 将投资论点从公告转向交付执行 |
| 2031/2032 目标 | 来源提及大规模生产 / 部署目标 | 规模化 | 来源中的 100,000 台目标和 Schaeffler 四位数部署周期 | Bosch; Schaeffler; Humanoid | 凸显上行空间和长期执行风险 |
时间线按公告日期编排。它保留 Bosch / Schaeffler 台数差异的未解状态,而不是强行归一。
[CO004, CO011, CO021, CO022, CO023, CO025]Humanoid 在两年的公开时间线里压缩完成了创立、高管招聘、产品验证、合作伙伴和独角兽融资。
时间线采用公开精确事件的公告日期;创立月份和部分产品里程碑经过取整。
[CO004, CO011, CO010, CO025, CO026, CO021]1.5 封面 KPI、缺口与反向框架
封面事实应该有力度,但必须谨慎加限定。估值为 $1.35 billion,累计融资为 $270 million,最新融资是 $152 million Series A 轮。总部和工程布局事实支撑较好。员工数不是:报告采用的标准区间约为 160–200,但抓取到的公开来源从超过 130 到超过 250 不等,因此任何精确员工数都应视为需要私有证据。客户牵引更敏感。报告基准使用 34,000 个预订单、约 $2.4 billion;CNET 则报道约 25,000 个预订单和 6 个 Fortune 500 试点。该差异应带入客户和估值章节。收入、ARR、毛利率、烧钱速度、现金跑道和股权结构经济条款仍未披露。因此,反向框架不是公司缺少动能,而是一家非常年轻、创始人主导、资本密集的机器人公司,在公开收入、交付质量或订单转化证据出现之前,就已经按独角兽估值定价。[CO012, CO013, CO015, CO016, CO017, CO031]
1.6 图表
02市场分析
2.1 边界:工业人形机器人,不是所有机器人
Humanoid 的可投资市场应划在面向人为设计工作流的、具有人形形态的移动机器人上,即部署进工业和商业流程的机器人。该范围包括用于制造物流、线边物料移动、仓库周转箱或包裹处理、巡检,并最终用于零售或服务任务的双足或轮式人形机器人;在人形外形能让机器人使用既有空间和工具时,这些任务都在范围内。该范围不包括更广泛的工业机器人装机基础、围栏单元内的传统机械臂、AMR、传送带,以及专用固定自动化,除非把它们作为 ROI 对比中的替代方案。这个边界重要,因为广义机器人统计会让市场看起来成熟,而人形机器人仍早期且商业上未被证明。Humanoid 自身定位和 Schaeffler 关系都指向更窄的工业劳动力市场,而不是家庭消费机器人市场。实用检验标准是:机器人能否接下原本需要人走过工厂或仓库完成的工作,而不是它是否属于宽泛得多的机器人类别。因此,边界让 AMR 和固定单元作为竞争基准可见,但不进入人形机器人 TAM 的分子。[CM001, CM002, CM003, CM004, CM005, CM006]
| 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 与 Humanoid 的相关性 |
|---|---|---|---|---|
| 工业人形机器人 | 面向工厂和仓库里人形机器人的机器人硬件、软件、服务、部署和车队支持 | 传统机械臂、输送机、AMR、无人机和消费机器人 | 工厂运营、工业工程、自动化、安全、IT | Alpha Wheeled 和 Schaeffler 式部署的主要边界 |
| 物流 / 仓储人形机器人 | 周转箱搬运、货物流转、补货、配套、退货和设施任务 | 现有 AMR、AS/RS、分拣、叉车和临时人工 | 3PL 运营、仓库自动化、用工规划 | 近端 SAM,有真实 GXO / Digit 试点证据 |
| 制造业人形机器人 | 线边物料移动、机床上下料、检测、装配辅助和厂内物流 | 围栏式工业机器人单元和定制固定自动化 | 工厂经理、制造工程、工业工程 | 通过 Schaeffler 和更广市场报告拿到最强伙伴证据 |
| 零售 / 商业服务人形机器人 | 后场搬运、货架支持、设施支持,以及有证据支撑的面客服务 | 点状机器人、亭机、货架扫描器和传统服务机器人 | 门店运营、零售创新、设施 | 纳入广义 TAM,但作为 Humanoid 近端 SAM 的置信度较低 |
| 消费 / 家用人形机器人 | 不纳入工业 SAM;只与更远期 Morgan Stanley 生态 TAM 相关 | 家用助手、仅面向养老的家用设备、娱乐机器人 | 家庭和消费渠道 | 对 Humanoid 当前工业估值案例而言过于投机 |
边界受证据约束:纳入支出必须是工业 / 商业工作流中的人形移动机器人;排除类别仍是 ROI 测试中的替代项。
[CM001, CM002, CM003, CM004, CM005, CM021]宽口径生态 TAM 必须收窄到人形机器人 TAM、工业 / 商业 SAM,以及 Humanoid 可转化 SOM。
只有最上两层有来源支持的数值;SAM / SOM 有意留空,因为公开来源没有单独拆出。
[CM013, CM008, CM017, CM021, CM022, CM041]2.2 规模测算:区间很宽,口径很多
这个市场很大,但数字尚未收敛。2026 年当前市场口径在 MarketsandMarkets、Fortune、Axis 和 Coherent 数据中集中于 USD 5.4-6.4 billion 左右,但 Precedence 明显更低,Global Market Insights 明显更高。2035 年终点分歧更大:Precedence 只有 USD 8.78 billion,Goldman Sachs 给出 USD 38 billion,MarketsandMarkets 达到 USD 50.27 billion,Global Market Insights 达到 USD 192.7 billion。Morgan Stanley 到 2050 年 USD 5 trillion 的测算,更像生态系统和长周期采用情景,不应作为近期产品市场规模输入。Humanoid 需要先把 SAM 收窄到工业和商业工作流,再套用份额假设;SOM 应绑定经过验证的伙伴部署、制造产能、正常运行时间和订单转化,而不是整个人形机器人梦想。这种矛盾不是写作错误,而是核心尽调点。高增长预测假设成本快速下降、供应商产能跟上、安全性被接受,以及客户愿意把人形机器人集群标准化;较低预测则意味着工作流认证更慢,且更便宜的任务专用自动化会替代更多需求。[CM008, CM009, CM010, CM011, CM012, CM013]
| 发布方 / 视角 | 年份或周期 | 地区 / 范围 | 价值 | CAGR / 台数 | 方法提示 | 置信度 | 局限 |
|---|---|---|---|---|---|---|---|
| MarketsandMarkets | 2026 至 2035 | 全球人形机器人市场 | USD 5.41B 至 USD 50.27B | 28.1% CAGR | 按类型、产品 / 服务、应用和地区分拆 | 中 | 广泛应用包括医疗、教育、酒店和零售,超出 Humanoid 工业 SAM |
| Fortune Business Insights | 2026 至 2034 | 全球人形机器人市场 | USD 6.24B 至 USD 165.13B | 50.60% CAGR | 采用硬件 / 软件生态框架的市场研究预测 | 中 | 高增长曲线依赖快速采购和零部件成本改善 |
| Precedence Research | 2026 至 2035 | 全球人形机器人市场 | USD 2.16B 至 USD 8.78B | 16.91% CAGR | 市场研究预测 | 中 | 与高增长 TAM 情景相矛盾,且可能低估 2026 年新工业部署 |
| Mordor Intelligence | 2026 至 2031 | 人形机器人市场 | USD 3.93B 至 USD 17.80B | 35.26% CAGR | 预测与老龄化、劳动力缺口、政策以及制造 / 物流采用挂钩 | 中 | 只覆盖到 2031 年,不能直接与 2035 年终点比较 |
| Goldman Sachs | 2035 | 全球人形机器人市场 | USD 38B | 1.4M 台出货 | Goldman 和 RobotCentral 引用的投行预测 | 高 | 仍是广义人形机器人视角,不是 Humanoid 特定收入 |
| Morgan Stanley | 2050 | 全球人形机器人生态 | USD 5T | 近 1B 台机器人;约 90% 为工业 / 商业 | 长期生态采用情景 | 中 | 对风险投资测算而言周期太长,除非加入分阶段转化假设 |
| Morgan Stanley China(经 ChinaBizInsider) | 2026 和 2030 | 中国人形机器人 | 2026 年 USD 2B;2030 年 USD 15B | 2026 年 50k 台;2030 年 446k 台 | 基于据报 Morgan Stanley 上调后的地区预测 | 中 | 二手文章;检索到的证据中,原始银行报告未公开 |
| Global Market Insights 数据 | 2026 至 2035 | 全球人形机器人市场 | USD 10.9B 至 USD 192.7B | 37.6% CAGR | 市场研究预测 | 中 | 2026 和 2035 年高端数值拉大了矛盾区间 |
| 工业 / 商业 SAM 视角 | 2026 至 2035 | 制造、物流、仓储、部分零售 | 广义 TAM 的子集 | 未被独立拆分 | 由重复应用证据和试点证明推导 | 低 | 没有公开来源能清晰拆分工业人形机器人 SAM |
| Humanoid SOM 口径 | 2026 至 2031 | 已验证的合作伙伴部署和转化订单 | 私有 / 不可得 | 取决于 beta 正常运行时间、安全、供应链和转化 | 证据受限的投资人模型 | 低 | 需要客户合同、交付排期、毛利率和流失证据 |
这些数值按相互冲突的估计保留;图中区间统一使用 USD 十亿美元,表格行说明宽口径 TAM 为什么不等于 Humanoid SAM/SOM。
[CM008, CM009, CM010, CM011, CM012, CM013]2035 年估计值相差超过 20 倍,基准情景估值应使用区间,而不是单一 TAM。
所有行均以十亿美元计量 2035 年全球人形机器人市场估计;Goldman 与 MarketsandMarkets 合并作为实务基准带。
[CM010, CM012, CM008, CM017, CM044]2.3 买方图谱与采用路径
最高置信度的早期买方,是那些在围绕人设计的空间里承担重复、人体工学负担高工作的制造和物流运营方。制造买方可能包括工厂运营、制造工程、工业工程、安全和 IT/安全负责人;物流买方则是 3PL 运营、仓库自动化和劳动力规划团队。零售出现在广义市场报告中,但 Humanoid 近期工业产品的证据基础更薄。采用路径应建模为带闸门的企业销售:识别受约束的工作流,跑概念验证,验证安全和 IT 集成,测量可用性和可维护性,然后才扩展到机器人集群。这条路径意味着收入不均匀、采购周期长;但如果试点转化为多站点部署,也会带来很高战略价值。预算归属也会改变证明标准。工厂经理可能关心正常运行时间和安全签核,物流运营方可能关心旺季每次搬运成本,IT/安全团队则可能否决不满足数据、网络和服务访问控制的部署。[CM006, CM007, CM023, CM024, CM025, CM026]
| 细分市场 | 经济买方 | 主要用户 | 付款方 / 预算负责人 | 工作流 | 采用触发因素 |
|---|---|---|---|---|---|
| 汽车和离散制造 | 制造副总裁、工厂经理、工业工程负责人 | 产线操作员、物料搬运员、维护团队 | 制造运营资本开支 / 运营开支,需安全和 IT 签字 | 线边物料搬运、看机、检测、套料 | 用工缺口、人体工学风险、生产回流、复用以人为中心的产线需求 |
| 合同物流和仓储 | COO、首席自动化官、现场总经理 | 仓库员工和主管 | 运营自动化预算、用工规划、3PL 客户账户经济性 | 周转箱搬运、货到人支持、退货、补货 | 高峰用工压力、安全、吞吐量、可度量的单次拣选 / 搬运成本 |
| 工业设施和维护 | 设施、EHS、运营卓越 | 技术员和维护计划员 | 设施资本开支 / 运营开支、可靠性工程 | 巡检、简单取送、危险或脏累任务 | 降低安全暴露、提升正常运行时间 |
| 零售和商业运营 | 门店运营、零售创新、设施 | 门店员工和后仓人员 | 门店运营和创新预算 | 后仓搬运、货架支持、客户服务、设施 | 劳动力可得性和服务一致性,但证据仍较弱 |
| 消费者家庭 / 照护 | 家庭或照护服务方 | 消费者和照护者 | 消费可选支出或医疗 / 照护预算 | 家务辅助和照护 | 排除在近期工业 SAM 之外;仅用于远期 TAM 敏感性 |
买方角色根据引用的工作流证据推断;在给销售周期长度或赢单概率赋值前,应在客户采购访谈中验证。
[CM006, CM007, CM023, CM024, CM025, CM026]制造和物流拥有最强的近期人形机器人采购证据。
行综合引用的客户和市场证据;预算所有者标签需要直接采购验证。
[CM023, CM024, CM025, CM006, CM007, CM039]Humanoid 的采用应建模为带闸门的企业部署漏斗。
漏斗数值是示意性指数值,不是市场份额;它用于可视化引用的采用证据所支撑的采购收窄过程。
[CM026, CM027, CM028, CM029, CM040, CM042]2.4 影响采用速度的驱动因素与约束
核心需求驱动因素包括劳动力稀缺、人口老龄化、制造回流或产业政策压力、视觉-语言-动作模型进步,以及组件成本下降。BLS 的物料搬运和生产劳动力池足够大,可以创造有意义的自动化机会;IFR 和 WEF 也把自动化与劳动力转型列为宏观主题。约束同样重要。买方必须看到相对人工、AMR 和固定自动化的回本;安全认证必须支持无围栏运行;电池、执行器、传感器、计算和可维护性必须改善;资本强度仍然很高。反向解读不是人形机器人没有市场,而是 TAM 估算假设可靠性、正常运行时间、成本和安全性会改善,但这些假设尚未在规模上被证明。估值工作之前,应把这些约束转成尽调里程碑:Beta 续航相对班次长度、实测任务成功率、工人周边无事故小时、现场服务响应时间、组件交期,以及扣除保修和支持后的毛利率。没有这些里程碑,市场可以继续显得很可观,但收入转化仍可能令人失望。[CM027, CM028, CM029, CM030, CM031, CM032]
| 因素 | 方向 | 时间 | 含义 | 尽调要求 |
|---|---|---|---|---|
| 劳动力短缺和替换性岗位空缺 | 驱动因素 | 当前至中期 | 大量物料搬运和生产岗位让自动化 ROI 有成立空间 | 量化目标客户的用工成本、人员流失率、工伤和空缺基线 |
| 劳动力老龄化和人口压力 | 驱动因素 | 中长期 | 支撑 G7 和中国对自动化的持续需求 | 区分照护行业人口因素和工业用工短缺 |
| AI / VLA 和仿真进展 | 驱动因素 | 当前至中期 | 提升任务泛化能力,降低编程负担 | 要求现场数据,证明演示之外的任务成功率 |
| 执行器、电池、传感器和算力成本下降 | 驱动因素 | 中期 | 可压缩回本周期,支撑舰队规模化 | 审核 BOM、供应商集中度、质保假设和毛利率 |
| 生产回流和产业政策压力 | 驱动因素 | 中期 | 提高本地工厂柔性自动化价值 | 梳理受影响客户细分和资本开支预算 |
| ROI 和回本证明 | 约束 | 当下 | 除非生产率超过人工、AMR 或固定自动化,采购会卡住 | 要求站点级 ROI 模型和试点后实测吞吐量 |
| 安全认证和无围栏运行 | 约束 | 当下至中期 | 限制人形机器人在工人周边作业的场景 | 审查风险评估、ISO/OSHA 合规计划、事故历史和保险 |
| 可靠性、正常运行时间和可维护性 | 约束 | 当下 | 正常运行时间低,会把 TAM 变成失败试点和高昂支持成本 | 要求 beta 正常运行时间、平均故障间隔、服务人员配置和备件数据 |
| 电池续航和载荷限制 | 约束 | 当下 | 在热插拔或充电运营成熟前,压缩可行班次和工作流 | 将续航与班次长度、节拍要求对比 |
| 资本强度和供应链爬坡 | 约束 | 中期 | 即使需求充足,市场也可能卡在产能和部件供应上 | 追踪执行器、减速器、电池、算力和组装供应商及交期 |
这些驱动因素和约束属于市场层面;尽调要求把它们转成承销 Humanoid 特定采用率和估值所需的证据。
[CM027, CM028, CM029, CM030, CM031, CM032]2.5 图表
03竞争格局
3.1 格局与资本规模
工业人形机器人的竞争格局已不再是一组实验室项目,而是一场资本密集竞赛,参与者包括美国领先公司、中国相关硬件规模玩家,以及规模较小但越来越可见的欧洲集群。Figure 是估值离群点,Series C 轮投后估值达 $39 billion,并有 BMW 工厂证明;Apptronik 则把最初公布的 Series A 轮扩展为超过 $935 million 的融资和约 $5 billion 估值。Agility 看起来没那么膨胀,但通过 GXO、Amazon、Toyota 以及超过 100,000 个周转箱搬运,拥有最清楚的现场物流部署证明。在这个背景下,Humanoid 的 $152 million Series A 轮、$270 million 累计融资和 $1.35 billion 估值,让它在欧洲具备可信度,但相对美国领导者仍偏小。战略问题在于,Schaeffler、Bosch 和轮式优先设计能否转化为比 Figure 和 Apptronik 大额融资更快的工业学习。这一顺序也把尽调重点放在部署速度,而不是新闻稿数量。[CP001, CP003, CP004, CP008, CP010, CP014]
| 竞争者 / 替代方案 | 类别 | 2026 年规模或融资信号 | 形态 | 主要牵引信号 | 差异化 | 限制 / 风险 |
|---|---|---|---|---|---|---|
| Humanoid / HMND | 欧洲工业人形机器人纯玩家 | $270M 累计融资;投后估值 $1.35B | 轮式优先,叠加双足路线图 | Schaeffler 数千台协议;简报提及 Bosch 制造关系 | 欧洲优先的工业锚定客户和快速执行 | 资本规模低于 Figure 和 Apptronik;公开合同经济性不透明 |
| Figure AI | 美国通用人形机器人龙头 | C 轮融资 >$1B;投后估值 $39B | 双足 Figure 03 | BMW 生产和物流排序工作流 | 最高私募估值和可见的 Helix 数据飞轮 | 资本预期极高;家庭和工业范围都很宽 |
| Apptronik | 美国人形机器人平台公司 | A 轮总额 >$935M;估值约 $5B-$5.3B | 双足和轮式 Apollo | Mercedes、GXO、Jabil、Google DeepMind 与 Robot Park | 合作伙伴阵容深,且明确提供轮式工业选项 | CEO 称行业仍处原型阶段;远程操作 / 数据采集负担重 |
| Agility Robotics | 物流 / 制造双足部署专家 | 拟与 Churchill 合并,毛收益 >$620M | 双足 Digit | GXO RaaS、Amazon 测试、Toyota Canada、100k 个周转箱 | 付费仓储部署证明最可见 | 首个用例更窄;存在规模化和公开市场执行风险 |
| 1X Technologies | 家庭优先的人形机器人,正扩展到工业 | 据报道的融资推进 / 估值在本章未获一手佐证 | 采用软体 / 肌腱设计的双足 NEO | NEO 预订单;EQT 框架最多 10k 台 | 消费者学习闭环和远程操作透明度 | 依赖 Expert Mode,且 EQT 不是有保障采购订单 |
| Tesla Optimus | 潜在进入者 / 一体化 EV-AI 平台 | Tesla 公司资产负债表支撑,无单独融资轮 | 双足 Optimus | 招聘和 Gen 3 生产雄心 | 若能执行,纵向整合和 AI / 制造规模是优势 | 公开客户证明有限;有供应链延迟和工厂质疑报道 |
| Boston Dynamics Atlas | 机器人研发既有龙头 | Hyundai 支持的既有玩家;未审阅独立融资轮 | 电动双足 Atlas | 产品 / 研发演示多于已审阅客户部署 | 品牌、控制技术和机器人信誉 | 商业化工业人形机器人市场进入路径不如 Digit/Figure/Apptronik 可见 |
| Sanctuary AI | 物理 AI / 灵巧操作玩家 | 审阅来源未量化融资 | 硬件无关 AI 加人形机器人路径 | 工业物理 AI 和液压手研发 | 聚焦灵巧手和多具身 AI | 审阅来源中的公开人形机器人舰队部署证明较少 |
| Unitree Robotics | 中国低成本人形机器人硬件 | G1 起价 $13.5K;零售清单中 G1 为 $21,600 | 双足 G1/H1 系列 | 开发者 / 研究采用和低价渠道可见度 | 激进低价和硬件可得性 | 工业企业支持和安全证明不够清晰 |
| UBTECH | 中国上市人形机器人 / 机器人公司 | 据报道 2026 年目标 5,000 台;HD 报道订单 | 工业 Walker S 系列加服务机器人 | 工厂 / 物流应用场景和生产目标 | 中国制造规模和宽产品组合 | 本章当前订单依赖二手报道 |
| Fourier Intelligence | 中国人形机器人和康复机器人玩家 | 据报道推出 N1 / 开源策略 | 包含 GR/N 系列的双足人形机器人产品线 | 产品发布和开发者生态信号 | 开源定位可能加快开发者采用 | 审阅来源中工业客户牵引不清晰 |
| NEURA Robotics | 欧洲物理 AI / 认知机器人玩家 | Humanoids Daily 报道 Tether 领投 $1.4B C 轮 | 下一代人形机器人和认知生态 | Automatica 2025 人形机器人 / 认知生态发布 | 欧洲大额融资信号和 SAP/NVIDIA 生态连接 | 此处证据为二手来源,并可能与人形机器人之外的更宽机器人业务重叠 |
| Flexion | 欧洲人形机器人软件 / AI 新进入者 | Humanoids Daily 报道融资 $50M | 聚焦仿人机器人 / 人形机器人智能 | 早期融资信号 | 专门的“大脑” / 软件切入角度 | 在已披露工业锚定客户上比 Humanoid 更早期 |
| 传统固定自动化 / AMR | 现状替代方案 | 既有资本开支预算和装机基础 | 非人形机器人和输送线 | 仓库 AMR 和工业机器人已经嵌入 | 窄任务技术风险更低 | 对人类空间中的移动操作灵活性较弱 |
| OEM 内部自建 | 买方自有替代方案 | 大型制造商可在内部出资试点 | 定制自动化加数据团队 | 汽车 / 物流工程组织 | 控制数据和流程集成 | 速度慢、人才消耗大,且不可能覆盖所有工作流 |
| 协作机器人 / 工业机械臂 | 相邻替代方案 | 成熟机械臂生态 | 固定式或移动式机械臂 | 在结构化工位中拣选、放置、焊接、看机 | 高可靠性和安全认证 | 需要重设计工作流,且不太适应人类空间 |
该枚举代表 2026 年工业人形机器人格局;融资和估值单元格是已审阅的最新公开信号,不是完整融资数据库。
[CP001, CP004, CP008, CP010, CP014, CP015]序数定位显示,Humanoid 工业聚焦度高,但公开规模化证明低于 Figure、Apptronik 和 Agility。
X 轴是公开工业牵引力 / 就绪度(1 低,5 高);Y 轴是公开资本 / 战略资源规模(1 低,5 高)。分数是基于引用证据的序数分析师估计,不是实测 KPI。
[CP001, CP004, CP008, CP010, CP014, CP015]Figure 和 Apptronik 把资本门槛抬到远高于 Humanoid 的位置,而 Unitree / 1X 的价格信号压低硬件预期。
柱状值混合估值、融资和拟募资金,因此图表传递的是规模信号,而不是资产负债表对比。
[CP001, CP008, CP014, CP015, CP025, CP038]3.2 能力、形态与 AI 路线
Humanoid 最可防守的技术定位是务实,而不是极致化:先从轮式工业机器人切入,与双足路线图共享上半身,并把 KinetIQ 用作编排和任务执行层。轮子并不独特,因为 Apptronik 也提供轮式和双足 Apollo 配置,独立报道还称工业客户不断要求轮式方案,以获得稳定性、电池续航和可预测性。Figure 更明显地追求以 Helix 为中心的双足 / 通用路径,其 BMW 排序工作流也强烈表明,双足操作正从演示走向工厂试验。Agility 的 Digit 为物流中的人类空间专门打造,Unitree 则推低成本双足开发者硬件,1X 则为家庭场景优化 NEO,并采用人在回路学习。含义是,功能宽度没有那么重要;关键在于机器人能否反复解决买方第一个付费工作流。[CP006, CP007, CP009, CP011, CP012, CP016]
| 采购标准 | Humanoid | Figure | Apptronik | Agility | 1X | 中国低成本 / 规模玩家 | 既有玩家 / 现状方案 |
|---|---|---|---|---|---|---|---|
| 形态 | 轮式优先;双足路线图 | 双足 Figure 03 | 双足和轮式 Apollo | 双足 Digit | 双足 NEO | Unitree/UBTECH/Fourier 双足系列 | AMR、机械臂、协作机器人、输送线 |
| 工业目标市场 | 制造和物流优先 | BMW 制造 / 物流,加家庭叙事 | 制造、物流、零售、医疗路线图 | 仓储 / 物流和制造 | 家庭优先;EQT 工业框架 | 工厂、研究、娱乐、物流信号 | 已知窄工作流 |
| AI 路径 | KinetIQ VLA / 舰队编排 | Helix / Helix 02 具身 AI | 具身 AI、Gemini Robotics 合作、Robot Park 数据 | Agility Arc 和物流机器人技能 | Redwood AI 加 Expert Mode | UnifoLM、开源和中国生态信号 | 基于规则、面向特定任务的成熟自动化 |
| 载荷 / 操作 | 简报称双臂载荷 15 kg | BMW 钣金和排序任务;未审阅确切载荷 | Business Insider 报道 Apollo 2 双臂可举 55 lb | 高载荷周转箱搬运;IEEE 称可举 16 kg | 25-DOF 手;家务 | Unitree G1 手臂载荷约 2-3 kg;UBTECH/Fourier 此处未知 | 工业机械臂载荷可超过人形机器人 |
| 商业证明 | Schaeffler 数千台;Bosch/SAP/NVIDIA/Siemens 简报信号 | BMW 30k 辆车和 Figure 03 回归 | Mercedes/GXO/Jabil 和 Robot Park | GXO RaaS、Amazon、Toyota、100k 个周转箱 | 预订单和 EQT 框架 | UBTECH 目标 / 产量报道;Unitree 零售渠道 | 装机基础最强 |
| 价格透明度 | 工业合同定价未披露 | 未披露 | 未披露 | RaaS,价格未披露 | 据报道 $20k 或 $499/月 | Unitree G1 起价 $13.5k;Walmart G1 $21.6k | 集成商报价已知,但高度定制 |
| 地域 | UK/欧洲,工程团队在 Boston/Vancouver | 美国 | 美国 | 美国 | 挪威 / 美国 | 中国权重高的玩家 | 全球既有玩家 |
| 信任 / 安全姿态 | 需要通过 Schaeffler beta 证明 | 已披露工厂经验;家庭使用扩大安全负担 | 人本提示;仍处原型 / 数据阶段 | 真实仓库安全证明正在出现 | 人在环隐私和可靠性问题 | 价格更低,但企业安全支持不清楚 | 认证工作单元和成熟安全制度 |
| 供应链 / 制造 | Bosch/Schaeffler 锚定;简报给出规模目标 | C 轮为 BotQ 制造提供资金 | Robot Park 与投资方供应链伙伴 | RoboFab 产能与 SPAC 路径 | 1X 提到 Hayward 工厂 | 垂直整合与中国规模压力 | 成熟供应商和集成商 |
单元格使用本章审阅的公开证据;缺乏支撑的精确载荷、价格或合同单元格标为未知,或作定性描述。
[CP006, CP007, CP009, CP011, CP012, CP016]Humanoid 的能力广度可信但不唯一;关键缺口是已披露自主率、价格和合同经济性。
矩阵在所有供应商规格或价格未完全公开处使用定性强度标签。
[CP006, CP007, CP009, CP012, CP016, CP017]3.3 商业牵引与打包信号
最强的竞争证明分布并不均匀。Figure 可以指向 BMW 生产贡献和后续 Figure 03 工作流;Apptronik 可以指向 Mercedes、GXO、Jabil、Google DeepMind 和专用数据工厂;Agility 可以指向与 GXO 的多年 RaaS 协议,以及公开描述的现场仓库工作。Humanoid 的 Schaeffler 协议和 Bosch 制造关系具有战略价值,因为它们提供工业触达、执行器 / 组件相邻性和潜在欧洲制造规模,但公开记录仍需要客户层面的合同经济条款、部署爬坡和验收标准。定价更不透明。1X 和 Unitree 公布或被报道有标价,而工业玩家通常披露试点、RaaS 或战略协议,却不披露单价、利用率、服务毛利率或取消条款。这种不透明让竞争定价成为尽调优先项,而不是已经解决的对比。[CP004, CP005, CP010, CP013, CP018, CP022]
| 供应商 | 公开价格或套餐信号 | 包含能力 | 未知项 / 注意事项 | 给 Humanoid 的启示 |
|---|---|---|---|---|
| Humanoid | 无公开单价;简报称 34,000 个预订单价值 $2.4B | 轮式 / 双足 HMND 01 Alpha、KinetIQ、工业部署 | 预订单质量、Bosch 1,000 台口径和服务经济性都需要一手文件 | 必须证明订单质量和工业 TCO,不能只看积压订单的标题数字 |
| Figure AI | 未见公开工业价格 | Figure 03 与 Helix;BMW 物流 / 制造工作流 | 客户合同金额和单位经济性未披露 | 若 BMW 成果放大,资金实力可支撑激进定价 |
| Apptronik | 无公开价格;估值和融资公开 | Apollo 2/3、轮式与双足形态、Robot Park 数据项目 | Robot Park 仍在原型 / 数据阶段;商业 SKU 时间不透明 | 直接竞争 Humanoid 的轮式工业逻辑 |
| Agility Robotics | 与 GXO 签有 RaaS 协议;价格未披露 | Digit 加 Agility Arc 车队平台 | RaaS 价格、利用率和毛利率未披露 | 付费工作流证据强,抬高 Humanoid 试点门槛 |
| 1X Technologies | 据报道一次性购买 $20,000,或 $499/月、$200 定金 | NEO 家用机器人、家务、AI 伴侣、Expert Mode | 消费端价格未必适用于工业部署;远程操作负担 | 给出透明低端价格锚,但未证明工业可靠性 |
| Unitree Robotics | G1 起价 $13.5K;据报道 Walmart 标价 $21,600 | 面向开发者 / 研究的双足硬件,23-43 个电机 | 企业支持、安全和集成成本未知 | 形成硬件价格锚,可能压低外界对人形机器人 ASP 的认知 |
| UBTECH / Fourier / NEURA / Flexion | 有融资、生产或发布信号;未见可比单价 | Industrial Walker、N1/开源、认知平台 | 依赖二手来源,产品范围不一 | 欧洲和中国入局者削弱独特性叙事 |
| 既有自动化 | 定制项目报价;通常已纳入预算 | AMR、输送线、机械臂、协作机器人、集成商服务 | 灵活性较低,但可靠性和安全论证更成熟 | Humanoid 必须证明首批工作流的 ROI 胜过更简单的自动化 |
价格不可简单横向比较:公开标价多是消费端 / 开发者硬件,工业人形机器人部署大多是私下 RaaS、试点或战略合同。
[CP001, CP004, CP008, CP014, CP015, CP017]3.4 护城河、威胁与现状替代品
持久优势可能来自工作流数据、客户集成权、安全证据和供应链执行,而不是单一头条人形机器人规格。这有利于拥有重复现场任务的公司:Agility 的周转箱处理、Figure 的 BMW 工作、Apptronik 的 Robot Park 和客户数据计划,以及 Humanoid 的 Schaeffler/Bosch 工业锚点。反向情景是,这些资产本身仍早。独立和半独立来源描述的行业仍充满原型、远程操作、赋能协议而非有约束力订单、供应链约束,以及前从业者的怀疑——他们认为工厂可能偏好更简单的自动化。Unitree 和 UBTECH 从中国压低硬件价格;NEURA 和 Flexion 表明 Humanoid 并不是唯一的欧洲主张者;既有厂商也能继续扩展固定自动化、AMR 和协作机器人。Humanoid 的护城河有合理性,但承保时应把它看作执行速度加上特权工业学习,而不是已成定局的技术垄断。下一个检查点是每个客户现场的自主生产小时证据,因为该指标能把可防守的运营护城河与仅仅资金充足的演示周期区分开。[CP019, CP020, CP031, CP037, CP038, CP039]
| 护城河主张 | 威胁 | 严重性 | 证据 | 缓释措施 / 尽调要求 |
|---|---|---|---|---|
| 欧洲优先的工业锚定客户 | NEURA、Flexion 和具全球业务的美国龙头也在争夺欧洲制造商 | 高 | Humanoid、NEURA 和 Flexion 均有欧洲定位 / 融资信号 | 索取已签客户合同、部署排期和地域排他条款 |
| 轮式优先的务实设计 | Apptronik 也交付轮式 Apollo,客户可能把轮式视为通用品类 | 高 | Forbes 和 BI 描述了工业场景对轮式稳定性的需求以及 Apollo 轮式版本 | 按任务核算 ROI、安全论证和正常运行时间,而不只看形态 |
| KinetIQ 自研 AI 大脑 | Figure Helix、Apptronik/Google DeepMind、1X Redwood 和 Tesla AI 都在争夺具身 AI 人才 | 高 | 多家竞争对手披露 AI / 数据飞轮策略 | 索取首批工作流基准、干预率、数据权利和模型更新节奏 |
| Schaeffler/Bosch 客户触达 | 已宣布订单未必等于约束性收入;Bosch 单量在此处未获一手佐证 | 高 | Schaeffler 有公开协议;Bosch 细节来自简报 | 取得主服务协议(MSA)、采购订单、验收标准、取消权和付款里程碑 |
| 快速执行文化 | 资金雄厚的对手能吸收失败,也能招揽稀缺机器人人才 | 中 | Figure 和 Apptronik 的融资 / 估值信号大得多 | 跟踪招聘速度、烧钱速度、硬件迭代周期和供应商承诺 |
| 工业部署数据 | 客户数据能形成专有资产,但试点可能始终规模小或依赖远程操作 | 中 | Figure BMW、Apptronik Robot Park、Agility GXO 和 Humanoid Schaeffler 都在形成数据循环 | 尽调数据所有权条款,以及自主运行小时与远程操作小时占比 |
| 硬件集成和供应链 | 中国低成本硬件与稀土 / 地缘政治冲击可能重置成本曲线 | 中 | Unitree 价格信号和 Tesla 供应链延迟报道 | 索取 BOM、关键供应商、执行器良率和中国敞口 |
| 通用人形机器人叙事 | 既有自动化可能更便宜、更可靠地解决目标工作流 | 高 | 前 Tesla Optimus 负责人质疑,以及成熟 AMR / 协作机器人替代方案 | 将每个付费工作流与 AMR / 协作机器人 / 固定式自动化报价和回本周期对比 |
风险严重性是基于公开证据的分析师判断;私下合同、自主性和成本数据可能大幅改变每项评级。
[CP018, CP019, CP020, CP031, CP033, CP034]Humanoid 的投资逻辑取决于能否先于资本雄厚的对手,把客户入口转化为自主吞吐量。
KPI 值是基于公开来源交叉验证的序数尽调判断。
[CP019, CP020, CP031, CP033, CP034, CP042]3.5 图表
04财务情况
4.1 资本结构与估值信号
Humanoid 的财务画像是一家资金充足、处于早期商业化阶段的工业硬件公司:公司有大量股权缓冲、独角兽估值和战略投资人,但没有公开收入报表。最干净、相互印证的融资事实是 2026 年 Series A 轮:独立报道显示,公司以 $1.35 billion 投后估值融资 $152 million,使披露的累计融资达到 $270 million。这意味着 Series A 轮前约融资 $118 million,但公开记录无法清楚区分种子轮、过桥、SAFE、优先股或战略合同经济条款。因此,本章把 Series A 轮视为现金能力和验证,而不是单位经济证明。Schaeffler 和 Bosch 在财务上重要,因为它们出现在故事两端:既是投资人或战略参与方,也是制造 / 部署对手方。建模时,应把融资拆成已披露现金流入、未披露资本条款和战略期权价值。公开事实可以支持流动性情景,但没有文件就无法支持股价瀑布或优先股调整后估值。[CI001, CI002, CI003, CI006, CI007, CI008]
| 轮次 / 指标 | 金额(USD) | 日期 / 状态 | 投资方 / 来源 | 投资判断含义 |
|---|---|---|---|---|
| 二手报道据称提及的种子 / 早期轮 | ~33M | 公开条款未披露 | Forbes 2026 年 2 月报道与创投资料线索 | 仅用作此前融资的方向性参考;无公开投资条款清单、证券类型或优先股堆叠 |
| 累计融资暗示的其他 A 轮前资金 | ~85M | 按此前资本扣除据报道种子轮推算 | $270M 累计融资减去 $152M A 轮和 ~$33M 种子轮 | 桥轮 / 延长期细节是重要股权结构表缺口 |
| A 轮 | 152M | 2026-07-21 宣布 | Prime Movers Lab 领投;Schaeffler、Bosch、Fubon、Aglaé 参投 | 新股现金垫和战略背书,不是收入证据 |
| 截至目前累计融资 | 270M | 截至 runDate | 多家独立融资报道 | 资本底座可观,但必须支撑硬件规模化和测试版部署 |
| 投后估值 | 1,350M | A 轮投后 | 融资报道和 Reuters 联合转载报道 | 已确认收入未披露,估值主要由预订额支撑 |
列举不完整:公开来源披露的是融资总额和投资方名单,但未披露完整证券条款、优先股堆叠、期权池、老股交易或确切种子 / 桥轮时间线。
[CI001, CI002, CI003, CI004, CI005, CI006]| KPI | 公开值 | 模型用法 | 置信度 | 尽调路径 |
|---|---|---|---|---|
| 投后估值 | $1.35B | 已知估值锚 | 高 | 确认股数、期权池、清算优先权和战略附属协议 |
| 累计融资 | $270M | 现金容量代理指标 | 高 | 获取银行余额、交割日期、费用和已花现金 |
| 已确认收入 | 保持空值;不要用预订额替代 | 中 | 索取按合同和季度拆分的 GAAP/IFRS 收入 | |
| ARR / RaaS 运行率 | 保持空值,直到看到经常性合同 | 中 | 索取 MRR/ARR 明细和已部署机器人利用率 | |
| 毛利率 | 保持空值;仅做情景模型 | 中 | 索取 BOM、人工、质保、服务和物流单位成本 | |
| 月度烧钱 | 估算情景区间;不要断言实际烧钱速度 | 低 | 索取现金流量表、采购承诺和招聘计划 | |
| 订单簿 / 预订额 | $2.4B 预订单价值 | 仅为未来收入信号 | 中 | 将每笔订单对应到已签合同、定金、取消权和交付排期 |
| 员工数 | ~160–200 人 | 仅为烧钱代理指标 | 中 | 核对薪资名单、承包商和制造用工责任 |
空值表示该指标未在审阅来源中公开披露,不是零。
[CI001, CI002, CI009, CI010, CI034, CI035]公开融资里程碑最终落到 $152M Series A,以及据报道累计 $270M 资本。
早期轮次时间和种子轮拆分在公开来源中不完整;Series A 与累计融资是高置信锚点。
[CI001, CI002, CI003, CI005]$152M Series A 资金在定性口径上用于部署、制造扩产和 AI 平台开发;具体金额分配未披露。
仅披露 $152M 流入;流出类别和金额是作者基于已说明资金用途主题和硬件扩产需求设定的情景。
[CI001, CI006, CI017, CI026, CI037]4.2 订单簿是前瞻信号,不是已确认收入
最强的收入线索是订单簿:公开来源反复提到 34,000 个预订单、价值约 $2.4 billion,意味着折扣、软件、服务、税费和流失前每台机器人约 $70,600。这是有价值的需求指标,但不是已确认收入、ARR 或毛利润。收入确认将取决于已签合同、交付与验收、取消权、定金、融资、保修,以及模式是硬件销售、租赁还是机器人即服务。Schaeffler 协议和 Bosch 制造合作提升了工业客户正在参与的置信度,但公开证据仍不清楚 Bosch 是制造商、客户、1,000 台订单对手方,还是三者皆是。合同条款得到审阅前,正确承保姿态是以订单簿驱动的上行空间,但带转化风险。因此,本章对已确认收入和 ARR 使用明确空值,同时把订单簿信号作为管线指标保留。这样可以避免硬件尽调常见错误:在交付和服务经济已知之前,就按软件倍数资本化订单标题。[CI010, CI011, CI012, CI013, CI014, CI015]
| 信号 | 公开数量 / 价值 | 收入状态 | 质量 | 关键尽调要求 |
|---|---|---|---|---|
| 预订单 | 34,000 台机器人 / ~$2.4B | 仅为预订额 | 规模大,但转化率未知 | 已签订单清单,含定金、取消权和交付日期 |
| 隐含 ASP | ~$70.6K / 台 | 由订单价值估算 | 有用,但为混合口径且未扣折扣 | 按 SKU、租期、软件、服务、税费和地区拆解价格桥 |
| Schaeffler 协议 | 数千台机器人 / 二手报道称 1,000+ 台 | 未来部署协议 | 战略性工业证据 | 主服务协议、验收标准和分阶段量产排期 |
| Bosch 伙伴关系 | 2026 年 3 月 PoC 后的制造伙伴;官方页面未列公开订单台数 | 制造赋能 | 降低自建工厂需求,但可能产生供应承诺 | 确认 Bosch 是否也是客户 / 订单对手方,以及台数 |
| Fortune 500 部署 | 报道称有 9 个工业部署 | 试点 / 证明信号 | 良好的销售效率代理指标 | PoC 到付费转化、周期长度和扩张经济性 |
| 测试客户机器人 | 计划于 2026 年晚些时候 | 未来交付触发点 | 不是当前收入 | 验收里程碑和质保义务 |
每一行都只按预订额或市场牵引力处理,除非合同、交付、验收和可收款性证据证明收入已确认。
[CI010, CI011, CI012, CI013, CI014, CI015]| 收入流 | 机制 | 当前公开状态 | 收入质量 | 尽调要求 |
|---|---|---|---|---|
| 机器人硬件销售 | 交付并验收后按台销售 | 订单簿存在;交付尚未公开确认为收入 | 客单价可能高,但不稳定 | 已执行 PO、发票、验收、付款、质保条款 |
| RaaS / 租赁 | 按机器人 / 时间 / 产出收取经常性费用 | 未公开披露为已上线合同模式 | 可能提升采用率,但拉长回本周期 | 租赁合同、利用率、残值、融资成本 |
| KinetIQ / AI 软件 | 软件嵌入机器人,或按能力层售卖 | 公开描述为自研 AI 大脑;价格未披露 | 若单独变现,可能撬动毛利 | 价目表、附加率、数据权利、升级条款 |
| 维护和支持 | SLA、备件、现场服务、正常运行时间保证 | 无公开毛利率或附加率数据 | 必需,但早期可能稀释毛利 | SLA 罚则、每台已部署机器人的支持成本 |
| 集成 / PoC 费用 | 付费试点和工作流集成 | PoC / 部署已有公开描述,但经济性未披露 | 若为付费则验证力强;若是免费试点则弱 | 试点发票、转化率、CAC / 回本周期代理指标 |
| 制造 / 伙伴经济性 | Bosch 和 Schaeffler 降低商业落地和供应摩擦 | 战略背书,不是直接收入披露 | 可降低资本开支,但可能要求承诺 | 伙伴条款、排他性、批量返利、部件定价 |
收入流是需要尽调的机制;没有私下合同证据,任何一项都不应建模为已实现收入。
[CI012, CI016, CI017, CI020, CI021, CI037]| 定价输入 | 数值 / 状态 | 置信度 | 模型处理 | 尽调要求 |
|---|---|---|---|---|
| 混合预订单 ASP | ~$70.6K | 中 | 作为由预订额推导的混合 ASP 使用,不当作标价 | 按 SKU、配置、软件、服务和地区拆分 |
| 标价 | 低 | 保持空值 | 官方价目表和折扣政策 | |
| 实际成交价 | 低 | 保持空值 | 已签客户发票和返利后净价 | |
| 定金 / 取消条款 | 低 | 未知前应用转化折扣 | 定金时间表、可退款性、取消权 | |
| RaaS 月费 | 低 | 仅为情景假设 | 签约月费、利用率下限、融资成本 | |
| 服务附加 / 维护 | 低 | 仅为情景假设 | SLA 定价、备件毛利、现场服务人员配置 |
唯一已量化的定价输入,是隐含的订单 ASP;所有实际成交价格字段仍需向公司索取私有数据。
[CI010, CI011, CI012, CI020, CI042]订单只有经过合同、交付、验收、可收回性和成本证据,才会变成收入。
这座桥是收入确认控制清单,不是在声称公开预订单就是收入。
[CI010, CI011, CI012, CI020, CI024, CI025]4.3 硬件单位经济决定现金流情景
Humanoid 不应按软件公司建模。公开成本研究显示,执行器占人形机器人成本的重要部分;计算、传感器、电池、机械结构、装配、测试、部署工程、支持和保修准备金还会继续消耗现金。Bosch 作为合同制造伙伴,可能降低 Humanoid 自建工厂 capex 需求,但无法消除库存融资、最低订购量、供应商承诺、返工或客户验收风险。由于没有公开来源披露 BOM、人工含量、毛利率、利用率、服务附加或 RaaS 回本,单位经济桥接基本仍是尽调模板。财务模型应把毛利率留空,对烧钱速度使用情景区间,并且只在看到转化证据后才给订单簿计入价值。规模化人形机器人集群还会制造时间错配:供应商可能要求客户验收机器人前先付定金,而保修和现场支持会在出货后消耗现金。第一批生产队列中,这些时间项可能主导会计毛利率。[CI024, CI025, CI026, CI027, CI035, CI036]
| 成本项 | 公开证据 / 代理指标 | 毛利压力 | 尽调要求 |
|---|---|---|---|
| 执行器 | K2 Capital 估计执行器占人形机器人成本的 30–50% | BOM 高度集中,供应商风险高 | 供应商报价、自制 / 外采拆分、良率和降本路线图 |
| 计算和 AI 硬件 | 共享简报的产品材料提到 NVIDIA/Intel 级计算栈 | 前期成本高,存在更新换代风险 | SKU 级计算 BOM 和淘汰假设 |
| 传感器和感知 | 工业机器人需要摄像头、力 / 扭矩、安全和校准栈 | 可靠性和校准成本中高 | 传感器 BOM、校准人工、更换率 |
| 机械框架和电池 | 机器人重量 / 载荷要求结构、电池、热设计和续航配套 | 中;推高运输和服务成本 | 电池循环寿命、框架良率、运输和安装成本 |
| 制造和测试 | Bosch 合作将生产推向代工制造 | 可能降低资本开支,但不能降低库存融资需求 | 代工制造协议、工装所有权、MOQ 和质保分摊 |
| 现场服务 / 质保 | 工业部署要求正常运行时间、支持、备件和安全响应 | 早期可靠性弱时会吃掉毛利 | SLA 条款、MTBF、支持团队人数、备件储备 |
成本项是公开代理分类;Humanoid 未披露 BOM、COGS、质保准备金或毛利率。
[CI024, CI025, CI026, CI037]| 指标 | 公开数值 | 置信度 | 重要性 | 尽调请求 |
|---|---|---|---|---|
| 单台机器人 COGS / BOM | 低 | 决定 $70K 订单 ASP 能否撑起正毛利 | 供应商报价、组装人工、良率、报废率与运费 | |
| BOM 中执行器占比 | K2 参考代理值 30–50% | 中 | 最大的降本杠杆,也是供应商瓶颈 | Humanoid 执行器自制 / 外采、供应商条款与降本曲线 |
| 毛利率 | 低 | 估值和烧钱速度的核心输入 | 按硬件、软件、服务与质保拆分的毛利率桥 | |
| 质保准备金 | 低 | 早期硬件可靠性问题会吃掉贡献毛利 | 质保索赔模型、MTBF、现场维修成本 | |
| RaaS 回本周期 | 低 | 订阅模式可能在回本前消耗现金 | 单台机器人 Capex、月费、融资利率、利用率、残值 | |
| 部署 CAC / 回本周期代理指标 | 低 | 企业试点周期可能很长,且大量占用工程师 | PoC 成本、销售周期、转化率、扩张率 | |
| 单台机器人营运资本 | 低 | 库存备货可能在收入确认前占用现金 | 客户预付款、供应商账期、在制品、产成品周转天数 |
空值是刻意保留的投资测算缺口;这张表是尽调请求清单,不是已填好的模型。
[CI009, CI011, CI024, CI025, CI026, CI037]现金余额和月度烧钱速度未披露,现金跑道只能按情景估算。
区间用 Series A 总额除以示意烧钱水平得出;不是公司指引,且未纳入此前现金、收入、债务和营运资金时点。
[CI034, CI035, CI036, CI037]关键现金流风险卡在零部件采购、制造扩产、客户验收和服务义务之间。
矩阵条目把公开信号转成现金流尽调议题;不假设任何私人合同经济条款。
[CI017, CI024, CI026, CI027, CI037, CI042]4.4 备案错配与私有指标是尽调闸门
UK Companies House 搜索有用,但不能一锤定音。可见的 HUMANOID LTD 记录显示一家存续私营公司、休眠账户,以及与机器人运营故事并不明显匹配的 SIC 代码 / 高管细节。这可能意味着运营集团使用另一实体,近期进行了重组,或存在无法用显而易见搜索词找到的备案。财务上,这是重大缺口,因为公开线索无法验证现金、债务、收入、客户定金、集团内 IP 所有权或优先股堆叠条款。因此,结论既不是由收入支撑,也不是纯粹投机:Humanoid 有可信的战略融资、大额订单标题和制造伙伴,但承保情景取决于在下一次融资触发前,把预订单转化为已交付机器人并实现正贡献毛利。下一轮尽调应从实体确认和私有财务模型开始,而不是继续搜新闻。如果管理层能对备案、合同可执行性和贡献毛利做出对账,融资故事会更可承保;否则,估值主要仍是期权价值。[CI009, CI028, CI029, CI030, CI031, CI032]
| 主题 | 公开证据 | 财务含义 | 严重性 | 下一步尽调 |
|---|---|---|---|---|
| 账面现金 | 累计融资 $270M;当前现金余额未披露 | 扣除既有支出和费用后,实际现金跑道无法判断 | 重大 | 银行流水、资金计划表与交割后现金余额 |
| 月度烧钱速度 | 未披露;仅有情景区间 | 没有烧钱速度,就无法判断下一轮融资时点 | 重大 | 按薪酬、研发、工装、库存和试点拆分的月度现金流 |
| 现金跑道情景 | $152M 在每月 $4–6M 烧钱速度下可支撑约 24–36 个月,在每月 $6–10M 下可支撑约 15–25 个月 | 这是有用的敏感性分析,不是管理层指引 | 重大 | 含招聘、资本开支、零部件预付款和回款的董事会计划 |
| 债务 / 项目融资 | RaaS 机队融资可能需要 | 重大 | 债务额度、租赁 SPV、留置权、契约条款与追索安排 | |
| Companies House 记录 | HUMANOID LTD 休眠申报 / SIC 不匹配 | 实体身份和运营公司财务仍未厘清 | 重大 | 确认法律实体、集团架构、IP 持有人与审计账目 |
| 股权结构表 / 优先权 | 可访问数据库中没有 | 估值标题可能掩盖优先权经济性 | 重大 | 融资文件、清算优先权、反稀释、认股权证、附函 |
| 客户集中度 | 大型协议集中在 Schaeffler、Bosch 与 Fortune 500 试点 | 订单可能依赖少数工业交易对手 | 重大 | 逐客户订单积压与集中度明细 |
资金充足性只能按情景呈现,因为实际现金、烧钱速度、债务和股权结构条款都未披露。
[CI001, CI002, CI009, CI016, CI029, CI030]4.5 图表
05产品与技术
5.1 产品定义与 HMND 01 Alpha 版本
HMND 01 Alpha 最好理解为工业劳动力自动化机器人,而不是通用消费级人形机器人。当前产品故事有两个形态,共享模块化上半身和 AI 栈:Alpha Wheeled 是更近期的工业平台,面向物流和制造工作流;Alpha Bipedal 是行走平台,用于研发、服务学习和未来更接近人类的导航。可投资产品证据在轮式版本上最强,因为 SAP/Martur Fompak、Siemens、Bosch 和 Schaeffler 的证明都围绕工厂或厂内物流工作。标准尽调基准是:不含末端执行器的 29 个自由度、15 kg 双臂载荷、高扭矩执行器、可更换手或夹爪,以及面向感知、操作和人机交互的传感器包。双足公开规格链条不够干净,因此本章把冲突公开页面视为尽调缺口,而不是改写共享报告基准。[CE001, CE002, CE003, CE004, CE005, CE006]
| 模块 / 资产 | 主要用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| HMND 01 Alpha Wheeled | 制造与物流运营方 | Alpha 已进入现场 PoC;路线图指向 Beta 和大规模制造 | 工业优先移动底盘,15 kg 载荷,四小时续航,轮式速度 2 m/s | 按站点拆分的生产可用时长、SLA 与验收标准 |
| HMND 01 Alpha Bipedal | 研发团队,以及未来服务 / 家庭工作流 | Alpha 研发平台;生产就绪度较低 | 不同形态共用上半身与 KinetIQ 学习能力 | 厘清公开速度 / 续航冲突与现场测试范围 |
| KinetIQ AI 大脑 | 机队运营方与机器人自主栈 | 技术架构已披露;性能声明来自公司 | 覆盖机队、单机、VLA 与全身控制的四层层级 | 独立基准测试与干预率证据 |
| KinetIQ Ascend | 操控策略训练团队 | 公司称生产任务上的 RL 有提升 | 从遥操作模仿推进到边练边学的 RL | 吞吐量 / 成功率提升的外部复现 |
| 末端执行器 | 工业任务设计方 / 维护团队 | 已披露可更换五指手或平行夹爪 | 模块化可按任务成本与可靠性匹配灵巧度 | 末端执行器寿命、维护与失效模式数据 |
| 传感器与计算包 | 控制、感知、安全团队 | 已描述丰富感知;计算依赖部分在外部 | 360° RGB、深度、腕部相机、力 / 力矩、麦克风、NVIDIA 计算 | 精确 BOM、安全等级传感器、Intel i9 配置 |
| 工业集成层 | 工厂 IT / 运营 | SAP、Siemens、Bosch、Schaeffler PoC 与协议 | 借助 SAP/Siemens 加制造伙伴完成企业任务下发 | 网络安全、变更管理与客户 IT 验收 |
公开产品模块和成熟度来自官方产品页、技术文章、合作伙伴 PoC 与独立规格数据库;私有验收标准仍不可得。
[CE001, CE002, CE003, CE004, CE005, CE007]| 规格 | Alpha Wheeled | Alpha Bipedal | 尽调解读 | 主要证据 |
|---|---|---|---|---|
| 客户角色 | 先切入工业物流 / 制造 | 研发 / 服务学习,以及未来类人导航 | 轮式形态更接近收入和部署证明 | 官方产品页与 NVIDIA 集成文章 |
| 高度 | 220 cm | 179 cm | 双足接近人形尺度;轮式更像设备尺度 | 官方 / 独立规格来源 |
| 重量 | ~300 kg | 90 kg | 轮式重量带来楼板承载和安全问题 | 官方 / 独立规格来源 |
| 自由度 | 29,不含末端执行器 | 29,不含末端执行器 | 标准口径不含五指手 / 夹爪 | 官方 / 独立规格来源 |
| 双臂载荷 | 15 kg | 15 kg | 足以搬运周转箱和小型工业零件;PoC 使用了更低载荷 | 官方 / 独立规格来源 |
| 最高速度 | 2 m/s | 1.5 m/s 标准口径 | 双足公开页面存在冲突;采用共享标准基线 | 共享简报、ChipSilicon、RoboZaps |
| 平均续航 | 4 hours | 3 hours 标准口径 | 电池续航是尽调闸门项 | 共享简报、ChipSilicon、RoboZaps |
| 末端执行器 | 12-DoF 五指手或 1-DoF 夹爪 | 12-DoF 五指手或 1-DoF 夹爪 | 按任务权衡灵巧度 / 成本 | 官方 / 独立规格来源 |
| 传感器 | 360° RGB、深度、腕部相机、6D F/T、麦克风 | 360° RGB、深度、腕部相机、6D F/T、麦克风 | 传感器冗余和安全评级未公开 | ChipSilicon 与 RoboZaps |
| 就绪度 | 工厂 PoC 与 Beta 路线图 | 原型 / 研发平台 | 不同形态的 TRL 差异很大 | PoC 与路线图来源 |
本表枚举当前产品来源披露的两款 HMND 01 Alpha 变体;双足续航 / 速度冲突按缺口处理。
[CE002, CE003, CE004, CE005, CE006, CE007]Alpha Wheeled 承载近期工业载荷 / 续航场景;Alpha Bipedal 更轻、成熟度更低。
轮式重量约 300 kg;尽管公开页面相互冲突,双足速度 / 续航仍采用共用标准基线。
[CE003, CE004, CE006, CE045]5.2 架构:硬件、计算、KinetIQ 与集群学习
技术逻辑是垂直整合:Humanoid 把机器人底盘、模块化末端执行器、机载感知、NVIDIA 级边缘计算、仿真,以及自研 KinetIQ 控制和编排栈组合起来。对一家年轻硬件公司来说,KinetIQ 的披露异常具体。它把集群编排、单机器人任务规划、VLA 位移操作和全身控制分成四层;披露的 System 1 与 System 0 频率为 10 Hz 和 50 Hz,让尽调有了营销语言之外可测试的内容。随后,KinetIQ Ascend 试图把操作从被动遥操作模仿推进到真实生产任务上的强化学习。这个方向令人鼓舞,因为公司承认行为克隆在速度和可靠性上的天花板;但也意味着自主成熟度取决于数据质量、干预捕捉、仿真到现实迁移和部署反馈回路,而这些仍大多是私有信息。[CE008, CE009, CE010, CE011, CE012, CE013]
| 层级 / 组件 | 作用 | 依赖 | 风险 / 尽调请求 |
|---|---|---|---|
| 传感器:360° RGB + 深度 + 腕部相机 | 用于导航、拣选和操控的感知 | 相机 / 深度供应商未披露 | 安全等级冗余与传感器退化数据 |
| 6D 力 / 力矩 + 触觉 | 接触控制与操控反馈 | 末端执行器与关节集成 | 校准漂移和过载 / 失效数据 |
| 6 麦克风阵列与 HRI 显示屏 / 扬声器 | 人机交互与工作流信号 | 音频隐私与工厂政策 | 隐私控制与噪声鲁棒性 |
| NVIDIA Jetson Orin / Thor | 板载 AI 推理与控制依赖 | NVIDIA 计算路线图与供应 | BOM 锁定、热余量、出口限制 |
| Intel i9 标准协同计算 | 共享简报和规格聚合站显示,它提供通用计算支持 | 精确 SKU 未公开 | 确认 SKU、生命周期与功耗预算 |
| Isaac Sim / Isaac Lab | 仿真、RL、验证、合成数据 | NVIDIA 软件生态 | sim-to-real 验证与可复现性 |
| KinetIQ System 3 / 2 | 机队与单机编排 | 企业任务 API、工厂系统 | 机队调度器安全与故障切换 |
| KinetIQ System 1 / 0 | VLA 移动操控与全身控制 | 机器人数据、遥操作、RL、实时控制 | 干预率、10 Hz / 50 Hz 时序余量 |
| KinetIQ Ascend | 基于 RL 的操控改进 | 持续训练基础设施 | 可靠性提升的独立证据 |
架构行综合了官方 KinetIQ 文章、NVIDIA 技术文档、招聘信息和独立规格聚合站;精确 BOM 与安全等级组件清单未公开。
[CE007, CE008, CE009, CE010, CE011, CE012]HMND 01 把模块化硬件、传感、算力、KinetIQ 控制、企业集成和仿真 / 机队学习拼在一起。
该栈是有证据支撑的架构抽象;精确生产 BOM 和数据平面拓扑仍属非公开信息。
[CE002, CE007, CE008, CE010, CE011, CE012]KinetIQ 把机器人智能从机队级目标一路拆到 50 Hz 全身控制和数据反馈。
该流程抽象自 Humanoid 公开的 KinetIQ 层级和学习闭环讨论。
[CE012, CE013, CE014, CE015, CE016, CE017]5.3 部署证明与客户工作流集成
最强的技术证明不是单台机器人演示,而是工业工作流内企业系统集成的模式。SAP/Martur Fompak POC 展示了机器人通过 Joule 和 SAP EWM 接收任务上下文,导航至托盘,抓取 KLT 箱,并交付到手推车。Siemens 提供第二个环境:HMND 01 Alpha 在 Erlangen 物流运营中测试,使用 Siemens Xcelerator 和 NVIDIA 的物理 AI 栈。Bosch 对制造准备度重要,因为它把故事从客户使用推进到合同制造和 DfX 支持。Schaeffler 重要,因为它把部署触达和执行器供应相邻性结合起来。上述内容尚不能证明规模化自主正常运行时间,但确实让 Humanoid 高于纯实验室演示状态。[CE023, CE024, CE025, CE026, CE027, CE028]
| 合作伙伴 / 场地 | 工作流 | 机器人 / 集成 | 公开结果 | 局限 |
|---|---|---|---|---|
| SAP + Martur Fompak | 现场生产设施中的 KLT 箱物流拣选 | HMND 01 Alpha Wheeled + SAP Joule/EWM 任务下发 | 机器人导航至托盘、取箱并送至手推车;SAP 称吞吐量提高、错误减少 | 没有公开 MTBF、干预率或验收阈值 |
| Siemens + NVIDIA | Erlangen 电子工厂的周转箱搬运物流 | HMND 01 Alpha + Siemens Xcelerator + NVIDIA 物理 AI 栈 | Siemens 称运营中达成目标指标 | 准确节拍、可用时长与安全论证未披露 |
| Bosch | 跨箱体尺寸的输送线到手推车内部物流 | HMND 01 进入 Bosch Bühl 动态物流环境 | PoC 先于合同制造和 DfX 协议 | 制造条款、单台成本与生产良率为私有信息 |
| Schaeffler | 箱体搬运与分阶段工厂部署 | Humanoid 机器人,加上 Schaeffler 作为首选执行器供应商 | 有约束力的分阶段部署与供应协议;第一阶段 Dec 2026-Jun 2027 | 公开合同经济性与验收标准缺失 |
| NVIDIA 生态 | 仿真、学习、推理与物理 AI 开发 | Jetson Thor、Isaac Sim、Isaac Lab、GR00T-like 构件 | Humanoid 称策略训练 / 部署可压缩到 24 小时 | 依赖合作伙伴栈,且可复现性证据为私有 |
部署证明包括官方以及合作伙伴 / 独立报道;这些结果是公开声明,不是经审计的生产 KPI。
[CE023, CE024, CE025, CE026, CE027, CE028]Humanoid 的产品成熟度取决于伙伴集成、算力平台、供应商 / 制造执行和安全验证。
依赖图把自有技术、公开伙伴依赖和运营依赖拆开。
[CE008, CE009, CE011, CE024, CE029, CE031]5.4 成熟度、路线图与技术就绪度
轮式工业楔子的公开成熟度约在 TRL 6-7:可工作的 Alpha 原型已进入相关工厂或物流场景,但稳定生产运行、客户验收标准和可重复集群经济的证据尚未公开。Alpha Bipedal 的就绪度更低,因为它被定位为开发和测试单元,而不是第一款规模化工业产品。路线图指向 Q4 2026 的 Beta 机器人、2026 年 12 月至 2027 年 6 月的 Schaeffler 现场阶段,以及轮式人形机器人的大规模制造。开发速度值得注意——Alpha 版本用时五到七个月——但速度也带来验证问题:可靠性增长、安全验证、电池耐久度、组件供应和现场服务流程必须比原型叙事更快成熟。[CE034, CE035, CE036, CE033, CE038, CE040]
| 能力 | 公开证据 | 估计就绪度 | 含义 | 下一步尽调请求 |
|---|---|---|---|---|
| 轮式移动 | SAP、Siemens、Bosch 的工厂 / 物流 PoC;2 m/s 规格 | TRL 6-7 | 已展示相关环境运行,但不是规模化机队证明 | 按站点和故障类别拆分的自主运行小时数 |
| 双足行走 | 五个月原型,48 小时后稳定行走;定位研发 | TRL 4-5 | 有前景,但不是第一收入切入点 | 地形 / 跌倒测试与恢复指标 |
| 双臂操控 | SAP 周转箱 / KLT 工作流、Bosch 箱体搬运、RL 提升 | TRL 6 | 特定物流任务具备可行性 | 按 SKU 拆分的节拍、成功率与载荷范围 |
| 企业集成 | SAP Joule/EWM 与 Siemens Xcelerator 集成 | TRL 6 | 任务下发桥是差异化点 | 网络安全、延迟与 ERP 异常处理 |
| 机队学习 | KinetIQ 记录遥操作和自主执行;DAgger 工作流 | TRL 5-6 | 可能形成数据飞轮,但主要由公司声称 | 干预频率与模型更新治理 |
| 安全与合规 | 招聘帖提到故障保护、watchdog、冗余、安全指标 | TRL 4-5 公开证据 | 已有流程信号;认证缺失 | ISO/ANSI/CE 路径、风险评估与事故日志 |
| 制造规模化 | Bosch DfX 制造合作与 Schaeffler 部署协议 | TRL 5-6 | 从 PoC 到量产的路径可信,但尚未证明 | 良率、供应商产能、服务人员配置、质保准备金 |
TRL 估计是基于公开证据的分析师判断,不是公司认证的就绪度等级。
[CE023, CE024, CE028, CE030, CE031, CE032]| 日期 / 阶段 | 功能或里程碑 | 状态 | 意义 | 证据 |
|---|---|---|---|---|
| 2025-01 | NVIDIA 技术栈整合对外发布 | 公司披露 | 仿真与 Jetson Thor 策略早早确立 | Humanoid NVIDIA 技术文章 |
| 2026-03 | SAP + Martur Fompak 现场生产物流 POC | POC 已完成 | 验证企业任务下发与轮式流程 | Humanoid、SAP、Robot Report 与 Logistics Manager |
| 2026-04 | Siemens + NVIDIA 工厂试点 | 测试已完成 | 第二个工业环境 + Xcelerator/NVIDIA 集成 | Humanoid 与 Siemens 新闻稿 |
| 2026-05 | Bosch POC 与制造合作 | 已宣布 | 把原型推进到量产规划 | Humanoid Bosch 文章 |
| 2026-05 | Schaeffler 部署与供应协议 | 公司称为有约束力的分阶段协议 | 执行器供应与工厂部署路径 | Humanoid Schaeffler 文章 |
| 2026 Q4 | Beta 机器人推出 | 路线图 | 客户设施部署的下一个关口 | Tech.eu 融资报道 |
| 2026-12 to 2027-06 | Schaeffler 场地首阶段 | 计划中 | 首次披露的大规模部署窗口 | Humanoid Schaeffler 文章 |
| 2027+ | 轮式人形机器人与下一代平台量产 | 路线图 | 取决于 Bosch/Schaeffler 供应链执行 | Tech.eu 与 Humanoid 路线图来源 |
路线图各行区分已完成的公开 POC 与未来路线图事项,投资交割前应重新验证。
[CE012, CE023, CE028, CE030, CE031, CE032]可见路线图从 2026 POC 走向 2026 Q4 Beta、Schaeffler 现场阶段,以及轮式大规模制造。
未来路线图条目是公开表述,不是已达成里程碑。
[CE012, CE023, CE028, CE030, CE031, CE032]轮式物流 POC 的就绪度最高;双足、认证和公开可靠性证据较低。
TRL 序位估计基于公开证据深度;一旦拿到客户验证数据,应替换。
[CE023, CE028, CE031, CE032, CE034, CE035]5.5 差异化与可防守性
Humanoid 的差异化产品策略务实,而不是科幻式。公司优先选择轮式工业机器人、现场工厂工作流,以及能在轮式和双足形态之间复用学习的模块化架构。可防守性建立在四项资产上:自研 KinetIQ 软件,与 SAP、Siemens Xcelerator 等企业系统的集成权,与 Schaeffler/Bosch/Siemens/Martur Fompak 的特权工业学习回路,以及垂直协调的制造 / 供应商栈。风险在于,每项资产都需要规模化证明。NVIDIA 提供的关键赋能基础设施也可供其他机器人公司使用;按公开披露,Bosch 和 Schaeffler 是战略伙伴,而不是自有工厂或独家组件来源;本次也未审阅专利资产。因此,护城河有合理性,但仍取决于执行。[CE012, CE017, CE024, CE029, CE031, CE032]
5.6 信任、安全、合规与技术风险清单
反向情景不是 Humanoid 缺少可信技术,而是人形机器人部署把难题压缩进一个会移动、很重、半自主的系统。IEEE Spectrum 的规模化批评强调电池寿命、可靠性、停机时间和安全标准;Robotics & Automation News 则强调,该品类距离通用自主仍很远。Humanoid 自己的文章也强化了这一担忧:公司解释了为什么行为克隆和遥操作演示无法达到近乎完美的工业可靠性。公开尽调因此应聚焦安全案例、正式认证路径、干预率、MTBF、电池化学和热数据、保修 / 服务义务,以及 KinetIQ 能否在多个班次中降低对遥操作的依赖,而不是只在一次 POC 中做到。在这些指标披露前,产品技术评级应是可信 Alpha,而不是已被证明的规模化平台。[CE021, CE022, CE038, CE039, CE040, CE041]
| 控制项 / 风险 | 公开状态 | 范围 | 缺口 / 尽调路径 |
|---|---|---|---|
| 故障安全、看门狗、冗余 | 控制软件招聘中提及 | 安全关键控制路径 | 要求提供架构安全论证与故障树分析 |
| 仿真、HIL、回归与性能测试 | 招聘与 NVIDIA/Isaac 材料中提及 | 控制与感知验证 | 要求提供测试覆盖率、通过 / 失败历史和发布关口 |
| 电池续航与充电作业周期 | 规格续航公开;电芯体系与循环寿命未公开 | 运营可用时长与安全 | 要求提供电池 BOM、热数据、充电循环和事故数据 |
| 自主性与远程操作 | 远程操作轨迹与共享自主路线图公开 | 数据采集与兜底操作 | 要求提供干预率、远程操作人力配比和升级日志 |
| 工业安全标准 | 已审阅来源未披露认证 | 人员近旁重型机器人部署 | 要求提供 CE/ISO/ANSI/RIA 认证计划和风险评估 |
| 可靠性 / MTBF / 可用时长 | 未披露生产级指标 | 客户 SLA 与 ROI 模型 | 要求提供 MTBF、停机类别和现场验收指标 |
| 隐私与网络安全 | SAP/Siemens 集成公开;控制措施未公开 | 企业 API、云端任务下发、摄像头 / 麦克风 | 要求提供安全架构、数据保留和渗透测试 |
| 知识产权归属 | KinetIQ 自研主张公开;专利未审阅 | 软件 / 控制护城河 | 要求提供专利申请、商业秘密范围和开源合规 |
信任表有意把未公开的认证与可靠性指标标为缺口,而不是从 POC 成功推断合规。
[CE017, CE021, CE022, CE038, CE039, CE040]5.7 图表
06客户情况
6.1 客户分层与按证据加权的买方图谱
Humanoid 的客户证据范围很窄,但战略上连贯:公开楔子是汽车制造、电子制造和物流相邻工厂流程中的工业工作。买方通常是制造运营、自动化或数字工厂高管;用户是车间物流或生产团队;付款方是企业采购组织,评估劳动力可得性、吞吐量、正常运行时间和安全性。Schaeffler 是锚定制造商,也是最强的收入质量证明点,因为公开来源描述了一项有约束力、分阶段的部署和供应协议。Bosch 最好归类为制造伙伴加 POC 对手方,而不是抓取到的公开记录中干净的终端客户订单。Siemens、SAP 和 Martur Fompak 显示了来自工厂现场和汽车物流运营方的试点 / POC 需求。零售是公司声明的未来目标,但未找到具名公开零售客户,因此不应作为已证明需求承保。[CU001, CU002, CU027, CU028, CU037, CU044]
| 分群 | 买方 / 用户 / 付款方 | 用例 | 公开证据 | 规模 / 价值信号 | 尽调缺口 |
|---|---|---|---|---|---|
| 锚定制造商:Schaeffler | 制造运营、工厂物流、采购 | 箱体搬运、内部物流、生产验证 | 有约束力的分阶段部署与供应协议 | 2032 年前四位数 / 数千台机器人;RaaS 模式 | 合同经济性、验收里程碑和场地级部署节奏 |
| 制造合作伙伴 / POC:Bosch | Bosch 制造管理层与工业内部物流团队 | 代工制造加内部物流 POC | Humanoid 与独立媒体确认 2026 年 3 月 POC 成功 | HMND 01 欧洲量产路径 | 单独客户订单规模与有约束力采购条款 |
| 工厂现场试点:Siemens + NVIDIA | 数字工厂运营、自动化工程师、物理 AI 平台负责人 | Erlangen 电子工厂周转箱搬运 | 伙伴公告与 Humanoid 带指标案例研究 | 每小时搬运 60 个周转箱、正常运行 >8 小时、抓放成功率 >90% | 试点后转化、机队规模和付费合同状态 |
| 汽车物流 POC:SAP + Martur Fompak | 汽车物流运营方与 SAP 企业工作流负责人 | 用 SAP AI 智能体拣选并配送 KLT 箱 | Humanoid 案例研究与 SAP 具身 AI 背景 | 2026 年 3 月现场生产物流 POC | 持续 ROI、可用时长、部署持续时间和商业条款 |
| 生态 / 计算合作伙伴:NVIDIA | 机器人技术栈与工业 AI 生态 | 仿真、开发与物理 AI 技术栈 | NVIDIA/Siemens 背景与伙伴公告 | 提升可信度并扩大集成面 | NVIDIA 参与究竟转化为客户需求,还是只提供技术赋能 |
| 潜在预购客户 | 物流、制造、零售等领域未具名企业买家 | 面向客户设施的未来 Beta 部署与量产管线 | Unite.AI 与 TNW 报道 34,000 份预订单;TNW 称未验证 | 据报道约 $2.4B 管线 | 客户身份、订金、转化率和取消权 |
分群按已抓取公开来源的证据权重划分;收入、账户数和地域拆分均未公开披露。
[CU001, CU002, CU027, CU028, CU037, CU044]企业客户从 POC 和集成推进到 RaaS 生产部署;Humanoid 的公开证据仍集中在大规模续约之前。
各阶段按公开客户公告加权;周期和转化率未披露。
[CU008, CU036, CU037, CU029]6.2 交易质量:Schaeffler 与预订单簿存在实质差异
核心尽调差异不是“有客户还是没客户”,而是有约束力部署与软性销售管线的差异。Schaeffler 的交易质量最好:Humanoid 和独立报道都称 5 月协议具有约束力、分阶段推进,首批德国系统计划在 2026 年底前部署,并随时间形成四位数机器人集群。协议还给出明确运营范围——箱体处理、物流和生产规模验证——并采用 RaaS 模式,把客户价值与持续服务交付绑定。Bosch 商业上重要,但抓取来源更强调合同制造和成功 POC,而不是单独记录的有约束力 1,000 台客户订单。Siemens 和 SAP/Martur Fompak 是有说服力的技术参考,但仍属于试点或概念验证证据。34,000 个预订单和约 $2.4 billion 销售管线是有用的需求信号,但在管理层披露定金、取消权、转化率、客户身份和收入确认时间前,应打折处理。[CU003, CU004, CU005, CU008, CU011, CU012]
| 日期 / 期间 | 指标或里程碑 | 数值 | 证据质量 | 意义 | 缺失分母 |
|---|---|---|---|---|---|
| 2026-01 | Schaeffler 战略合作 | 五年内数百台机器人 | 中高:官方与 The Robot Report | 早期锚定客户 / 合作伙伴关系确立 | 1 月范围是否被 5 月有约束力协议取代 |
| 2026-03 | SAP + Martur Fompak POC | 披露一个现场生产物流 POC | 中:公司案例研究;SAP 背景为间接证据 | 汽车物流工作流在 POC 层面得到验证 | POC 时长、机队规模、付费状态和 ROI |
| 2026-04 | Siemens 工厂现场试点 | 每小时搬运 60 个周转箱;正常运行 >8 小时;抓放成功率 >90% | 试点指标可信度高:伙伴新闻稿 + Humanoid 案例研究 | 最强的量化客户结果证据 | 商业转化与跨场地可重复性 |
| 2026-05 | Schaeffler 有约束力的分阶段部署 | 四位数 / 数千台机器人;首批德国系统于 2026 年底 | 协议存在可信度高;交付台数仍属未来事项 | 超越演示阶段的实质商业牵引 | 合同金额、取消条款和验收测试 |
| 2026-05 | Bosch 制造合作 / POC | 规模化生产伙伴;POC 于 2026 年 3 月完成 | 合作可信度高;订单规模归因可信度中 | 制造路径提升可部署性 | Bosch 是否也是有约束力客户,以及对应多少台 |
| 2026-07 | 预购管线 | 34,000 份预订单;据报道约 $2.4B 管线 | 独立验证前为中低 | 需求信号与投资者叙事支撑 | 订金、客户身份、有约束力状态和预期转化 |
| Q4 2026 | 向客户设施推出 Beta | Beta 机器人计划开始推出 | 中:公司 / 媒体路线图 | 从 POC 过渡到客户现场部署 | 哪些客户接收 Beta 机器人,以及适用哪些验收标准 |
已报道里程碑混合了有约束力协议、试点、POC 和预订单;本表不意味着收入确认。
[CU005, CU016, CU021, CU022, CU023, CU024]| 交易对手 | 角色 | 垂直 / 工作流 | 状态 | 结果 / 信号 | 局限 |
|---|---|---|---|---|---|
| Schaeffler | 锚定客户 + 战略技术伙伴 + 执行器供应商 | 制造、工厂物流、部件 | 有约束力分阶段部署;首批系统 2026 年底 | 数千台 / 四位数部署与首选执行器供应 | 交付机队、收入和留存尚不可观察 |
| Bosch | 制造伙伴与 POC 交易对手 | 工业内部物流与代工制造 | 2026 年 3 月 POC 后建立规模化生产合作 | 工作流 POC 加 Bosch 制造可信度 | 公开来源无法清楚证实 Bosch 是 1,000 台付费客户 |
| Siemens | 工厂现场试点客户 / 伙伴 | 电子制造物流 | Erlangen 工厂试点 / 测试 | 每小时 60 次搬运、>8h 正常运行、抓放成功率 >90% | 未披露商业化推出或续约数据 |
| NVIDIA | 计算与物理 AI 生态伙伴 | 仿真、AI 技术栈、工业 AI | 生态 / 技术赋能方 | 支撑 Siemens 物理 AI 试点与工业 AI 技术栈 | 本身不是直接客户部署 |
| SAP | 企业工作流与具身 AI 伙伴 | 仓储 / 制造机器人业务场景 | POC 生态伙伴 | Martur Fompak POC 中,SAP AI 智能体下发任务指令 | SAP 来源未披露 Humanoid 专属付费合同条款 |
| Martur Fompak | 汽车物流 POC 用户 | 汽车生产物流 | 现场生产物流环境中的 POC | 完成 KLT 箱取回与手推车配送工作流 | 未公开持续 KPI、推出计划或采购订单 |
| 未具名预购客户 | 潜在买家 | 物流、制造、零售及其他行业 | 预购管线 | 34,000 份预订单 / 据报道约 $2.4B 管线 | 未经独立验证;有约束力状态不明 |
枚举包括商业交易对手和战略合作伙伴;状态区分有约束力部署、制造伙伴、试点、POC 和生态赋能。
[CU002, CU003, CU011, CU015, CU018, CU023]公开管线从头条预订单到具名量化试点,再到有约束力的商业部署,收窄很快。
数值是以披露数量为锚的示意指数值,不是转化率或收入。
[CU023, CU024, CU016, CU039, CU045]从数字看,醒目的预订单撑起叙事;但证据质量最高的,是 Schaeffler 那一份有约束力协议和具名试点。
条形混合了单位、计数和指数值;图中展示只用于对比证据阶段强弱,不是财务瀑布图。
[CU022, CU023, CU024, CU002, CU016, CU045]6.3 部署状态与客户结果
最高质量的运营结果证据来自 Siemens-Humanoid-NVIDIA 工厂现场试点,因为它报告了吞吐量、正常运行时间和拣放成功率。机器人在 Siemens Erlangen 以每小时 60 次搬运处理周转箱,正常运行时间超过八小时,自主拣放成功率高于 90 percent。这仍不是完整商业留存证明,但已经是一组与客户相关的运营指标。SAP/Martur Fompak 也有价值,因为它发生在现场汽车制造物流环境中,并把机器人连接到 SAP AI 智能体;但公开页面没有量化持续正常运行时间或 ROI。Bosch 同样证明了围绕传送带到手推车箱体移动的复杂工作流,但来源没有披露时长、集群规模或经济转化。Schaeffler 第一阶段时间在 2026 年底和 2027 年,当前承保必须把已签承诺与已交付集群表现分开。[CU006, CU007, CU013, CU015, CU016, CU017]
| 交易对手 / 项目 | 阶段 | 量产 / 试点 | 客户结果证据 | 下一里程碑 | 尽调要求 |
|---|---|---|---|---|---|
| Schaeffler 第一阶段 | 已签署,部署日期在未来 | 有约束力分阶段部署;生产验证于 2026 年底开始 | 用例具体,但尚无已交付机队指标 | 2026 年 12 月至 2027 年 6 月德国场地第一阶段 | 审阅已签主服务协议、场地 SOW 和验收测试 |
| Bosch POC / 制造 | POC 已完成,同时为制造伙伴 | POC 与规模化生产合作 | 内部物流展示了工作流能力 | HMND 01 规模化制造启动 | 获取产能、良率和客户订单排期 |
| Siemens Erlangen | 运营试点 / 测试 | 电子工厂试点 | 每小时搬运 60 个周转箱、>8h 正常运行、成功率 >90% | 多场地商业部署决策 | 要求 Siemens 参考客户访谈和试点后转化备忘录 |
| SAP + Martur Fompak | POC 已完成 | 现场生产物流 POC | 通过 SAP AI 智能体拣选 KLT 箱并用手推车配送 | 商业部署决策 | 要求提供 POC 报告,覆盖时长、异常、安全事件和 ROI |
| 预订单簿 | 管线 | 预购 / 预留 | 无客户层面使用证据 | 转化为已签署 Beta 部署,2026 Q4+ | 确认订金、取消权和具名交易对手 |
状态基于 2026-07-21 抓取的公开来源评估,可能滞后于私下部署进展。
[CU006, CU013, CU016, CU019, CU021, CU032]Schaeffler 交易质量最高;Siemens 量化试点结果最高;预订单验证最低。
定性评级仅使用公开证据;私人合同可能改变排名。
[CU003, CU015, CU018, CU024, CU038, CU041]6.4 留存、扩张与集中度风险
已审阅的公开来源均未披露 Humanoid 客户基础的 NRR、GRR、流失率、客户满意度、续约率、ARR、收入运行率或已确认收入。这意味着客户耐久性只能从交易对手质量、集成深度以及是否存在实测结果来推断。Schaeffler 既是优势,也是集中度风险:它是最大的公开部署、战略技术伙伴和首选执行器供应商。如果关系顺利转化,它会变成客户引用飞轮;如果生产经济不达预期,它会同时威胁客户证明、供应链可信度和投资人叙事。Bosch 也带来依赖,因为 Humanoid 的规模化依赖外部制造伙伴。开放尽调问题很直接:已签合同排期、定金、取消条款、集群验收里程碑、服务水平承诺、头部客户收入集中度、续约队列和 POC 后转化管线。[CU009, CU029, CU030, CU031, CU032, CU036]
| 指标 | 公开值 | 可信度 | 重要性 | 尽调要求 |
|---|---|---|---|---|
| 净收入留存率(NRR) | 未披露 | 低 | 判断 RaaS 部署在试点后能否扩张 | 要求按部署启动季度和客户垂直行业拆分队列 NRR |
| 总收入留存率(GRR)/ 流失 | 未披露 | 低 | 反映首个机器人场地部署后的持续性 | 要求提供流失、取消和不续约历史 |
| Schaeffler 续约 / 扩张条款 | 分阶段推出之外未披露 | 中 | 锚定客户扩张会推动参考客户飞轮,也放大集中度风险 | 审阅合同最低承诺、可选阶段和验收里程碑 |
| Bosch 重复生产承诺 | 未披露 | 中低 | 制造伙伴可靠性影响交付客户需求的能力 | 审阅制造 SLA、产能和最低量承诺 |
| 客户满意度 / 愿意接受参考访谈 | 未披露 | 低 | 早期硬件部署中,具名参考客户访谈至关重要 | 完成与 Schaeffler、Siemens、Bosch 和 Martur Fompak 的参考客户访谈 |
所有财务留存字段均为空,因为公开来源未披露队列留存、流失、NRR、GRR 或满意度指标。
[CU029, CU030, CU031, CU043, CU045]6.5 反向证据与对人形机器人部署风险的怀疑
反向来源并不否定 Humanoid 的牵引,但限制了今天能给多少信用。VaaSBlock 认为该行业已明显改善,但当前试点能力与营销暗示的自主通用人形工人之间仍有很大缺口。Physical AI Guide 区分公告、可订购性和可行性计划,与真实试点或广泛部署并不相同。IoT Digital Twin PLM 增加了方法论警告:供应商和媒体数字没有经过独立审计。The Next Web 把这个谨慎直接用于 Humanoid,指出 34,000 个预订单主张没有独立验证。这些来源支持一种纪律性折扣:Schaeffler 可视为锚定商业协议,Siemens/SAP/Bosch 可视为高质量客户试点或 POC,更广泛预订单簿在转化为有约束力部署和已确认收入前只能算销售管线。[CU022, CU024, CU025, CU033, CU034, CU035]
| 驱动因素或风险 | 当前证据 | 影响 | 缓释因素 | 尽调路径 |
|---|---|---|---|---|
| Schaeffler 多角色集中 | 最大公开协议、技术伙伴和执行器供应商 | 高:单一关系同时影响客户证明和供应链 | 若执行成功,可成为蓝筹工业参考客户 | 量化头部客户收入敞口和供应独家性 |
| Bosch 制造依赖 | POC 后 Bosch 将为欧洲制造 HMND 01 | 高:量产规模依赖外部合作伙伴 | Bosch 的工业化背书降低执行风险 | 审查产能、保修责任分配和生产成本曲线 |
| 试点到量产转化 | Siemens / SAP / Bosch 案例有说服力,但并非都披露为付费部署 | 中高:试点未必转化为 ARR | Siemens 的运营指标降低技术不确定性 | 跟踪截至 Q4 2026 概念验证(PoC)转为付费机队的比例 |
| 预订单疲软 | 34,000 份预订单 / 约 $2.4B 需求管线未获独立验证 | 高:公开管线可能高估有约束力需求 | Schaeffler 的约束性合同提供更硬的锚点 | 审查定金、取消权、客户身份和管线账龄 |
| 垂直行业集中 | 具名验证集中在汽车 / 工业制造 | 中:切入点聚焦,但多元化有限 | 深扎垂直行业有助于产品市场匹配 | 索取按垂直行业拆分的管线和前 10 大客户敞口 |
| 未披露留存 | 未找到净留存率(NRR)、总留存率(GRR)、流失、续约或满意度指标 | 高:持久性还无法写进承销假设 | 若持续跟踪,RaaS 模式可以产出经常性证据 | 要求队列仪表盘和部署后正常运行时间经济性 |
风险评级为定性评估,基于公开证据,而非未披露的收入集中度。
[CU025, CU030, CU031, CU033, CU034, CU035]6.6 图表
07风险
7.1 商业与财务风险:订单簿驱动的故事缺少收入证明
最高概率的投资风险是,Humanoid 的公开商业信号仍由订单簿驱动,而不是收入驱动。公司拥有异常强的融资标题支撑:多家独立来源报道了 $152 million Series A 轮、$1.35 billion 投后估值和 $270 million 累计融资。这些资本给团队留下执行空间,但也抬高了业绩门槛,因为估值似乎已经在公开认可收入、ARR、毛利率、现金烧钱或现金跑道证据出现前,资本化了未来机器人部署。34,000 台预订单主张具有战略重要性,但最强的反向来源明确表示,该数字没有经过独立验证,并且报道时 Humanoid 尚未交付商业单元。Schaeffler 和 Bosch 降低了进入市场风险,但也把证明集中在两个战略交易对手身上。因此,承保姿态应是“有条件上行”:只有在合同可执行性、定金、客户验收、服务经济和已交付单元利用率得到证明后,才给销售管线估值。[CR001, CR002, CR003, CR004, CR005, CR006]
| 风险 | 证据信号 | 可能性 | 影响 | 缓释成熟度 | 剩余敞口 | 投资含义 |
|---|---|---|---|---|---|---|
| 预订单 / LOI 无约束力或可取消 | 34,000 台管线规模很大,但公开层面未获独立验证。 | 高 | 高 | 在审查合同、定金、取消权和验收条款前,成熟度低。 | 管线可能大幅缩水或延后。 | 不要把 $2.4B 按收入承销。 |
| 订单额被误当作收入 | 未找到公开收入、ARR、毛利率、烧钱速度或现金跑道披露。 | 高 | 高 | 低;需要管理层数据室。 | 估值可能是期权价值,而不是收入支撑。 | 采用里程碑定价,并要求财务报表。 |
| 客户集中在 Schaeffler / Bosch | 公开商业和制造验证主要由 Schaeffler 与 Bosch 支撑。 | 高 | 高 | 中;交易对手可信,但高度集中。 | 任一合作伙伴流失或延后,都会冲击验证、供应和叙事。 | 在更广泛客户验收拿出证据前,应用集中度折价。 |
| 工业销售周期长 | 工厂部署需要 EHS、集成、采购和验收。 | 中 | 高 | 若合作伙伴牵头的试点推进,成熟度为中。 | 收入确认可能比订单额滞后数个季度甚至数年。 | 融资计划绑定已交付并验收的机器人,而不是已宣布的试点。 |
| 估值与接近零的公开收入 | 尽管没有公开收入披露,融资后估值仍达到 $1.35B。 | 高 | 高 | 缺少单位经济性时成熟度低。 | 里程碑延后可能带来下轮降估值融资或稀释风险。 | 要求价格纪律,并审查清算优先权。 |
| 硬件营运资金牵引 | 规模制造需要库存、供应商预付款、返工、保修和服务支持。 | 中 | 高 | 若 Bosch 吸收部分工厂资本开支,成熟度为中。 | 客户验收前,现金消耗可能加速。 | 建模现金转换周期和保修准备金。 |
| 炒作驱动的客户预期 | 人形机器人市场围绕工厂和家庭应用存在广泛炒作。 | 中 | 中 | 低;信息口径可控,市场周期不可控。 | 叙事倍数可能快速压缩。 | 把分阶段投资放在实际部署之后。 |
| 竞争性资本军备竞赛 | TNW 将 Figure AI 与 $39B 估值相连,Unitree 则是中国低成本参照。 | 中 | 高 | 低到中;合作伙伴策略有帮助,但不等于资产负债表规模。 | Humanoid 可能面临价格、人才和算力竞争。 | 要求差异化证据和毛利率韧性。 |
可能性和影响是基于公开证据的尽调判断;“接近零收入”指没有公开收入证据,并非确认会计收入为零。
[CR001, CR003, CR004, CR005, CR006, CR011]| 情景 | 触发模式 | 概率信号 | 风险传导 | 投资动作 |
|---|---|---|---|---|
| 乐观 | 合同验证定金、取消保护、CE 路径、beta 稳定运行,以及 Bosch/Schaeffler 的规模准备度。 | 多个部署完成验收,单台贡献为正。 | 订单额转成可信收入,估值有里程碑支撑。 | 继续尽调;若价格按里程碑调整,可考虑按比例跟投。 |
| 基准 | 销售管线仍然很大,但 beta 推出期间,合同条款、安全文件和单元经济性仍不公开。 | 部分试点扩张,但没有公开的车队经济性。 | 估值仍是期权价值,信息风险很高。 | 观察;只有拿到资料室访问权并有下行保护时才投资。 |
| 悲观 | 预订单延期、beta 验收偏弱,或安全 / 制造问题拖慢规模化。 | Q4 2026 beta 延期,或利用率低迷。 | 订单额被下修,烧钱加速,稀释 / 估值下调轮风险上升。 | 回避,或要求大幅重定价。 |
| 压力 | 地缘政治 / 出口管制问题、合作伙伴退出,或严重安全事故中断部署。 | 客户采购受阻,或监管叫停。 | 收入、合作伙伴验证和融资财团信心同时断裂。 | 暂停;除非问题已补救且获得保险覆盖。 |
由于收入、利润率、烧钱速度、合同条款和现金余额均未公开,情景分析只能定性。
[CR001, CR004, CR011, CR012, CR017, CR023]商业转化、安全 / 监管、制造规模化和估值风险,都落在残余风险最高的象限。
定性评分把公开证据映射到相对可能性和影响,不是统计损失模型。
[CR003, CR004, CR017, CR023, CR026, CR033]7.2 技术、安全与监管风险:自主主张遇上工厂安全法规
Humanoid 的产品风险不只是机器人能不能动,而是工业人形机器人能否在工人附近可靠运行,并具备客户、监管机构、保险公司和劳资委员会都接受的安全控制。公开产品页面声称最长四小时续航、15 kg 双臂载荷、运行在 NVIDIA 计算上的 KinetIQ AI,以及自主的人类速度操作。这些规格有用,但反向机器人来源说明了为什么演示不能等同于持久自主:IEEE 曾警告,机器人演示可能有人在回路中;更早的人形机器人项目也依赖工程师、系绳和缓慢任务执行。安全负担同样正式。ISO 10218、ISO/TS 15066、OSHA 机器人指南、EU Machinery Regulation 2023/1230 和 CE-marking 义务,共同形成围绕风险评估、协作运行限制、合规文件、护栏、急停、验证日志和客户现场流程的尽调路径。在认证姿态、事故历史和 beta 可靠性数据可得前,剩余安全 / 监管敞口仍然重大。[CR013, CR014, CR015, CR016, CR017, CR018]
| 规则 / 来源 | 司法辖区 | 风险状态 | 可能性 | 影响 | 缓释措施 | 剩余敞口 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| ISO 10218-1 工业机器人安全 | 全球标准 / 客户要求 | 适用于工业机器人机械安全论证 | 高 | 高 | 将 HMND 01 的危险源、控制架构、急停、验证测试和集成商责任映射到 ISO 10218-1。 | 在拿出第三方安全论证和集成商签字前,敞口重大。 | 索取安全文件、风险评估、测试证据和集成商验收文件。 |
| ISO/TS 15066 协作机器人 | 全球技术规范 | 机器人靠近人员或以协作模式运行时适用 | 高 | 高 | 定义协作运行模式、力 / 功率限制、间距监测和现场任务边界。 | 人形机器人在按人类设计的空间里作业,因此敞口重大。 | 索取协作运行风险评估、防护设计和工人试运行日志。 |
| 欧盟《机械法规》2023/1230 | 欧盟 / EEA | 新机械法律框架影响机器投放欧盟市场 | 中 | 高 | 维护合格评定计划、基本健康与安全要求矩阵和技术文档。 | CE 路径拿出证据前,欧洲工厂部署敞口高。 | 索取 CE 策略、欧盟符合性声明草案,以及适用时的公告机构意见。 |
| CE 标志义务 | 欧洲经济区 | 制造商必须完成评定、技术文件、声明和 CE 标志 | 中 | 高 | 在 Humanoid、Bosch、Schaeffler 和客户之间分配制造商 / 进口商 / 分销商职责。 | 若责任在多个合作伙伴之间拆分,敞口重大。 | 审查供应和部署合同中的 CE 责任分配。 |
| OSHA 机器人工作场所危害 | 美国客户现场 | OSHA 没有机器人专门标准,但机器人危害识别和机器防护仍适用 | 中 | 高 | 现场危害分析、防护、LOTO、培训、事故报告和急停程序。 | 美国试点的工作场所安全责任落在部署环节,因此敞口重大。 | 索取美国试点安全计划和客户 EHS 签字。 |
| 出口管制、视同出口、制裁观感 | 美国、英国、欧盟、跨境团队 | 受控技术和涉俄观感可能影响算力、代码、招聘和客户 | 中 | 高 | 出口分类、受限方筛查、视同出口审查和治理记录。 | 合规流程形成文件前,敞口重大。 | 索取出口管制备忘录、制裁筛查、客户国家清单和董事会监督纪要。 |
列举并不完整:仅覆盖公开证据中与工业人形机器人部署最相关的法律和安全制度,并非针对每个目标国家的完整法律意见。
[CR018, CR019, CR020, CR021, CR022, CR023]| 失效模式 | 证据信号 | 可能性 | 影响 | 缓释成熟度 | 剩余敞口 | 未解决缺口 |
|---|---|---|---|---|---|---|
| 自主能力在演示之外表现不及预期 | IEEE 和持怀疑态度的来源提示,演示可能包含人在环支持。 | 高 | 高 | 未知;Humanoid 声称有 KinetIQ,但未发布机队自主指标。 | 在 beta 机器人无需远程操作运行前,敞口高。 | 索取自主脱离、人工介入和远程操作日志。 |
| 规模化可靠性不及预期 | 可比人形机器人过往常见系绳、工程师陪跑和执行缓慢。 | 高 | 高 | 公开证据低;据称有 9 个部署,但日志未公开。 | 在拿出多班次正常运行时间和 MTBF 数据前,敞口高。 | 按试点索取正常运行时间、故障、维护和客户验收数据。 |
| 续航限制利用率 | 官网称 Alpha Wheeled 续航最高 4 小时。 | 中 | 中 | 未知;可换电池 / 充电方案未公开。 | 每项任务可能需要更多机器人,或接受停机时间。 | 索取任务级占空比和充电工作流。 |
| 工人附近作业的安全性未通过验收 | 工业人形机器人靠近人员作业,需要 ISO / OSHA / 欧盟安全控制。 | 中 | 高 | 未知;认证状态未公开。 | 在审查 CE 文件和客户 EHS 签字前,敞口高。 | 索取安全论证、事故历史和现场风险评估。 |
| 网络 / 机队管理敞口 | Schaeffler 协议包括机队软件和 24/7 支持,意味着联网机队运营。 | 中 | 中 | 未知;未公开找到信任 / 安全页面。 | 若机器人接入客户工厂系统,敞口重大。 | 索取安全架构、SBOM、SOC2 / ISO27001 状态和事故流程。 |
| 保修和现场服务负担 | 硬件机器人需要维护、更新和持续性能管理。 | 中 | 高 | 若 Bosch / Schaeffler 支持规模化,成熟度为中,但责任不清。 | 早期队列可能吃掉利润率。 | 审查服务级别协议和保修准备金模型。 |
| Beta 推出延后 | Beta 机器人计划在 Q4 2026 推出,但安全、生产和客户验收在公开层面仍未证实。 | 中 | 高 | 公开证据低。 | 延迟会冲击收入转化和融资叙事。 | 跟踪 beta 发货日期、验收率和前 3 大客户扩张。 |
运营风险行综合公司声明、独立机器人质疑和标准义务;缺失遥测被视为尽调缺口,而不是失败证据。
[CR013, CR014, CR015, CR016, CR017, CR018]7.3 执行、依赖与竞争风险:伙伴规模是优势,也是卡点
Humanoid 的规模化模式刻意依赖战略伙伴,而不是把生产每一层都握在自己手里。这样可能更省资本,尤其是 Bosch 承担工业化、Schaeffler 供应执行器时;但风险也被推到供应商产能分配、质量追责、保修协同、工程变更控制和交易对手优先级上。Schaeffler 已公开表示,其到 2031 年期间将承担超过 50% 执行器需求的优先供应商角色,这是明确的集中点。NVIDIA 级算力是另一项依赖:机器人平台的 AI 技术栈和更广的人形机器人生态都靠专用加速器,其供给、路线图和出口分类都会影响业务。BIS 关于视同出口、出口合规计划和再出口的指引说明,算力和受控技术不能当作普通零部件处理。竞争压力还会叠加:Unitree 提供了一个可见的中国低成本参照,Tesla 和 Figure 影响客户预期,TNW 提到 Figure 的 $39 billion 估值,也凸显资本差距。因此,要走向量产,Humanoid 必须拿出供应冗余、良率、现场服务和出口管制合规基本功的证据。[CR007, CR008, CR009, CR010, CR024, CR025]
| 依赖项 | 交易对手 | 角色 | 集中度 | 失效情景 | 严重性 | 缓释措施 | 剩余敞口 |
|---|---|---|---|---|---|---|---|
| 执行器 | Schaeffler | 到 2031 年覆盖 >50% 需求的优先供应商 | 高 | 供应短缺、涨价或质量问题拖慢机器人制造。 | 高 | 为执行器导入第二来源,并为关键关节保留安全库存。 | 在替代执行器完成认证前,敞口重大。 |
| 合同制造 | Bosch | 规模制造和工业化合作伙伴 | 高 | Bosch 优先级、良率问题或合同范围变化拖慢生产。 | 高 | 明确质量归属、良率关口、工装责任和产量承诺。 | 在审查主制造协议前,敞口重大。 |
| 算力平台 | NVIDIA Jetson / 物理 AI 技术栈 | 支撑 KinetIQ 和人形机器人工作流的边缘 AI 算力 | 高 | 配额、路线图延迟、出口分类或成本飙升会约束制造。 | 高 | 认证 BOM 替代方案并锁定供应;分类出口敞口。 | AI 算力是产品承诺的核心,因此敞口重大。 |
| 战略验证客户 | Schaeffler、Bosch、Siemens、SAP 与 Martur Fompak | 客户验证、试点和部署可信度 | 高 | 一个高曝光试点失败或延后,就会影响所有下游销售。 | 高 | 扩大具名客户基础,并发布结果指标。 | 客户多元化改善前,敞口高。 |
| 监管验收 | 欧盟 / OSHA / 客户 EHS | 机器人靠近人员作业的许可和验收 | 中 | 安全审批或客户职工委员会验收拖慢铺开。 | 高 | 必要时,提前与客户和监管方确认安全论证。 | 欧洲工厂敞口重大。 |
| 出资方信心 | Series A 投资人 / 未来银团 | 为硬件规模化和营运资金提供资金 | 中 | 市场风险或错失里程碑会推高稀释或下轮降估值风险。 | 高 | 基于里程碑的融资计划和透明 KPI 报告。 | 若订单额到 2027 年仍未转化,敞口高。 |
集中度仅使用公开证据;合同排他性、最低采购量和终止权未能审查。
[CR002, CR006, CR007, CR008, CR009, CR010]技术和合作伙伴失败会传导到交付、收入确认、烧钱速度、融资条款和估值。
边代表基于公开风险证据形成的因果尽调假设,不是实测弹性。
[CR012, CR023, CR026, CR043, CR044, CR047]后果最重的依赖集中在 Schaeffler 执行器、Bosch 制造、NVIDIA 算力、安全监管和后续资本。
图反映公开依赖;实际合同集中度需要查看私下协议。
[CR006, CR007, CR008, CR024, CR026, CR028]7.4 治理、地缘政治观感、市场周期风险与缓释
治理风险不来自某一条负面指控,而在于缺少通常用来降低关键人、股权结构、合规和执行不确定性的公开证据。公开报道把故事集中在创始人兼 CEO Artem Sokolov,以及公司两年内从想法冲到独角兽的速度上。Forbes 2 月的标题用「前俄罗斯珠宝大亨」的视角包装他的背景;即便没有指称不当行为,制裁、出口管制、客户采购和外部观感尽调也因此更重要。本报告审阅的公开记录没有披露董事会构成、优先权结构、关键人保险、继任计划、制裁筛查或出口管制治理。市场周期风险同样真实:Ars、Vox、LA Times、Compare Robots 和 RoboHorizon 都提供了对人形机器人热潮、演示落差和部署难度的负面或怀疑背景。实际缓释办法是把叙事里程碑改成证据里程碑:已签且可执行的合同、安全文件、Beta 正常运行时间、生产良率、供应商第二来源、治理文件和现金续航。任何一项不过关,都应触发价格重置或停止条件。[CR035, CR036, CR037, CR038, CR039, CR040]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 影响 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 创始人兼 CEO | 叙事和执行故事高度围绕 Artem Sokolov。 | 中 | 高 | 关键人计划、授权运营节奏、董事会监督和接班梯队。 | 访谈 CEO / CTO / COO 等角色,并审查董事会纪要。 |
| 技术领导层 | 公开团队页面覆盖面广,但未披露留存、薪酬或关键人映射。 | 中 | 中 | 留任方案、关键角色图谱,以及与 beta 里程碑绑定的招聘计划。 | 索取组织架构图、离职率、期权刷新计划和开放岗位。 |
| 股权结构表 / 优先权 | 未公开披露优先股堆叠、期权池、董事会或 side letter 条款。 | 高 | 高 | 完整股权结构表、投资人权利、清算优先权和关联方审查。 | 为任何新投资定价前,先取得法律文件。 |
| 地缘政治观感 | Forbes 对创始人的表述叠加俄罗斯制裁 / 出口管制制度,可能让客户采购复杂化。 | 中 | 中 | 制裁筛查、出口管制政策、KYC 和客户沟通计划。 | 开展受限方检查,并审查合规备忘录。 |
| 治理控制 | 没有公开证据显示其具备出口管制治理、安全委员会或事故升级流程。 | 中 | 高 | 董事会层面的风险委员会、EHS 负责人、出口管制官和事故报告机制。 | 索取治理章程和保险覆盖。 |
治理风险被界定为证据缺失和观感敞口,而不是不当行为指控。
[CR035, CR036, CR037, CR038, CR039, CR030]| 风险 | 可监控触发项 | 阈值 / 事件 | 动作含义 |
|---|---|---|---|
| 预订单转化 | 具名客户签署的合同和定金 | 资料室中可执行条款支持的名义订单额低于 50% | 对订单额打折,或暂停投资。 |
| 收入质量 | 已交付并验收的机器人 | 计划 beta 启动后两个季度仍无已验收的商业部署 | 将估值视为过度透支。 |
| 安全认证 | CE / 安全文件完整性 | 没有针对欧盟场址的风险评估、EHS 签批或合规计划文件 | 不再承销部署假设。 |
| 自主性成熟度 | 遥操作 / 干预率 | 关键任务需要约定安全监控之外的隐性人工协助 | 重置产品就绪度假设。 |
| 可靠性 | 车队正常运行时间和 MTBF | beta 客户现场任务成功率低于 90%,或多班次正常运行表现差 | 后续跟投前要求补救计划。 |
| 制造良率 | 试生产良率和返工 | 连续两个生产批次良率或返工未达计划 | 上调现金需求,推迟收入确认。 |
| 供应商集中度 | Schaeffler 执行器或 NVIDIA 算力配额 | 规模化生产前没有合格第二来源或配额协议 | 计入供应链风险折扣。 |
| 出口管制 | 分类与筛查 | 针对算力、源代码和员工,没有形成 EAR / UK 军民两用 / 出口管制流程文件 | 投资前要求完成合规补救。 |
| 治理 | 股权结构表和董事会权利 | 管理层无法提供股权结构表、优先权条款、董事会会议纪要和关键人计划 | 不对新资金出价。 |
| 市场炒作 | 可比公司情绪和融资市场 | 里程碑滑坡时,人形机器人同行估值下修,或硬件机器人融资窗口关闭 | 使用估值下调轮条款,或回避。 |
否决标准是尽调触发器,不是最终法律或经营结论。
[CR004, CR011, CR012, CR018, CR023, CR026]| 时间窗口 | 缓释工作线 | 负责人 / 证据 | 目标材料 | 延误后的剩余风险 |
|---|---|---|---|---|
| 0-30 days | 合同与股权结构表审计 | CFO / 法律顾问 / 投资人 | 已签订单明细、取消条款、股权结构表、优先股堆叠 | 订单额和估值仍无法支撑投资判断。 |
| 0-60 days | 安全与 CE 准备度 | 安全负责人 / 监管法律顾问 | ISO 10218/15066 风险评估、CE 技术文件计划、EHS 签批模板 | 欧盟部署延迟或责任风险。 |
| 30-90 days | 可靠性与自主性验证 | CTO / 部署负责人 | 遥操作日志、正常运行时间、故障、客户验收、beta 测试计划 | 产品就绪度说法仍停留在个案层面。 |
| 60-120 days | 供应链与出口管制控制 | COO / 合规负责人 | BOM 配额、第二来源计划、EAR / UK 出口分类、筛查记录 | 规模化时出现制造或合规瓶颈。 |
| 到 Q4 2026 beta | 投资人里程碑复核 | 董事会 / 领投方 | 已交付 beta 机器人、已验收客户工作流、单元经济性、现金跑道 | 后续融资面临稀释或估值下调轮风险。 |
时间表采用公开 beta 时间点和风险控制顺序;具体负责人是建议的尽调负责人,并非 Humanoid 已确认岗位。
[CR006, CR008, CR011, CR013, CR018, CR021]7.5 附录
08估值
8.1 价格敏感结论:跟踪,不能盲目投资测算
Humanoid 以 $1.35 billion 投后估值完成 $152 million Series A,可信度足以进入投委会观察名单,但公开收入支撑还不够,不能直接买入。支撑价格的是一套宏大战略叙事:欧洲第一家专注人形机器人的独角兽、累计 $270 million 资本基础、Schaeffler 与 Bosch 关系,以及据报 $2.4 billion 的订单簿。反方同样重要。公开记录基本没有披露已确认收入和 ARR,预订单数字应视为订单潜力,而不是可入账收入。也就是说,估值赌的是转化、交付、正常运行时间和制造规模,而不是今天财报的倍数。当前价格下,实际立场应是跟踪或继续研究;只有在合同可执行性、定金、贡献毛利和实体结构得到验证之后,才明确愿意承做投资测算。因此,本估值章节把每个头条都当作有条件输入:融资确认投资人愿意下注,合作关系确认产业兴趣,同业估值确认品类热度,但这些事实单独都不能证明收入可收回或现金流耐久。投资测算标准应是:客户接受可工作的机器人,在可执行条款下付款,且扣除制造、支持和保修成本后仍留下足够毛利,支撑下一轮部署。[CV001, CV002, CV005, CV008, CV015, CV038]
| 决策项 | 评估 | 证据基础 | 投资含义 |
|---|---|---|---|
| 投资建议 | 观察 / 继续研究 | Series A 和合作伙伴真实存在,但收入和利润率未披露 | 没有私下合同和单元经济性尽调,不要承销新的买入 |
| 置信度 | 中低 | 融资和同行估值标记公开;经营财务仍是私人信息 | 用里程碑约束入场 |
| 风险评级 | 高 | 硬件规模化、供应链、RaaS 融资、客户验收和价格压缩 | 要求下行保护,或等待验证 |
| 估值立场 | 偏高 | 0.56x 订单额看似合理,但公开披露的已确认收入为空 | 只有订单转化且利润率为正,估值才算合理 |
| 入场纪律 | 以里程碑为门槛 | Q4 2026 beta 推出、Schaeffler/Bosch 验收、实体与优先权审查 | 现在观察;拿到交付证据后再复盘 |
评估仅使用公开来源;已确认收入、现金余额和优先权条款仍属私人信息。
[CV001, CV002, CV005, CV008, CV015, CV038]| 论点 | 支撑估值 | 什么会改变判断 |
|---|---|---|
| 欧洲先发独角兽信号 | 欧洲最大、聚焦人形机器人的 Series A 验证了投资人胃口 | 轮次条款或优先股堆叠显著差于普通股 |
| 战略合作伙伴 | Schaeffler 和 Bosch 支撑需求、执行器和制造路径 | 合同被证实无约束力或仅限试点 |
| 订单额口径 | 34,000 份预订单和 $2.4B 订单簿带来很大的转化期权 | 定金很小、可取消,或高度集中 |
| 同行估值膨胀 | Figure 和 Apptronik 的估值标记让 $1.35B 在品类内显得温和 | 同行估值标记压缩,或无法转化为部署 |
| 市场 TAM | Goldman、Morgan Stanley 和 MarketsandMarkets 勾勒出庞大的长期市场 | 工业买家推迟采用,或安全约束拖慢部署 |
| 收入质量 | 若正常运行时间高,RaaS 可能形成经常性经济性 | 车队融资和现场服务吃掉贡献毛利 |
| 价格压缩 | Unitree 价格锚挑战高端 ASP 假设 | Humanoid 证明工业可靠性和集成价值足以支撑溢价 |
| 证据质量 | 官方和高可信来源相互印证融资 | 私下尽调无法验证实体、收入或客户条款 |
论点有意成对呈现投资逻辑与反向逻辑,因为估值对里程碑敏感。
[CV001, CV003, CV005, CV011, CV017, CV020]低 / 基准 / 高三档估值区间,把 Series A 视作围绕当前 $1.35B 价格的里程碑期权。
这些区间是基于公开同业估值标记和已披露 Series A 价格做的情景交叉校验,不是正式 DCF。
[CV037, CV038, CV039]Humanoid 在市场和战略证明上得分较高,但收入质量和证据完整度偏弱。
评分是基于公开证据的 IC 启发式打分,尺度 0-10;不是统计模型。
[CV030, CV038, CV039, CV040, CV050]8.2 可比公司和隐含倍数显示估值偏紧,但并非不理性
同业样本解释了为什么这个估值既偏紧,也说得过去。Figure 的 $39 billion Series C 和 Apptronik 据报超过 $5 billion 的估值,把人形机器人推成资本军备竞赛品类;放在这些标尺旁边,Humanoid 的 $1.35 billion 显得不算夸张。但从运营证据看,画面就没那么宽松。Figure 有 BMW 生产证据,Apptronik 资本基础大得多,Agility 已有真实物流部署。Humanoid 最强的公开指标是据报 $2.4 billion 的预订单簿,对应估值约为订单额的 0.56 倍、每份预订单约 $39,700。只有订单能转化,这些数字才显得便宜;如果订单可取消、RaaS 需要大量车队融资,或 Unitree 式定价在 Humanoid 起量前拉低硬件 ASP,这些数字就显得昂贵。[CV006, CV007, CV016, CV017, CV018, CV019]
| 可比对象 | 2026 年估值 / 融资信号 | 商业验证 | 对 Humanoid 的意义 | 局限性 |
|---|---|---|---|---|
| Humanoid | $1.35B 投后估值;$152M Series A;累计融资 $270M | 34,000 份预订单;Schaeffler 和 Bosch 合作关系 | 目标公司估值锚 | 无已确认收入或利润率披露 |
| Figure AI | Series C 投后估值 $39B;承诺资本超过 $1B | BMW 生产与物流场景中的 Helix 证据 | 人形机器人品类的上行可比对象 | 资本基础大得多,美国平台能见度更高 |
| Apptronik | 超过 $935M Series A;据报道投后估值约 $5.3B | Apollo 借 Google/Mercedes/John Deere 投资人网络扩张 | 更接近的私募轮可比对象 | 公开收入披露仍有限 |
| Agility Robotics | S-4 流程,SPAC 路径预计总募资 >$620M | GXO 多年物流协议 | 商业部署基准 | 已审阅来源未清晰披露公开估值 |
| 1X | 报道称 NEO 以 $20,000 或订阅方式提供 | 消费者预订单模式 | 价格与打包信号 | 家庭用例不同于工业 RaaS |
| Sanctuary AI | 已审阅首页未披露当前估值 | 工业具身 AI 定位 | 定性工业可比对象 | 公开估值数据不足 |
| Unitree | G1 标价 $13.5K 起;Walmart 相关报道引用 $21.6K 标价 | 低成本硬件可得 | 对 ASP 和商品化形成反向压力 | 开发者 / 消费者硬件不同于工业服务机器人 |
| Tesla Optimus | 无法从 Tesla 上市公司估值中拆分 | Tesla 内部战略项目 | 战略威胁,也是招聘 / 注意力可比对象 | 无法推导干净的私人人形机器人倍数 |
| Humanoid 市场 TAM | Goldman 预计到 2035 年 $38B;MarketsandMarkets 预计到 2035 年 $50.27B | 品类整体采用预测 | 退出上限与市场拉动背景 | TAM 不是公司收入 |
| 融资数据平台 | PitchBook / Dealroom / Crunchbase / CB Insights 档案覆盖 | 档案与生态发现 | 未来刷新时的交叉核对来源 | 公开页面不能替代私人股权结构表 |
| UBTECH | 已审阅来源无公开估值 | 商业机器人和人形机器人组合 | 与中国相关的竞争广度 | 不是私募轮估值可比对象 |
列举的是基于 runDate 可得公开来源的代表性同行 / 市场样本;私人二级市场估值标记可能缺失。
[CV001, CV017, CV018, CV019, CV020, CV021]| 指标 | 公式 | 隐含值 | 解读 |
|---|---|---|---|
| 投后估值 | 已披露 Series A 投后估值 | $1.35B | 当前价格锚 |
| 累计融资 | 已披露累计资本 | $270M | 现金实力与投资人背书的代理指标 |
| Series A 前资本 | 累计融资减去 Series A | $118M | 既有资本结构的代理指标 |
| 估值 / 累计融资 | $1.35B / $270M | 5.0x | 相比前沿 AI 同行并不过高,但不是经营验证 |
| 估值 / 订单额 | $1.35B / $2.4B | 0.56x | 只有订单额转化,表面吸引力才成立 |
| 每份预订单对应订单额 | $2.4B / 34,000 | ~$70.6K | 折扣和取消前的名义 ASP 代理 |
| 每份预订单对应估值 | $1.35B / 34,000 | ~$39.7K | 看起来低于订单 ASP,但忽略毛利率 |
| 每名工程师对应估值 | $1.35B / 160–200 名工程师 | ~$6.75M–$8.44M | 高人力资本期权价值 |
| 收入倍数 | 估值 / 已确认收入 | 无意义 | 已确认收入未公开披露 |
所有倍数均由公开数字推导;订单额不是已确认收入,也不是 ARR。
[CV001, CV002, CV003, CV004, CV005, CV006]| 检查点 | 公开事实 | 估值读数 | 尽调注意事项 |
|---|---|---|---|
| 2024 年成立 | Humanoid 成立于 2024 年 | 快速进入独角兽价格区间 | 公司年轻,执行不确定性更高 |
| 既有资本 | Series A 前约 $118M | A 轮前融资基础不小 | 轮次构成和优先权未公开 |
| Series A | $152M,投后估值 $1.35B | 新投资人价格验证品类需求 | 优先股堆叠和期权池未披露 |
| 累计融资 | 累计资本 $270M | 支撑产品和部署的现金跑道 | 烧钱速度、债务和营运资本未披露 |
跳升分析根据已披露累计融资和 Series A 规模推断此前资本;不推断股价或所有权。
[CV001, CV002, CV003, CV004, CV015, CV017]按名义估值,Humanoid 远低于 Figure 和 Apptronik,但已披露融资额和部署证明也落后于它们。
Agility 数值是总募集额参考,不是企业价值;价格锚是硬件价格,不是公司估值。
[CV017, CV020, CV022, CV023, CV024, CV047]8.3 情景估值是里程碑期权,不是传统 DCF
DCF 式框架仍然有用,但应作为情景纪律,而不是制造虚假的精确感。乐观情景给 Humanoid 的公开 2031 年 100,000 台目标、强 RaaS 附着和可信的工业软件 / 服务层计入价值;如果客户接受度和毛利被证明成立,数十亿美元级结果可以成立。基准情景把本轮价格作为锚点,因为公开事实验证了战略需求,但没有验证出货经济性。悲观情景承认,硬件延迟、保修成本、低利用率和价格压缩,能很快把一个庞大的预订单数字变成下一轮降价融资。关键驱动不是 TAM:Goldman、Morgan Stanley 和 MarketsandMarkets 都支持长期人形机器人市场很大。真正关键的是,在下一轮融资窗口前,Humanoid 能交付、能收款、可支持且具备正贡献毛利的机器人份额。[CV027, CV028, CV029, CV030, CV031, CV034]
| 情景 | 2031 年经营假设 | 估值逻辑 | 指示性价值 | 概率信号 / 触发项 |
|---|---|---|---|---|
| 乐观 | 交付或签约 50,000–100,000 台机器人;RaaS 附着强;工业正常运行时间已验证 | 相对 Figure 打折的品类领导者,但较当前轮次有溢价 | ~$4.0B 高情景 | Schaeffler/Bosch 转化、毛利率验证、客户重复下单 |
| 基准 | 累计 10,000–25,000 台机器人;部分订单额转化;硬件 / RaaS 经济性混合 | 当前轮价格仍是合理期权价值 | ~$1.35B 基准情景 | 里程碑达成,但利润率和流失率仍未公开 |
| 悲观 | 已验收机器人 <5,000 台;取消率高;ASP 压缩;beta 推出延迟 | 与 IP 和人才绑定的估值下调轮期权价值 | ~$0.5B 低情景 | Q4 2026 beta 失败、定金薄弱、保修成本高 |
情景价值是指示性交叉核对,不是 DCF;公开来源未披露单台利润率、流失率或 RaaS 利用率。
[CV011, CV012, CV024, CV031, CV034, CV035]| 驱动因素 | 下行情景输入 | 基准情景输入 | 上行情景输入 | 估值影响 |
|---|---|---|---|---|
| 预订单转化 | 10% | 35% | 70% | 最大波动因素:预订不等于收入 |
| 平均实际 ASP | $35K | $70K | $100K | 影响收入池和价格压缩风险 |
| 扣除保修后的毛利率 | 为负 | 15% | 35% | 决定规模化是吞噬现金还是产生现金 |
| RaaS 利用率 | 单班低利用 | 混合班次 | 多班高利用 | 左右经常性经济性和回本周期 |
| 交付时点 | 延至 2028+ | 2027-2028 年爬坡 | 2026 年 beta,2027 年部署 | 改变资本需求和客户信心 |
| 同业倍数环境 | 机器人下轮降估值行情 | 当前私募估值标记 | Figure / Apptronik 式溢价 | 左右退出和下一轮估值 |
| 资本密集度 | 车队融资不可得 | 仅有战略支持 | 资产融资可得 | 决定规模化前的稀释 |
| Unitree 式 ASP 压力 | 严重 | 中等 | 靠工业集成消化 | 检验溢价防守 |
| 实体与合同清晰度 | 未解决 | 部分核验 | 完全核验 | 决定 IC 能否下投资判断 |
敏感性区间是根据公开证据缺口推导的承做假设,并非管理层指引。
[CV024, CV031, CV034, CV035, CV036, CV042]里程碑表现可能让 Humanoid 落入下轮降价融资风险,也可能打开数十亿美元上行空间。
瀑布图增量是说明方向的估值示例,不是经审计模型输出。
[CV034, CV035, CV036, CV046]只有订单转化和毛利证明一起改善,估值才算公允。
定性矩阵把公开证据缺口映射到估值立场;毛利率和转化率未披露。
[CV031, CV036, CV045, CV049]8.4 尽调要求与终止触发条件
下一轮尽调应以文件为主。媒体报道可以确认本轮融资、同业估值和合作伙伴公告,但回答不了优先权结构、客户取消权、定金覆盖、制造成本、保修敞口或运营实体对账。第一项投资条件,是证明 Schaeffler、Bosch 和更广泛的预订单关系有足够约束力,能支撑生产融资。第二项是证明 HMND 01 能以可接受的正常运行时间跑付费工作流,且现场服务成本低于 RaaS 贡献毛利。第三项是价格韧性:Unitree 的低标价和中国相关供应链意味着,Humanoid 必须守住工业可靠性、安全和集成价值,而不能依赖硬件稀缺。如果 Beta 部署延误、订单可执行性不及预期,或下一里程碑后贡献毛利仍不披露,正确动作就是等更低入场价,或直接放弃。[CV022, CV024, CV031, CV032, CV040, CV041]
| 触发项或要求 | 所需门槛 / 证据 | 如何传导至投资逻辑 | 行动含义 |
|---|---|---|---|
| 合同可执行性 | Schaeffler / Bosch / 订单簿条款、定金、取消权的签署文件 | 把新闻口径预订转为可承做的积压订单 | 仅在文件审查后推进 |
| Beta 推出 | 2026 年 Q4 beta 机器人交付并获工业用户验收 | 验证产品时点和客户接受度 | 延误触发估值折扣 |
| 单位经济性 | BOM、制造成本、保修准备金、毛利率、服务成本 | 判断 RaaS 和硬件规模化是否产生现金 | 没有毛利桥就不买入 |
| RaaS 融资 | 车队所有权、租赁 / SPV 条款、利用率、正常运行时间 SLA | 可能带来经常性收入,也可能变成资本消耗 | 要求融资方案 |
| 实体结构 | 运营实体、IP 所有权、Companies House 核对 | 影响法律所有权和股权结构表尽调 | 解决前为阻断项 |
| 优先股堆叠 | 清算优先权、期权池、反稀释、战略权利 | 决定经济入场价格 | 调整估值或放弃 |
| 竞争定价 | Humanoid 能相对 Unitree 低成本硬件守住溢价的证据 | 守住 ASP 和贡献利润率 | ASP 坍塌则下调估值 |
| 同业估值标记压缩 | Figure / Apptronik / Agility 估值重置或部署失败 | 改变赛道退出倍数 | 入场前重新定价 |
这些尽调要求按其把新闻支撑的期权转化为可投资证券的能力排序。
[CV015, CV024, CV031, CV042, CV043, CV044]8.5 附录
免责声明
本报告仅基于截至 2026-07-21 可获得的公开信息做研究综合,不构成投资建议。Humanoid 是私营公司;大量财务和合同事实未披露,报告已标为尽调缺口。预订单、估值等数字来自公司表述和媒体报道,无法独立审计。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Humanoid, branded HMND, is a UK AI and robotics company building industrial humanoid robots for commercial deployment. | 高 | SO001, SO010 |
| CO002 | The company positions its model as commercially scalable humanoid labour for industrial workflows rather than a research-only robotics program. | 高 | SO001, SO005 |
| CO003 | Humanoid is headquartered in London and publicly describes engineering satellites or offices in Boston and Vancouver. | 中 | SO007, SO018 |
| CO004 | Humanoid was founded in 2024 by Artem Sokolov; one public founder profile specifies May 2024. | 高 | SO007, SO015 |
| CO005 | The HMND 01 Alpha product family has wheeled and bipedal variants aimed first at industrial, logistics, manufacturing, retail, and related settings. | 高 | SO001, SO005, SO006, SO022 |
| CO006 | The wheeled HMND 01 Alpha is listed at 220 cm, about 300 kg, 29 degrees of freedom, up to 2 m/s, about four hours runtime, and 15 kg payload. | 高 | SO001, SO005, SO022 |
| CO007 | The bipedal HMND 01 Alpha is described by independent product summaries as 179 cm, 90 kg, 1.5 m/s, three-hour runtime, 29 DoF, and 15 kg payload, while the official page uses some shared Alpha specs. | 中 | SO006, SO022 |
| CO008 | KinetIQ is Humanoid’s proprietary AI framework or brain for end-to-end humanoid fleet orchestration and trial-and-error learning. | 高 | SO001, SO004, SO005 |
| CO009 | Artem Sokolov is Founder and CEO, with public profiles tying him to manufacturing, investment, and a family-business background before Humanoid. | 高 | SO015, SO016, SO019 |
| CO010 | Forbes framed Sokolov as a former Russian jewelry magnate in a February 2026 fundraising story, making founder-background and geopolitical perception a diligence topic. | 中 | SO014 |
| CO011 | Humanoid appointed Jarad Cannon as CTO in May 2025, and coverage says he brought more than 13 years of robotics and AI execution experience. | 中 | SO017 |
| CO012 | Public team-size evidence is inconsistent: company partnership pages say over 200 engineers, a founder interview says over 130 professionals, and funding coverage says more than 250 employees. | 中 | SO007, SO016, SO012 |
| CO013 | For cover purposes this chapter carries the canonical headcount band as approximately 160–200, with a gap noting higher and lower public reports. | 中 | SO007, SO016, SO012 |
| CO014 | Humanoid says its team includes engineers, researchers, and innovators from top global technology companies, but the public pages reviewed do not provide a complete roster by prior employer. | 中 | SO007, SO016 |
| CO015 | Humanoid announced a $152 million Series A financing on July 21, 2026. | 高 | SO010, SO012, SO020 |
| CO016 | The Series A was reported at a $1.35 billion post-money valuation. | 高 | SO010, SO012, SO020 |
| CO017 | The Series A brought total capital raised to date to $270 million. | 高 | SO010, SO012, SO021 |
| CO018 | The $270 million total and $152 million Series A imply approximately $118 million of prior rounds or pre-Series-A capital. | 高 | SO010, SO012 |
| CO019 | Prime Movers Lab led the Series A, with Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures named as participating investors. | 高 | SO010, SO012, SO020 |
| CO020 | Silicon Republic identifies Aglaé Ventures as the investment firm of LVMH chairman Bernard Arnault, adding strategic brand-capital signal to the investor list. | 中 | SO012 |
| CO021 | Schaeffler signed a binding phased deployment and supply agreement for a four-digit number of wheeled units across global facilities, with German production deployments beginning in late 2026. | 高 | SO007, SO010 |
| CO022 | Bosch is reported as a contract-manufacturing partner providing strategic consulting and technical expertise in hardware design, production, and supply chain. | 高 | SO012, SO021 |
| CO023 | Phemex reported a 1,000-robot Schaeffler order and Bosch-supported production plan for 100,000 units, while official and tech.eu materials emphasize thousands of Schaeffler robots rather than a Bosch order count. | 中 | SO020, SO007, SO010 |
| CO024 | Company materials name SAP, NVIDIA, Bosch, and Siemens among customers, builders, or partners around the HMND 01 Alpha industrial proof points. | 高 | SO001, SO008, SO009, SO012 |
| CO025 | The company and SAP/Martur Fompak completed an automotive manufacturing logistics proof of concept in a live production environment on March 30, 2026. | 高 | SO009, SO001 |
| CO026 | Humanoid, Siemens, and NVIDIA announced a physical-AI factory-floor deployment at Siemens Erlangen that used Siemens industrial integration and NVIDIA simulation tooling. | 高 | SO008, SO001 |
| CO027 | Tech.eu states the financing proceeds are intended to support Q4 2026 beta-version robot rollout, mass manufacturing of wheeled robots, and KinetIQ development. | 中 | SO010 |
| CO028 | The Siemens/NVIDIA announcement says virtual simulation helped cut prototype development from a typical 18–24 months to seven months. | 中 | SO008, SO019 |
| CO029 | Tech.eu quotes Sokolov saying Humanoid became Europe’s first pure-play humanoid robotics unicorn in two years. | 高 | SO010, SO012 |
| CO030 | Silicon Republic described the financing as the largest-ever Series A for a humanoid-first robotics company in Europe, according to Humanoid. | 中 | SO012 |
| CO031 | The canonical report input uses 34,000 pre-orders worth about $2.4 billion as the cover customer/pre-order figure, but this chapter could not corroborate that exact count in fetched public pages. | 中 | SO024, SO010 |
| CO032 | CNET reported around 25,000 HMND 01 preorders and six Fortune 500 pilots based on Sokolov’s comments, creating a material variance against the 34,000 canonical baseline. | 中 | SO024 |
| CO033 | Revenue run-rate, ARR, gross margin, burn, runway, board seats, and financing preference terms were not disclosed in the public sources reviewed for this chapter. | 中 | SO010, SO012, SO026 |
| CO034 | Dealroom provides a third-party company profile and investor/funding surface, but it does not replace primary cap-table, board, or financing-document diligence. | 中 | SO026 |
| CO035 | Guardian coverage of AI and jobs underscores that humanoid deployment narratives will face workforce-impact and policy scrutiny even when framed as labor augmentation. | 中 | SO027 |
| CO036 | The milestone sequence now spans founding in 2024, CTO appointment, product prototypes, SAP/Martur Fompak POC, Siemens/NVIDIA pilot, Schaeffler partnership, Bosch manufacturing reports, and the July 2026 Series A unicorn event. | 高 | SO007, SO008, SO009, SO010, SO012, SO017 |
| CO037 | The stakeholder map separates financial investors from strategic industrial partners because Schaeffler and Bosch matter as deployment/manufacturing validators, not just cap-table names. | 高 | SO007, SO012, SO021 |
| CO038 | No public source reviewed provides a full board roster, voting-control terms, founder dilution, or secondary-sale information for the Series A. | 中 | SO010, SO012, SO026 |
| CO039 | The official newsroom shows a dense 2026 cadence of Bosch, Schaeffler, Siemens/NVIDIA, SAP/Martur, KinetIQ, and product announcements. | 中 | SO003 |
| CO040 | Independent and semi-independent product directories corroborate the HMND 01 positioning as a modular industrial humanoid platform, but they are not substitutes for verified customer acceptance data. | 中 | SO022, SO023 |
| CM001 | The relevant market is human-form mobile robots for industrial or commercial work in human-centric environments rather than the entire robotics market. | 高 | SM010, SM016, SM019 |
| CM002 | Conventional industrial robots are an adjacent excluded category because OSHA defines industrial robots as programmable multifunctional mechanical devices performing variable programmed motions. | 高 | SM019, SM008 |
| CM003 | Autonomous mobile robots, fixed industrial arms, conveyors, and manual labor remain status-quo substitutes for humanoid material movement and assembly jobs. | 中 | SM008, SM011, SM018 |
| CM004 | MarketsandMarkets lists manufacturing, warehousing and distribution centers, healthcare, education, hospitality, retail, and public relations as humanoid applications. | 中 | SM001 |
| CM005 | Humanoid’s homepage describes Alpha Wheeled as industrial-grade humanoid labor designed to operate autonomously at human speed. | 中 | SM016 |
| CM006 | Humanoid says the Schaeffler partnership will bring hundreds of robots into production facilities over five years with 2026-2027 beta validation. | 中 | SM017 |
| CM007 | GXO piloted Agility Robotics’ Digit at a SPANX warehouse to test logistics tasks and free employees for higher-value work. | 中 | SM018 |
| CM008 | MarketsandMarkets estimates the humanoid robot market at USD 5.41 billion in 2026 and USD 50.27 billion by 2035, a 28.1% CAGR. | 中 | SM001 |
| CM009 | Fortune Business Insights estimates the market at USD 6.24 billion in 2026 and USD 165.13 billion by 2034, a 50.60% CAGR. | 中 | SM002 |
| CM010 | Precedence Research estimates a much lower USD 2.16 billion 2026 market and USD 8.78 billion 2035 market, contradicting higher-growth TAM lenses. | 中 | SM004 |
| CM011 | Mordor Intelligence estimates the humanoids market at USD 3.93 billion in 2026 and USD 17.80 billion by 2031, a 35.26% CAGR. | 中 | SM005 |
| CM012 | Goldman Sachs’ 2035 lens is USD 38 billion of global humanoid market value with about 1.4 million unit shipments. | 高 | SM006, SM013 |
| CM013 | Morgan Stanley frames humanoids as a potential USD 5 trillion market by 2050 with nearly 1 billion robots in use. | 中 | SM007 |
| CM014 | Morgan Stanley’s China lens, as reported by ChinaBizInsider, forecasts 50,000 China humanoid shipments and a USD 2 billion China market in 2026. | 中 | SM015 |
| CM015 | Axis Intelligence reports approximately USD 4.89 billion of humanoid revenue in 2025, USD 6.24 billion in 2026, and 16,000-18,000 units shipped in 2025. | 中 | SM014 |
| CM016 | Grand View Research’s archived page estimates only USD 1.55 billion in 2024 and USD 4.04 billion by 2030, a lower path than the 2035 mega-TAM cases. | 中 | SM003 |
| CM017 | Global Market Insights publishes a high-side forecast of USD 10.9 billion in 2026 and USD 192.7 billion by 2035. | 中 | SM021 |
| CM018 | Coherent Market Insights estimates USD 6.40 billion in 2026 and USD 104.95 billion by 2033 at a 49.0% CAGR. | 中 | SM022 |
| CM019 | The Business Research Company reports USD 5.44 billion in 2025 and USD 39 billion by 2030 at a 47.1% CAGR. | 中 | SM024 |
| CM020 | SkyQuest estimates USD 1.46562 billion in 2025 and USD 35.41439 billion by 2033 at a 48.9% CAGR. | 中 | SM025 |
| CM021 | Industrial and commercial SAM should initially prioritize manufacturing, logistics, and warehousing because those uses appear in analyst application lists and in real pilots. | 高 | SM001, SM005, SM017, SM018 |
| CM022 | Humanoid’s own SOM should be constrained to validated industrial beta deployments and partner conversion rather than consumer-home adoption. | 中 | SM016, SM017 |
| CM023 | Manufacturing adoption is likely owned by plant operations, manufacturing engineering, industrial engineering, safety, and IT/security stakeholders. | 中 | SM017, SM028 |
| CM024 | Warehouse adoption is likely owned by 3PL operations, warehouse automation leaders, and labor-planning teams that manage material-movement workflows. | 中 | SM018, SM027 |
| CM025 | Retail appears in broad market reports, but the cited evidence is weaker than manufacturing and logistics evidence for Humanoid’s near-term industrial boundary. | 低 | SM001, SM002, SM016 |
| CM026 | The adoption path runs from workflow selection to proof-of-concept, safety and IT integration, availability validation, and then phased fleet procurement. | 高 | SM012, SM017, SM018 |
| CM027 | Enterprise buyers are expected to require measurable productivity gains and total-cost-of-ownership proof rather than buying humanoids for experimentation alone. | 中 | SM002, SM012 |
| CM028 | Safety is a hard constraint because OSHA highlights robot accidents during non-routine work and ISO 10218-1:2025 defines safety requirements for industrial robots. | 高 | SM019, SM020 |
| CM029 | McKinsey identifies cost reduction, dexterity and mobility, sustained uptime, and safety for fenceless operation as bridges to commercial viability. | 中 | SM012 |
| CM030 | Actuators, batteries, sensors, compute, and control platforms are material cost and performance constraints for humanoid scaling. | 中 | SM001, SM002, SM012 |
| CM031 | Labor gaps and demographic aging are explicit demand drivers in market reports and IFR’s humanoid position-paper framing. | 中 | SM005, SM010, SM030 |
| CM032 | BLS reports 6.95 million U.S. hand laborer and material mover jobs in 2024, creating a large logistics labor pool for automation comparison. | 中 | SM027 |
| CM033 | BLS projects production occupations to decline overall but still generate about 963,400 openings per year due to replacement needs. | 中 | SM028 |
| CM034 | IFR reported 4,281,585 operational industrial robots worldwide and 541,302 annual installations in 2023, showing a large automation installed base. | 高 | SM008, SM011 |
| CM035 | The 2026 Barcelona Declaration signatories framed robotics as a strategic priority for industrial competitiveness and talent development. | 中 | SM009 |
| CM036 | WEF’s Future of Jobs Report 2025 surveyed over 1,000 employers representing more than 14 million workers on technology-led workforce transformation through 2030. | 中 | SM029 |
| CM037 | UN World Population Prospects 2024 provides population projections to 2100 that underpin the aging-workforce demand thesis. | 中 | SM030 |
| CM038 | IFR explicitly warns that humanoid robots are surrounded by hype and says its position paper separates vision from reality. | 中 | SM010 |
| CM039 | Humanoids must beat existing tools and infrastructure because buyers can keep using installed automation, manual labor, and workflow-specific machines. | 中 | SM005, SM012, SM018 |
| CM040 | Robotics-as-a-service can reduce deployment-cost friction but does not remove the need to prove uptime, serviceability, and payback. | 中 | SM001, SM017, SM012 |
| CM041 | Asia is likely to be an important manufacturing and component hub for humanoids because Goldman Sachs cites the region’s supply-chain breadth and lower manufacturing costs. | 高 | SM006, SM007 |
| CM042 | Early logistics humanoid use cases are narrow material-handling tasks rather than full general-purpose labor replacement. | 中 | SM018, SM027 |
| CM043 | Deloitte’s 2026 manufacturing outlook says targeted technology investments could be essential to maintaining competitiveness. | 中 | SM031 |
| CM044 | The retained sizing set spans roughly USD 2.16 billion to USD 10.9 billion for 2026 and USD 8.78 billion to USD 192.7 billion for 2035, so TAM is an estimate range rather than a settled fact. | 中 | SM001, SM004, SM021, SM022 |
| CP001 | Humanoid announced a $152 million Series A at a $1.35 billion post-money valuation and $270 million total raised. | 高 | SP002, SP003 |
| CP002 | Prime Movers Lab led Humanoid’s Series A, with Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures participating. | 中 | SP002 |
| CP003 | Humanoid is described in the funding coverage as Europe’s first pure-play humanoid robotics unicorn. | 高 | SP002, SP003 |
| CP004 | Humanoid’s announced Schaeffler relationship covers large-scale deployment of thousands of humanoid robots in manufacturing environments. | 高 | SP002, SP004 |
| CP005 | Humanoid’s initial Schaeffler deployments are scheduled around beta-stage robots in 2026-2027. | 中 | SP004 |
| CP006 | Humanoid’s product strategy emphasizes wheeled robots for manufacturing, logistics, and other industrial environments. | 高 | SP001, SP002, SP005 |
| CP007 | Humanoid’s KinetIQ stack is presented as a proprietary AI brain for reasoning and executing physical tasks in industrial environments. | 高 | SP001, SP002 |
| CP008 | Figure AI announced more than $1 billion in committed Series C capital at a $39 billion post-money valuation. | 中 | SP007 |
| CP009 | Figure’s official site positions Figure 03 as a general-purpose humanoid for everyday home help powered by Helix. | 中 | SP006, SP007 |
| CP010 | Figure 02 contributed to production of more than 30,000 BMW X3 vehicles and logged more than 1,250 operational hours. | 高 | SP008, SP009 |
| CP011 | Figure 03 returned to BMW Plant Spartanburg for a more complex sequencing workflow powered by Helix 02. | 中 | SP009 |
| CP012 | Figure’s Helix logistics policy report says average handling time fell from about 5.0 seconds to 4.31 seconds as package diversity increased. | 中 | SP010 |
| CP013 | Figure announced a Brookfield partnership tied to real-world humanoid pretraining data across a portfolio including 100,000 residential units. | 中 | SP011 |
| CP014 | Apptronik announced over $935 million in Series A financing after a $520 million Series A-X extension. | 高 | SP014, SP015 |
| CP015 | TechCrunch and CNBC reported Apptronik’s 2026 valuation at roughly $5.0 billion to $5.3 billion. | 高 | SP016, SP017 |
| CP016 | Apptronik lists Apollo as a humanoid platform for manufacturing, warehousing, logistics, retail, and other physically demanding work. | 高 | SP013, SP014 |
| CP017 | Apptronik publicly says Apollo has both bipedal and wheeled mobility options. | 高 | SP013, SP018, SP019 |
| CP018 | Apptronik cites partnerships with Mercedes-Benz, GXO Logistics, Jabil, and Google DeepMind. | 高 | SP014, SP015, SP017 |
| CP019 | Forbes reported Apptronik’s Robot Park as a 90,000-square-foot physical-AI data factory and quoted its CEO saying the industry is essentially all prototypes. | 中 | SP018 |
| CP020 | Business Insider reported that Apptronik Robot Park uses remote human control much of the time to generate training data for autonomy. | 中 | SP019 |
| CP021 | Agility Robotics describes Digit as a commercially deployed bipedal humanoid for logistics and manufacturing work. | 高 | SP020, SP021, SP022 |
| CP022 | GXO and Agility announced a multi-year Robots-as-a-Service agreement deploying Digit and Agility Arc in a live warehouse environment. | 高 | SP021, SP022 |
| CP023 | Agility announced Amazon testing of Digit for tote recycling and noted Agility was part of the Amazon Industrial Innovation Fund. | 中 | SP023 |
| CP024 | Agility announced a 2026 commercial agreement with Toyota Motor Manufacturing Canada after a successful pilot. | 中 | SP024 |
| CP025 | Agility and Churchill Capital Corp XI announced a confidential S-4 filing for a proposed business combination expected to provide more than $620 million gross proceeds. | 中 | SP025 |
| CP026 | Agility reported Digit had moved more than 100,000 totes in commercial deployment. | 高 | SP026, SP020 |
| CP027 | 1X positions NEO as a home robot that can work autonomously by default and use 1X Expert remote guidance for unknown chores. | 高 | SP028, SP030 |
| CP028 | Humanoids Daily reported NEO pre-orders at a $20,000 purchase price or $499 monthly subscription with U.S. deliveries slated to begin in 2026. | 中 | SP030 |
| CP029 | 1X says NEO has 25-degree-of-freedom tendon-driven hands and a vertically integrated Hayward humanoid factory. | 中 | SP029 |
| CP030 | Humanoids Daily reported 1X and EQT outlined up to 10,000 NEO robots across EQT portfolio companies between 2026 and 2030. | 中 | SP031 |
| CP031 | The 1X-EQT source cautions that the arrangement appears to be an enabling agreement rather than a guaranteed purchase order. | 中 | SP031 |
| CP032 | Boston Dynamics presents Atlas as an electric humanoid platform, but the public evidence reviewed is product and R&D material rather than customer deployment proof. | 中 | SP032, SP033 |
| CP033 | Unitree lists G1 from US$13.5K with about 35 kg weight, 23 to 43 joint motors, and about two hours battery life. | 中 | SP034 |
| CP034 | Humanoids Daily reported a Unitree G1 appeared on Walmart’s U.S. website for $21,600, reinforcing low-cost channel pressure. | 中 | SP040, SP034 |
| CP035 | Sanctuary AI now emphasizes hardware-agnostic production-ready Physical AI for industrial work and a path to humanoids. | 中 | SP036 |
| CP036 | UBTECH’s official site lists Industrial Walker S2, Walker S1, and Walker S products, plus warehousing and logistics application scenarios. | 中 | SP037 |
| CP037 | Humanoids Daily reported UBTECH targeting a ten-fold output jump to 5,000 humanoids in 2026. | 中 | SP044 |
| CP038 | Humanoids Daily reported NEURA Robotics raised a $1.4 billion Series C led by Tether, expanding Europe’s competitive set. | 中 | SP041 |
| CP039 | Humanoids Daily reported Flexion raised $50 million to build humanoid robotics technology in Europe. | 中 | SP042 |
| CP040 | Fourier Intelligence launched an N1 humanoid with an open-source strategy, adding another China-linked platform to the landscape. | 中 | SP043 |
| CP041 | Tesla Optimus public reporting reviewed here emphasizes recruitment, Gen 3 speculation, and production ambitions rather than disclosed industrial customer deployments. | 中 | SP039, SP045 |
| CP042 | An ex-Tesla Optimus lead publicly questioned factory humanoid value and favored simpler automation, reinforcing status-quo displacement risk. | 中 | SP046 |
| CP043 | Humanoid’s total raised is materially smaller than Figure’s Series C and Apptronik’s Series A total. | 高 | SP002, SP007, SP014, SP015 |
| CP044 | Humanoid’s $1.35 billion valuation is far below Figure’s $39 billion and below Apptronik’s roughly $5.0 billion to $5.3 billion valuation. | 高 | SP002, SP003, SP007, SP016, SP017 |
| CP045 | Humanoid’s wheeled-first industrial wedge aligns with independent reporting that industrial customers ask for wheels for stability, battery life, and predictability. | 高 | SP005, SP018, SP019 |
| CP046 | The clearest near-term moat across humanoid competitors is access to real customer workflows and robot-generated training data, not a disclosed hardware spec alone. | 高 | SP004, SP008, SP010, SP018, SP019, SP022, SP026 |
| CP047 | Only Unitree and 1X provide reviewed public list-price signals; industrial contracts for Humanoid, Figure, Apptronik, and Agility remain opaque. | 高 | SP030, SP034, SP040, SP002, SP007, SP014, SP022 |
| CP048 | Industrial humanoid deployment remains early because many public examples are pilots, data factories, or first deployments rather than broad multi-site production fleets. | 高 | SP018, SP019, SP021, SP022, SP030, SP031, SP038, SP046 |
| CI001 | Humanoid announced a $152 million Series A financing on 2026-07-21 at a $1.35 billion post-money valuation. | 高 | SI004, SI005, SI015 |
| CI002 | Multiple independent reports state that Humanoid has raised $270 million in total capital to date. | 高 | SI004, SI009, SI013 |
| CI003 | The $270 million total raised implies roughly $118 million of pre-Series-A capital when the $152 million Series A is subtracted. | 中 | SI004, SI009 |
| CI004 | Forbes reported in February 2026 that Humanoid had been in talks to raise about $200 million before the final Series A announcement. | 中 | SI008 |
| CI005 | The reported seed/early financing history is incomplete publicly, with the shared diligence brief indicating a seed round of about $33 million and other early rounds comprising the balance of prior capital. | 低 | SI008, SI029, SI030 |
| CI006 | Prime Movers Lab led the Series A, with Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures reported as participants. | 高 | SI004, SI010, SI016 |
| CI007 | Strategic investors Schaeffler and Bosch are also tied to Humanoid’s manufacturing and deployment path, making the round partly a supply-chain and customer-channel financing. | 中 | SI003, SI004, SI017, SI021 |
| CI008 | The Series A represents about 56% of the $270 million reported lifetime capital base. | 中 | SI001, SI004, SI009 |
| CI009 | Public sources reviewed do not disclose recognized revenue, ARR, gross margin, monthly burn, cash balance, debt, or preference-stack terms. | 中 | SI004, SI009, SI027, SI029, SI030, SI031 |
| CI010 | Humanoid publicly claims or is reported to have 34,000 robot pre-orders worth about $2.4 billion. | 中 | SI009, SI023, SI024, SI025 |
| CI011 | The 34,000 pre-orders and $2.4 billion stated order value imply an average order value of about $70,600 per robot before discounts, services, taxes, and cancellation risk. | 中 | SI009, SI023, SI024 |
| CI012 | Pre-orders and order value are bookings signals rather than recognized revenue unless robots are delivered, accepted, priced, and collectible under contract terms. | 中 | SI009, SI014, SI023 |
| CI013 | TechStartups reported that Humanoid says it has completed nine industrial deployments with Fortune 500 companies. | 中 | SI009 |
| CI014 | Tech.eu reported that Humanoid signed what it called the industry’s largest publicly announced commercial agreement with Schaeffler for thousands of robots. | 中 | SI002, SI004 |
| CI015 | Forbes and Robonomy reported a four-digit or 1,000-plus Schaeffler robot deal, creating a more specific but still secondary view of the Schaeffler order scale. | 中 | SI018, SI019, SI020 |
| CI016 | Humanoid’s official Bosch page confirms a May 2026 partnership and a successful March 2026 proof of concept, but it does not itself disclose a Bosch unit-order count. | 中 | SI003 |
| CI017 | The Next Web and Forbes report that Bosch is expected to manufacture Humanoid’s wheeled robots at scale. | 高 | SI017, SI021, SI022 |
| CI018 | Forbes coverage of the Schaeffler agreement cites a pathway that could reach 100,000 units by 2031, which is a scale target rather than a current delivered-unit count. | 中 | SI018, SI019, SI020 |
| CI019 | The Economic Times reported that beta robot deployments at customer sites were planned for later in 2026. | 中 | SI014 |
| CI020 | Humanoid’s likely revenue streams are hardware sales, leasing or robot-as-a-service, software/AI capability, maintenance, spares, integration, and pilot fees, but public sources do not disclose realized mix. | 中 | SI001, SI003, SI009, SI017 |
| CI021 | Official and press descriptions position the robots for manufacturing, logistics, and industrial labor-shortage use cases rather than consumer deployments. | 中 | SI001, SI004, SI016 |
| CI022 | Goldman Sachs projected the global humanoid robot market could reach $38 billion by 2035. | 中 | SI040 |
| CI023 | Roland Berger described humanoid robots in 2026 as moving from science fiction toward industrial reality as hardware, AI, and ecosystems converge. | 中 | SI037, SI038 |
| CI024 | K2 Capital’s 2026 actuator research describes actuators as accounting for roughly 30–50% of humanoid robot cost. | 中 | SI039 |
| CI025 | Humanoid robot cost structure should be modeled around actuators, compute, sensors, batteries, mechanical structure, assembly, testing, warranty, integration, and field service rather than software gross margin alone. | 中 | SI039, SI041, SI043, SI044 |
| CI026 | McKinsey warned that humanoid supply-chain constraints can become a major value bottleneck, making component access and scale manufacturing central to financial outcomes. | 中 | SI041 |
| CI027 | Robotics & Automation News highlighted that commercial success depends on effective output, supporting skepticism that robot demos automatically convert to customer ROI. | 中 | SI042 |
| CI028 | 36Kr Europe framed Humanoid’s 34,000 orders and IPO ambitions skeptically, asking whether the company is a genius or a fraud. | 中 | SI025 |
| CI029 | Companies House search results surface HUMANOID LTD company number 14835233, incorporated on 29 April 2023. | 中 | SI026, SI027 |
| CI030 | The HUMANOID LTD Companies House profile lists active private limited status, registered office at 2 Foulden Terrace, London, and accounts made up to 1 May 2025. | 中 | SI027, SI028 |
| CI031 | The filing history for HUMANOID LTD includes dormant company accounts made up to 1 May 2025 and dormant accounts made up to 29 April 2024. | 中 | SI027, SI028 |
| CI032 | The Companies House profile’s SIC codes and officer listing do not obviously match the robotics operating profile described by Humanoid’s website and press coverage. | 中 | SI001, SI027, SI028 |
| CI033 | The public filing mismatch means Companies House data is useful for entity-search diligence but not sufficient to underwrite the robotics company’s revenue, cash, or cap table. | 中 | SI026, SI027, SI028 |
| CI034 | Humanoid’s team size is reported around 200 people by multiple sources, while the shared brief frames the diligence range as roughly 160–200 engineers and researchers. | 中 | SI024, SI025, SI009 |
| CI035 | A $152 million Series A could support roughly 24–36 months only under a $4–6 million monthly burn scenario; actual runway remains unknown because burn and cash are undisclosed. | 中 | SI001, SI004, SI041, SI034 |
| CI036 | A more hardware-intensive $6–10 million monthly burn scenario would reduce the same $152 million raise to roughly 15–25 months before considering revenue, working capital, and capex offsets. | 中 | SI004, SI039, SI041 |
| CI037 | Manufacturing through Bosch may reduce owned factory capex but does not remove working-capital exposure for inventory, components, warranty reserves, and customer acceptance. | 中 | SI003, SI017, SI021, SI041 |
| CI038 | The $2.4 billion order-book headline is large relative to the $1.35 billion post-money valuation, but it is not comparable to ARR or revenue without conversion and margin evidence. | 中 | SI001, SI004, SI009, SI023 |
| CI039 | Apptronik’s reported $935 million Series A and $5 billion-plus valuation show that high private valuations are common among humanoid peers before broad public revenue disclosure. | 中 | SI045 |
| CI040 | Venture database pages from PitchBook, Dealroom, Crunchbase, CB Insights-style sources, and Tracxn do not provide a complete public cap table, revenue, burn, or preference stack in the accessible fetches. | 中 | SI029, SI030, SI031, SI032 |
| CI041 | The publicly visible financing story is therefore stronger on capital availability and strategic validation than on revenue quality, gross margin, and cash-flow predictability. | 中 | SI001, SI004, SI009, SI025, SI042 |
| CI042 | The most important diligence blocker is a private data room tying pre-orders to signed contracts, deposits, cancellation terms, delivery schedules, bill of materials, and warranty-service cost curves. | 中 | SI009, SI023, SI039, SI041, SI042 |
| CE001 | HMND 01 Alpha is positioned as a next-generation labor automation unit for industrial and logistics tasks. | 高 | SE001, SE002 |
| CE002 | Humanoid publicly describes two HMND 01 Alpha embodiments: Alpha Wheeled for industrial use and Alpha Bipedal for R&D/service-learning use. | 高 | SE003, SE004, SE012 |
| CE003 | The canonical Alpha Wheeled specification is 220 cm height, about 300 kg weight, 2 m/s speed, four-hour runtime, 29 DoF excluding end effectors, and 15 kg bimanual payload. | 高 | SE003, SE016, SE017 |
| CE004 | The canonical Alpha Bipedal specification is 179 cm height, 90 kg weight, 1.5 m/s speed, three-hour runtime, 29 DoF excluding end effectors, and 15 kg bimanual payload. | 中 | SE016, SE018 |
| CE005 | HMND 01’s disclosed manipulation package supports modular end effectors including 12-DoF five-finger hands and 1-DoF parallel grippers. | 中 | SE003, SE016, SE018 |
| CE006 | Independent robot databases corroborate a 29-DoF and 15 kg payload baseline for HMND 01. | 中 | SE016, SE017, SE018 |
| CE007 | The reviewed sources describe a sensor suite that includes 360-degree RGB vision, depth sensors, wrist cameras, 6D force/torque sensing, haptics, and a microphone array. | 中 | SE016, SE018 |
| CE008 | Humanoid says its alpha robots use NVIDIA Jetson Thor, Isaac Sim, and Isaac Lab; the chapter treats NVIDIA Jetson Orin/Thor plus Intel i9 as the canonical compute dependency from the shared brief and spec aggregators. | 中 | SE012, SE016, SE026, SE027 |
| CE009 | NVIDIA positions Jetson Orin and Jetson Thor as edge AI platforms for next-generation robotics and physical robotics workloads. | 中 | SE026, SE027 |
| CE010 | NVIDIA Isaac GR00T bundles robot foundation models, simulation frameworks, middleware, CUDA-X runtime libraries, and Jetson Thor for real-time robot inference and control. | 中 | SE028 |
| CE011 | NVIDIA Isaac Sim is a simulation, testing, and synthetic-data framework that can ingest CAD, URDF, and real-world captures into robotics validation pipelines. | 中 | SE029 |
| CE012 | KinetIQ is publicly described as a four-layer architecture spanning fleet orchestration, individual robot planning, VLA locomanipulation, and whole-body control. | 高 | SE006, SE008, SE025 |
| CE013 | KinetIQ System 3 coordinates multiple robots against external fleet-level goals on timescales of seconds to minutes and beyond. | 中 | SE006 |
| CE014 | KinetIQ System 2 coordinates a single robot by invoking capabilities such as navigation or locomanipulation as agentic tools. | 中 | SE006 |
| CE015 | KinetIQ System 1 is a VLA-based locomanipulation network that translates natural-language goals into target poses and runs at 10 Hz. | 中 | SE006 |
| CE016 | KinetIQ System 0 is a whole-body control layer that executes target poses while preserving dynamic stability at a 50 Hz control rate. | 中 | SE006 |
| CE017 | Humanoid’s KinetIQ tooling records both teleoperation and autonomous policy execution for workflows such as DAgger and training on successful policy rollouts. | 中 | SE006 |
| CE018 | KinetIQ Ascend extends the stack with end-to-end, vision-guided reinforcement learning for manipulation VLAs rather than relying only on behavior cloning. | 中 | SE007 |
| CE019 | Humanoid claims KinetIQ Ascend improved industrial machine-feeding throughput by 42%. | 中 | SE007 |
| CE020 | Humanoid claims KinetIQ Ascend improved item handover throughput by 85% and success from 80% to 98%. | 中 | SE007 |
| CE021 | Humanoid’s own technical writing says behavior-cloned policies inherit demonstrator speed and quality and can fall short of industrial reliability. | 中 | SE007 |
| CE022 | Humanoid’s deployment-ready RL discussion frames sim-to-real transfer, model mismatch, action space design, and reward tuning as practical deployment risks. | 中 | SE013 |
| CE023 | The SAP and Martur Fompak POC tested HMND 01 Alpha Wheeled in a live production logistics workflow, where it navigated to a pallet, retrieved a KLT box, and delivered it to a trolley. | 高 | SE008, SE023, SE025 |
| CE024 | Humanoid says its robot API connected to SAP APIs through Joule so SAP Extended Warehouse Management could send tasks to the robot over the internet. | 高 | SE008, SE021, SE025 |
| CE025 | SAP described the Martur Fompak deployment as SAP S/4HANA and SAP EWM enriching humanoid robots with task, attribute, and exception-handling knowledge. | 高 | SE021, SE020 |
| CE026 | SAP reported early results of increased throughput, fewer errors, and a target of up to five times greater work efficiency for mass production. | 中 | SE021 |
| CE027 | Humanoids Daily reported the SAP trial used an 8 kg dual-arm payload limit while executing the Martur Fompak picking tasks. | 中 | SE024 |
| CE028 | Siemens says HMND 01 Alpha was successfully tested at Siemens’ Erlangen electronics factory, performing logistics operations with the NVIDIA physical-AI stack. | 高 | SE009, SE022 |
| CE029 | Siemens positioned Xcelerator as the industrial integration layer and NVIDIA as the simulation/development accelerator for the factory pilot. | 高 | SE009, SE022 |
| CE030 | Bosch and Humanoid completed a March 2026 intralogistics POC transferring boxes from a conveyor to a trolley across five box sizes and multiple footprints. | 中 | SE010 |
| CE031 | Bosch will act as Humanoid’s contract manufacturing partner and apply a Design for Excellence framework covering hardware design, production, supply chain, and cost optimization. | 中 | SE010 |
| CE032 | Schaeffler signed a binding phased deployment and supply agreement and becomes Humanoid’s preferred actuator supplier. | 中 | SE011 |
| CE033 | The initial Schaeffler deployment phase is described as running from December 2026 through June 2027 across two German sites. | 中 | SE011 |
| CE034 | The Alpha Wheeled prototype was built in seven months, and the Alpha Bipedal page says the bipedal robot was developed in five months and walked stably after 48 hours of training. | 高 | SE003, SE004 |
| CE035 | RoboZaps classifies HMND 01 deployment status as piloting and notes a later Beta version on the roadmap. | 中 | SE018 |
| CE036 | Tech.eu reported that Humanoid plans Beta version robot rollouts in Q4 2026, mass manufacturing of wheeled humanoids, next-generation platform development, and continued KinetIQ work. | 中 | SE035 |
| CE037 | Humanoid’s vertical-integration claim combines proprietary hardware, proprietary KinetIQ AI, NVIDIA-enabled simulation, Bosch manufacturing, and Schaeffler actuator supply. | 中 | SE010, SE011, SE012, SE035 |
| CE038 | Humanoid’s control-software recruiting emphasizes modular real-time architectures for locomotion, manipulation, and teleoperation plus reliability, latency, safety metrics, HIL, regression, and performance testing. | 中 | SE014 |
| CE039 | Humanoid’s RL manipulation recruiting calls for simulation task suites, teleoperation trajectory collection, behavior cloning, and sim-to-real experimentation. | 中 | SE015 |
| CE040 | Public safety evidence is currently process-oriented rather than certification-oriented: disclosed controls include fail-safes, watchdogs, redundancy, simulation, HIL testing, and safety metrics in recruiting and technical posts. | 中 | SE013, SE014, SE029 |
| CE041 | IEEE Spectrum warns that humanoid scaling depends on battery life, reliability, safety, and standards that are still evolving for dynamically balancing systems. | 高 | SE030, SE031 |
| CE042 | Robotics & Automation News says humanoid platforms remain impressive in demos and narrow tasks but far from general-purpose autonomy. | 中 | SE031 |
| CE043 | Unite.AI’s interview with Artem Sokolov framed autonomy versus control, data privacy, safety protocols, and ethical audits as key unresolved robotics questions. | 中 | SE033 |
| CE044 | Humanoids Daily described Humanoid’s autonomy roadmap as staged: shared autonomy, increased independence in 2026-2027, and full autonomy in targeted industrial settings in late 2027. | 中 | SE034 |
| CE045 | Humanoid.guide still lists older HMND 01 dimensions and DoF, creating a public-spec conflict against newer official and robot-database numbers. | 低 | SE019, SE018 |
| CE046 | No reviewed source disclosed production-grade MTBF, intervention rate, uptime SLA, battery chemistry, formal safety certification, or patent-family evidence for KinetIQ. | 低 | |
| CU001 | Humanoid publicly targets industrial customers in manufacturing, logistics, retail and other critical sectors rather than consumers. | 高 | SU001, SU018 |
| CU002 | The publicly named partner/customer set includes Schaeffler, Bosch, Siemens, NVIDIA, SAP and Martur Fompak. | 高 | SU001, SU018, SU020 |
| CU003 | Schaeffler is the strongest public commercial proof point because Humanoid calls the May 2026 agreement binding and phased. | 中 | SU002, SU005 |
| CU004 | The Schaeffler deployment is described across company and independent sources as a four-digit or thousands-of-robots manufacturing rollout. | 中 | SU002, SU004, SU006, SU007 |
| CU005 | Schaeffler deployments are scheduled to begin in live German production environments before the end of 2026. | 中 | SU002, SU005, SU006 |
| CU006 | Forbes reports phase one of the Schaeffler contract runs from December 2026 through June 2027 across Herzogenaurach and Schweinfurt. | 中 | SU028 |
| CU007 | The first Schaeffler use cases focus on box handling, logistics tasks and validation toward stable continuous operation at production scale. | 高 | SU006, SU028 |
| CU008 | The Schaeffler agreement is structured as Robot-as-a-Service with fleet management, maintenance and support rather than a simple hardware sale. | 高 | SU004, SU028 |
| CU009 | Schaeffler also serves as a strategic technology partner and preferred actuator supplier for Humanoid wheeled platforms. | 中 | SU002, SU003, SU010 |
| CU010 | The January 2026 Schaeffler partnership initially described hundreds of robots over five years before the later May announcement framed thousands. | 中 | SU003, SU029, SU002 |
| CU011 | Humanoid announced a Bosch partnership after a March 2026 proof of concept in a complex industrial intralogistics workflow. | 中 | SU008, SU009 |
| CU012 | Bosch is positioned as Humanoid’s manufacturing partner for scaled production of HMND 01 rather than solely as an end customer. | 高 | SU008, SU012, SU013 |
| CU013 | The Bosch proof of concept included transferring boxes between conveyor and trolley and handling intralogistics workflow variability. | 中 | SU008, SU009 |
| CU014 | The task brief attributes an approximately 1,000-unit order to Bosch, but public fetched sources more consistently tie the 1,000-plus unit framing to Schaeffler. | 中 | SU012, SU013, SU028 |
| CU015 | Siemens and Humanoid tested HMND 01 Alpha at Siemens electronics factory in Erlangen, Germany. | 高 | SU014, SU015, SU027 |
| CU016 | The Siemens pilot reported 60 tote moves per hour, more than eight hours uptime and above 90 percent autonomous pick-and-place success. | 高 | SU014, SU027 |
| CU017 | NVIDIA’s physical AI stack is part of the Siemens-Humanoid factory pilot and the broader Siemens industrial AI positioning. | 高 | SU014, SU015, SU016 |
| CU018 | Humanoid, SAP and Martur Fompak completed a proof of concept in a live automotive manufacturing logistics environment in March 2026. | 高 | SU026, SU001 |
| CU019 | The SAP/Martur Fompak proof of concept had the robot receive instructions from an SAP AI agent, retrieve a KLT box and deliver it to a trolley. | 中 | SU026 |
| CU020 | SAP separately frames embodied AI as connecting robots to business context across warehouse, manufacturing and inspection workflows. | 中 | SU017 |
| CU021 | Humanoid reports Beta-version robot rollouts at customer facilities beginning in Q4 2026. | 中 | SU028, SU030 |
| CU022 | Unite.AI reports Humanoid has completed nine industrial deployments. | 中 | SU029 |
| CU023 | Unite.AI reports Humanoid has accumulated 34,000 robot pre-orders representing approximately $2.4 billion of commercial pipeline. | 中 | SU029 |
| CU024 | The Next Web reports the 34,000 pre-order figure is not independently verified and dwarfs revenue of humanoid robotics companies currently shipping hardware. | 中 | SU030 |
| CU025 | The 34,000 pre-order book should not be treated as recognized revenue or contracted backlog absent disclosure of binding purchase orders, deposits or cancellation rights. | 中 | SU022, SU023, SU030 |
| CU026 | Tech.eu reports Humanoid recently signed the industry’s largest publicly announced commercial agreement with Schaeffler. | 中 | SU018 |
| CU027 | Humanoid’s named public customer proof is concentrated in automotive manufacturing, industrial manufacturing and factory logistics. | 高 | SU002, SU008, SU014, SU026 |
| CU028 | Retail appears in Humanoid’s stated target sectors and customer-facility roadmap but lacks a named public customer proof point. | 中 | SU018, SU028 |
| CU029 | No public source reviewed discloses Humanoid net revenue retention, gross revenue retention, churn or renewal rates. | 低 | |
| CU030 | The Schaeffler customer relationship creates concentration risk because it is simultaneously the largest public deployment, a strategic partner and a preferred actuator supplier. | 高 | SU002, SU003, SU028 |
| CU031 | Bosch creates partner-dependence risk because scaled production is tied to an external manufacturing partner rather than wholly captive production. | 高 | SU008, SU012, SU013 |
| CU032 | The public customer evidence is strongest for pilots and phased deployments, while full production revenue realization remains mostly forward-looking into late 2026 and beyond. | 高 | SU002, SU014, SU026, SU028 |
| CU033 | Physical AI Guide cautions that announced feasibility plans and orderability are not the same as active pilots or broad deployments. | 中 | SU023 |
| CU034 | VaaSBlock argues the sector still has a significant gap between current pilot capability and the autonomous general-purpose humanoid worker implied by marketing. | 中 | SU022 |
| CU035 | IoT Digital Twin PLM cautions that vendor and press deployment figures should not be treated as independently audited. | 中 | SU024 |
| CU036 | Humanoid’s go-to-market is enterprise-led and partner-led, combining RaaS deployment, contract manufacturing and strategic industrial partnerships. | 高 | SU004, SU008, SU012, SU028 |
| CU037 | Customer segments can be separated into anchor manufacturers, manufacturing partner/OEM, physical-AI ecosystem partners, automotive logistics pilot users and prospective unnamed pre-order customers. | 高 | SU002, SU008, SU014, SU026, SU029 |
| CU038 | The SAP/Martur Fompak and Siemens examples are proof-of-concept or pilot evidence, not disclosed multi-site commercial rollouts. | 高 | SU014, SU026, SU015 |
| CU039 | The Schaeffler relationship has better deal quality than ordinary LOI-style pipeline because public sources describe it as binding and phased. | 高 | SU002, SU005, SU028 |
| CU040 | The 20,000-to-34,000 pre-order variance remains an open diligence issue because fetched sources corroborate 34,000 while broader market search results also surface 20,000-unit humanoid preorder claims for other companies. | 低 | |
| CU041 | Customer outcomes are concrete in the Siemens pilot because throughput, uptime and pick-and-place success were disclosed. | 高 | SU014, SU027 |
| CU042 | Customer outcomes are less quantified in the Bosch and SAP/Martur Fompak POCs because public sources describe workflow completion but do not disclose sustained uptime or ROI. | 中 | SU008, SU009, SU026 |
| CU043 | No fetched source discloses signed revenue, ARR, recognized sales or realized gross margin from Humanoid’s named customer relationships. | 低 | |
| CU044 | Schaeffler, Bosch and Siemens all sit in or around automotive and industrial manufacturing, creating a vertical concentration that may be useful for wedge focus but risky for diversification. | 高 | SU002, SU008, SU014, SU028 |
| CU045 | Humanoid’s customer pipeline quality should be risk-weighted by stage: Schaeffler binding phased agreement highest, Bosch manufacturing/POC medium-high, Siemens and SAP/Martur POCs medium, and general pre-orders lower. | 高 | SU002, SU008, SU014, SU026, SU030 |
| CR001 | Humanoid announced a $152 million Series A at a $1.35 billion post-money valuation, bringing total raised to $270 million. | 高 | SR002, SR005, SR007 |
| CR002 | Prime Movers Lab led the Series A, with Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures participating. | 高 | SR002, SR005 |
| CR003 | The Next Web reported that Humanoid was barely two years old and had not yet shipped a commercial unit at the time of the July 2026 funding article. | 中 | SR004 |
| CR004 | The Next Web reported that Humanoid claimed 34,000 robot pre-orders worth nearly $2.5 billion and that the figure had not been independently verified. | 中 | SR004 |
| CR005 | Unite.AI reported 34,000 robot pre-orders representing an approximately $2.4 billion commercial pipeline. | 中 | SR007 |
| CR006 | Humanoid says it signed a binding, phased deployment and supply agreement with Schaeffler, with deployment beginning in live German production environments in late 2026. | 高 | SR008, SR002 |
| CR007 | Humanoid says Schaeffler will be its preferred supplier for more than 50% of joint-actuator demand for wheeled platforms through 2031. | 中 | SR008 |
| CR008 | Humanoid says its Bosch partnership followed a March 2026 proof of concept and marked the transition toward scaled production. | 高 | SR009, SR010 |
| CR009 | Bosch is quoted by Humanoid as having the infrastructure and industrialization expertise to help move from prototype to volume manufacturing. | 中 | SR009 |
| CR010 | Forbes described Humanoid as pursuing a distributed partner strategy rather than a fully vertically integrated manufacturing model. | 中 | SR010 |
| CR011 | The reviewed public sources do not disclose recognized revenue, ARR, gross margin, monthly burn, runway, or working-capital commitments. | 中 | SR002, SR004, SR007 |
| CR012 | Bookings should not be treated as revenue until signed terms, deposits, cancellation rights, delivery, acceptance, warranty, and service economics are reviewed. | 中 | SR004, SR007 |
| CR013 | Humanoid’s homepage states that HMND 01 Alpha Wheeled handles up to 15 kg, has up to four hours of runtime, uses KinetIQ on NVIDIA compute, and has 29 degrees of freedom excluding end effectors. | 中 | SR001 |
| CR014 | Humanoid’s homepage claims Alpha is designed to operate autonomously at human speed with a high level of reliability. | 中 | SR001 |
| CR015 | IEEE Spectrum warned that impressive robot demonstrations can include a human in the loop. | 中 | SR019 |
| CR016 | IEEE Spectrum noted that earlier advanced humanoids in DARPA-style tasks were controlled by engineers, used overhead safety tethers, and moved slowly. | 中 | SR021 |
| CR017 | Ars Technica, Vox, and Compare Robots each framed humanoid-robot narratives as vulnerable to hype, demo selection, or teleoperation caveats. | 中 | SR023, SR024, SR026 |
| CR018 | ISO 10218-1 addresses safety requirements for the industrial robot as a machine, while ISO/TS 15066 addresses collaborative robots. | 高 | SR013, SR014 |
| CR019 | ISO’s page states that ISO/TS 15066:2016 covers robots and robotic devices for collaborative robots and was last reviewed and confirmed in 2022. | 中 | SR014 |
| CR020 | OSHA’s robotics overview points employers to hazard recognition, hazard evaluation, solutions, and machine-guarding resources for robotics in the workplace. | 中 | SR015 |
| CR021 | Regulation (EU) 2023/1230 is the EU machinery regulation replacing the Machinery Directive and setting rules for machinery placed on the market. | 高 | SR031, SR016 |
| CR022 | The European Commission states that manufacturers must carry out conformity assessment, prepare technical files, issue EU declarations of conformity, and affix CE marking before EEA placement. | 中 | SR016 |
| CR023 | Industrial humanoids deployed near workers in Europe carry residual safety and certification risk until the risk assessment, CE file, validation data, and customer-site controls are shown. | 中 | SR013, SR014, SR015, SR016, SR031 |
| CR024 | NVIDIA says Jetson Thor targets physical AI and humanoid robotics and delivers up to 2070 FP4 TFLOPS with 128 GB of memory. | 中 | SR029 |
| CR025 | NVIDIA’s Jetson Orin page lists embedded AI performance tiers up to 275 TOPS, illustrating dependence on specialized accelerator roadmaps. | 中 | SR030 |
| CR026 | If Humanoid’s bill of materials relies on NVIDIA-class U.S.-origin compute, allocation, export classification, and customer geography can become operational constraints. | 中 | SR029, SR030, SR034 |
| CR027 | BIS says export compliance programs are procedures and tools intended to mitigate export-violation risk for organizations subject to the EAR. | 中 | SR033 |
| CR028 | BIS says U.S.-origin items subject to the EAR remain subject to the EAR wherever located, and some foreign-made items can be subject under de minimis or foreign-direct-product rules. | 高 | SR032, SR034 |
| CR029 | UK government guidance says an export licence is needed before exporting controlled military goods, software, technology, and dual-use-list items from the UK. | 中 | SR018 |
| CR030 | OFAC maintains a Russian Harmful Foreign Activities Sanctions program and lists OFAC-BIS sanctions/export-control alerts relevant to Russia-related compliance diligence. | 中 | SR017 |
| CR031 | Unitree’s G1 page lists a roughly 35 kg humanoid platform with 23 to 43 degrees of freedom, creating a visible low-cost Chinese competitor reference point. | 中 | SR028 |
| CR032 | Unitree states that the G1 is a civilian robot and asks users not to make dangerous modifications or use the robot hazardously. | 中 | SR028 |
| CR033 | The Next Web described Prime Movers Lab as the venture firm behind Figure AI’s $39 billion valuation, indicating a better-capitalized competitive benchmark. | 中 | SR004 |
| CR034 | Humanoid’s homepage and funding coverage list SAP, NVIDIA, Bosch, Siemens, and Schaeffler as important partner or customer proof points, increasing dependence on a small set of strategic counterparties. | 中 | SR001, SR002, SR008, SR009 |
| CR035 | Humanoid’s team page presents a broad technical organization focused on engineering, AI, machine learning, and robotic applications. | 中 | SR012 |
| CR036 | A February 2026 Forbes article framed Artem Sokolov’s background with the headline “A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million.” | 中 | SR011 |
| CR037 | The reviewed public evidence does not disclose Humanoid’s cap table, board composition, preference stack, key-person plan, sanctions screening, or export-control governance. | 中 | SR002, SR011, SR012, SR017, SR018 |
| CR038 | Funding coverage centers the company narrative on Sokolov’s quotes about going from idea to European humanoid unicorn in two years. | 中 | SR002, SR005 |
| CR039 | Russian-linked founder optics can compound diligence burden when combined with U.S., UK, and sanctions export-control regimes affecting controlled technology and cross-border deployments. | 中 | SR011, SR017, SR018, SR032, SR033, SR034 |
| CR040 | The Los Angeles Times article quoted Rodney Brooks’ view that having ideas is easy but turning them into reality is hard in the context of inflated technology hype. | 中 | SR025 |
| CR041 | Vox reported that humanoid robots have been hyped as the future of factory work, household chores, elder care, and more. | 中 | SR024 |
| CR042 | Compare Robots argued that the gap between a demo reel and a robot trusted in real settings remains wide and that many demos are teleoperated. | 中 | SR026 |
| CR043 | RoboHorizon argued that large-scale deployment, industrial strategy, social trust, and regulatory skill matter more than spectacular prototypes. | 中 | SR027 |
| CR044 | Residual investment risk remains high until Humanoid verifies contract enforceability, safety certification, beta deployment reliability, production yield, and unit economics. | 中 | SR004, SR008, SR009, SR013, SR016, SR031 |
| CR045 | The near-term mitigation plan should prioritize contract audit, safety case evidence, reliability data, supplier allocation, governance documents, and financing runway before increasing price exposure. | 中 | SR004, SR013, SR016, SR033, SR034 |
| CR046 | Public sources reviewed for this chapter did not verify field incident history, recall history, customer acceptance criteria, or certification status for HMND 01 deployments. | 低 | |
| CR047 | A downside financing scenario is plausible if the 34,000-unit pipeline does not convert into accepted deployments before cash needs for manufacturing, inventory, warranty, and support increase. | 中 | SR004, SR007, SR009, SR010 |
| CR048 | The most important commercial monitoring indicators are signed customer deposits, cancellation rights, delivered robots, acceptance rates, utilization, and renewal/service attach. | 中 | SR004, SR006, SR007, SR008 |
| CV001 | Humanoid announced a $152 million Series A financing at a $1.35 billion post-money valuation on 2026-07-21. | 高 | SV001, SV007, SV008 |
| CV002 | Humanoid reported total capital raised to date of $270 million after the Series A. | 高 | SV001, SV007, SV009 |
| CV003 | The Series A implies approximately $118 million of prior capital raised before the $152 million round. | 中 | SV001, SV007 |
| CV004 | At a $1.35 billion post-money valuation and $270 million raised, the valuation is about 5.0 times lifetime capital raised. | 中 | SV001, SV007 |
| CV005 | Public sources repeat a 34,000 pre-order book valued at approximately $2.4 billion. | 高 | SV007, SV008 |
| CV006 | The $2.4 billion pre-order book implies roughly $70,600 of headline value per pre-order. | 中 | SV007, SV008 |
| CV007 | A $1.35 billion post-money valuation equals about 0.56 times the $2.4 billion disclosed pre-order book. | 中 | SV001, SV007, SV008 |
| CV008 | Recognized revenue and ARR are not publicly disclosed, so revenue multiples are not meaningful from public evidence. | 中 | SV001, SV017, SV018 |
| CV009 | The correct public-market multiple proxy is bookings or backlog potential, not recognized revenue. | 中 | SV003, SV005, SV007 |
| CV010 | Humanoid’s public homepage describes HMND 01 Alpha as an industrial humanoid platform with 15 kg payload, four-hour runtime, KinetIQ, and 29 degrees of freedom. | 中 | SV002 |
| CV011 | Humanoid says the Schaeffler agreement targets a four-digit number of wheeled units across Schaeffler facilities by 2032 under a Robot-as-a-Service model. | 中 | SV003 |
| CV012 | Forbes reported the Schaeffler deal as 1,000-plus robots and connected actuator demand to a possible path toward 100,000 units by 2031. | 高 | SV038, SV039 |
| CV013 | Humanoid says Bosch became a partner after a successful proof of concept and is part of the transition to scaled production. | 中 | SV005 |
| CV014 | Humanoid’s Siemens and NVIDIA factory pilot is evidence of industrial validation but not evidence of recognized revenue. | 中 | SV006 |
| CV015 | The visible Companies House HUMANOID LTD record shows public filings but does not disclose operating revenue, gross margin, customer deposits, or preference terms. | 高 | SV018, SV019 |
| CV016 | Humanoid’s $1.35 billion valuation implies about $6.75 million to $8.44 million per engineer or researcher if the team is approximately 160 to 200 people. | 中 | SV001, SV007, SV008 |
| CV017 | Figure AI announced more than $1 billion of Series C committed capital at a $39 billion post-money valuation. | 中 | SV020 |
| CV018 | Figure’s BMW deployment evidence provides a stronger factory proof point than valuation alone. | 中 | SV021, SV022 |
| CV019 | Apptronik announced more than $935 million of Series A capital after a new $520 million extension round. | 高 | SV023, SV024 |
| CV020 | TechCrunch reported Apptronik’s post-money valuation at about $5.3 billion, while CNBC framed the round at about a $5 billion valuation. | 高 | SV024, SV025 |
| CV021 | Agility Robotics’ GXO agreement is a live commercial deployment proof point for humanoid logistics workflows. | 中 | SV026 |
| CV022 | Agility Robotics announced a confidential Form S-4 submission connected to a proposed business combination with Churchill Capital Corp XI. | 中 | SV027 |
| CV023 | 1X’s NEO page and related coverage show a consumer-oriented robot offered with a purchase or subscription price, not an industrial factory valuation comp. | 中 | SV028, SV029 |
| CV024 | Unitree lists the G1 humanoid at a much lower hardware price anchor than Humanoid’s implied pre-order ASP. | 高 | SV031, SV032 |
| CV025 | Sanctuary AI positions its embodied AI systems for industrial work but does not publish a comparable valuation on its homepage. | 中 | SV030 |
| CV026 | Tesla Optimus is not separable from Tesla’s public-company valuation, so it is a strategic threat rather than a clean private-round comparable. | 中 | SV033, SV025 |
| CV027 | Goldman Sachs Research estimated the global humanoid robot market could reach $38 billion by 2035. | 中 | SV034 |
| CV028 | Morgan Stanley projected a long-term humanoid market scenario of about one billion robots and a $5 trillion ecosystem by 2050. | 中 | SV035 |
| CV029 | MarketsandMarkets estimated the humanoid robot market at $5.41 billion in 2026 and $50.27 billion by 2035. | 中 | SV036 |
| CV030 | Market-size evidence supports a large exit option but does not validate near-term revenue, gross margin, or order conversion for Humanoid. | 中 | SV034, SV035, SV036 |
| CV031 | McKinsey highlights humanoid supply-chain constraints and scaling requirements as central to commercialization, which supports a discount for execution risk. | 中 | SV037 |
| CV032 | IEEE Spectrum’s review of humanoid deployment history underscores the long gap between impressive demos and reliable work in real environments. | 中 | SV040 |
| CV033 | PitchBook, Dealroom, Crunchbase, and CB Insights provide company-profile coverage, but public profile pages do not disclose Humanoid’s preference stack or cash balance. | 中 | SV014, SV015, SV016, SV017 |
| CV034 | A bull case can underwrite material upside only if Humanoid converts pre-orders into tens of thousands of delivered or contracted RaaS robots by 2031. | 中 | SV003, SV005, SV012, SV038 |
| CV035 | A base case should assume partial conversion of bookings, slower delivery than the 100,000-unit ambition, and valuation support around the current round price. | 中 | SV003, SV005, SV007, SV037 |
| CV036 | A bear case should assume pre-order attrition, slower customer acceptance, and price compression from lower-cost competitors. | 中 | SV031, SV032, SV037, SV040 |
| CV037 | The chapter’s low/base/high valuation range is $0.5 billion, $1.35 billion, and $4.0 billion, reflecting downside option value, current round price, and a scaled-production success case. | 中 | SV001, SV007, SV020, SV024, SV037 |
| CV038 | At the current public evidence level, the recommended stance is track or research-more rather than buy because the valuation is bookings-backed but not revenue-backed. | 中 | SV001, SV007, SV018, SV037 |
| CV039 | The valuation stance is stretched: the price is below Figure and Apptronik but high for a pre-revenue industrial hardware company with private contract terms. | 中 | SV001, SV020, SV024, SV037 |
| CV040 | Strategic investors and partners reduce financing and manufacturing risk but do not remove customer enforceability, acceptance, warranty, or margin risk. | 中 | SV003, SV004, SV005, SV037 |
| CV041 | Schaeffler and Bosch are valuation-supportive because they provide demand, actuator, and manufacturing signals from industrial counterparties. | 中 | SV003, SV004, SV005 |
| CV042 | A fresh entry becomes more attractive if primary diligence verifies binding orders, deposits, gross margin, delivery milestones, and a clean entity structure. | 中 | SV003, SV018, SV019, SV037 |
| CV043 | A thesis-break trigger is failure to ship beta robots in Q4 2026 or to produce customer-accepted pilots with uptime evidence. | 中 | SV002, SV003, SV006 |
| CV044 | Another thesis-break trigger is evidence that the 34,000 pre-orders are mostly cancellable, unfunded, or materially lower than the reported $2.4 billion value. | 中 | SV005, SV007, SV008 |
| CV045 | A further thesis-break trigger is Unitree-style price compression that pulls industrial ASP below the economics needed for positive contribution margin. | 中 | SV031, SV032, SV037 |
| CV046 | RaaS could increase lifetime value and customer adoption if utilization and uptime are high, but it also increases fleet-financing and service-capital needs. | 中 | SV003, SV026, SV037 |
| CV047 | Humanoid’s current valuation is more attractive than Figure’s on headline valuation, but Figure has a much larger capital base and stronger disclosed BMW deployment proof. | 中 | SV020, SV021, SV022 |
| CV048 | Humanoid’s valuation is much lower than Apptronik’s reported $5 billion-plus valuation, but Apptronik has raised more than three times Humanoid’s total capital. | 中 | SV001, SV023, SV024 |
| CV049 | The valuation cannot be supported by a conventional DCF without private evidence of ASP, gross margin, acceptance timing, RaaS utilization, and reinvestment needs. | 中 | SV018, SV019, SV037 |
| CV050 | The investment KPI score should be high on market and strategic proof but low-to-medium on revenue quality, evidence completeness, and valuation margin of safety. | 中 | SV001, SV003, SV020, SV024, SV037 |
| CV051 | Figure’s company and careers pages support its status as a well-funded AI robotics competitor with an active hiring platform, but they do not disclose comparable revenue. | 中 | SV041, SV042 |
| CV052 | Apptronik’s homepage positions Apollo as an intelligent, mobile, dexterous humanoid robot, reinforcing its relevance as an industrial humanoid peer. | 中 | SV043 |
| CV053 | 1X’s about page positions the company around humanoid companions for homes, reinforcing that its valuation relevance differs from Humanoid’s industrial wedge. | 中 | SV044 |
| CV054 | Unitree and UBTECH official pages reinforce China-linked competitive breadth, adding adverse pressure to premium valuation and ASP assumptions. | 中 | SV045, SV046 |
| CV055 | Figure’s newsroom provides a live peer-milestone feed that can be refreshed for secondary-stock, BMW, and logistics evidence before a Humanoid entry decision. | 中 | SV047 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | Humanoid is creating the world’s leading, commercially scalable, and safe humanoid robots. |
| SO002 | Humanoid | Team - Humanoid | |
| SO003 | Humanoid | Humanoid Blog: Robotics News and Industry Insights | |
| SO004 | Humanoid | Technology - Humanoid | |
| SO005 | Humanoid | HMND 01 ALPHA WHEELED - Humanoid | Built in just 7 months, HMND 01 Alpha Wheeled is the fastest-developed humanoid. |
| SO006 | Humanoid | HMND 01 ALPHA BIPEDAL - Humanoid | |
| SO007 | Humanoid | Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots | Deployment begins in live German production environments in late 2026. |
| SO008 | Humanoid | Siemens and Humanoid bring Physical AI to the factory floor with NVIDIA | |
| SO009 | Humanoid | HMND 01 Alpha goes to work with SAP and Martur Fompak | |
| SO010 | tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | Humanoid... announced a $152 million Series A financing at a $1.35 billion post-money valuation. |
| SO011 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | |
| SO012 | Silicon Republic | UK-based Humanoid secures $152m Series A funding | |
| SO013 | SiliconANGLE | Humanoid raises $152M at $1.35B valuation to bring human-like robots into factories | |
| SO014 | Forbes | A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million | A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million |
| SO015 | Forbes Business Council | Artem Sokolov | Founder & CEO - Humanoid | Humanoid was founded in May 2024 with a clear focus. |
| SO016 | Unite.AI | Artem Sokolov, Founder of Humanoid – Interview Series | Today, he leads a team of over 130 professionals from some of the world’s leading tech companies. |
| SO017 | Humanoids Daily | Humanoid Founder Artem Sokolov on Pragmatic Path to Commercial Robotics | Humanoid announced in May 2025 the appointment of Jarad Cannon as its new Chief Technology Officer. |
| SO018 | AI Magazine | Artem Sokolov | Our headquarters is in London, with engineering satellites in Boston and Vancouver. |
| SO019 | TechTimes | Artem Sokolov: A Strategy-First Career Across Manufacturing and Robotics | |
| SO020 | Phemex | Humanoid Robotics Secures $152M Series A, Valued at $1.35B | |
| SO021 | Traders Union | Humanoid secures $152 million Series A as robotics investment accelerates | |
| SO022 | ChipSilicon | HMND 01 | Humanoid Robot | |
| SO023 | Aparobot | HMND 01 - Robot Details, Use Case and Specifications | |
| SO024 | CNET | I’ve Seen It With My Own Eyes: The Robots Are Here and Walking Among Us | The company has already signed around 25,000 preorders for HMND 01 and completed pilots with six Fortune 500 companies, he said. |
| SO025 | Humanoid | Careers - Humanoid | |
| SO026 | Dealroom.co | Humanoid AI company information, funding & investors | |
| SO027 | The Guardian | AI will transform the human job and enhance skills, says science minister | |
| SM001 | MarketsandMarkets | Humanoid Robot Market Size, Share, Latest Trends & Growth Analysis, 2026 - 2035 | The humanoid robot market size is estimated to be USD 5.41 billion in 2026 and is projected to reach USD 50.27 billion by 2035. |
| SM002 | Fortune Business Insights | Humanoid Robot Market Size, Share, & Growth Report [2034] | Organizations are prioritizing measurable productivity gains over experimental deployments. |
| SM003 | Grand View Research | Humanoid Robot Market Size & Share | Industry Report, 2030 | |
| SM004 | Precedence Research | Humanoid Robot Market Size to Hit USD 8.78 Billion by 2035 | |
| SM005 | Mordor Intelligence | Humanoids Market Size, Forecast Report (2026-2031) | Rapid cost declines in artificial intelligence hardware, demographic ageing in the G7 and China, and widening industrial labour gaps are converging. |
| SM006 | Goldman Sachs | The global market for humanoid robots could reach $38 billion by 2035 | The global market for humanoid robots could reach $38 billion by 2035. |
| SM007 | Morgan Stanley | Future State: 1 Billion Humanoids | By 2050, about 90% of humanoids, or about 930 million units, will likely be used for repetitive, simple, and structured work. |
| SM008 | International Federation of Robotics | Record of 4 Million Robots in Factories Worldwide | |
| SM009 | International Federation of Robotics | Global robotics industry joins forces to advocate for common automation strategies | |
| SM010 | International Federation of Robotics | New IFR position paper on humanoid robots published | The latest IFR position paper takes a closer look at humanoid robots und tries to separate vision and reality. |
| SM011 | Statista | Industrial robots worldwide - statistics & facts | |
| SM012 | McKinsey & Company | Turning humanoid supply chain constraints into billion-dollar wins | The strategic question is no longer whether humanoid robots will work but whether they can scale economically—and at what cost, speed, and reliability. |
| SM013 | RobotCentral | Goldman Sachs Predicts $38 Billion Humanoid Robot Market by 2035 – But Likely an Underestimate | |
| SM014 | Axis Intelligence | Humanoid Robot Statistics 2026: Market Size, Deployments, and Who is Actually Winning | |
| SM015 | ChinaBizInsider | Morgan Stanley Raises China Humanoid Robot 2026 Forecast to 50,000 Units | |
| SM016 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | |
| SM017 | Humanoid | Humanoid and Schaeffler Announce Strategic Technology Partnership | |
| SM018 | GXO Logistics | GXO conducting industry-leading pilot of human-centric robot | The robot is being tested on logistics tasks at SPANX’s facility in Flowery Branch, Ga. |
| SM019 | Occupational Safety and Health Administration | Robotics - Overview | There are currently no specific OSHA standards for the robotics industry. |
| SM020 | International Organization for Standardization | ISO 10218-1:2025 | ISO 10218-1 provides the safety requirements for the robot as a machine itself. |
| SM021 | Global Market Insights | Humanoid Robot Market Size, Forecasts Report 2026-2035 | |
| SM022 | Coherent Market Insights | Humanoid Robot Market Size and YoY Growth Rate, 2026-2033 | |
| SM023 | Allied Market Research | Humanoid Robot Market Size, Share, & Trends - 2031 | |
| SM024 | The Business Research Company | Humanoid Robot Market Size, Share and Growth Report, Forecast 2035 | |
| SM025 | SkyQuest Technology | Humanoid Robot Market Size, Share | Forecast Report [2033] | |
| SM026 | KBV Research | Humanoid Robot Market Size - Industry Trends to 2023-2030 | |
| SM027 | U.S. Bureau of Labor Statistics | Hand Laborers and Material Movers | |
| SM028 | U.S. Bureau of Labor Statistics | Production Occupations | |
| SM029 | World Economic Forum | The Future of Jobs Report 2025 | |
| SM030 | United Nations Department of Economic and Social Affairs | World Population Prospects 2024: Dataset | |
| SM031 | Deloitte | 2026 Manufacturing Industry Outlook | |
| SP001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | KinetIQ and KinetIQ Ascend form the foundation of our AI framework for end-to-end orchestration of humanoid robot fleets. |
| SP002 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | Humanoid announced a $152 million Series A financing at a $1.35 billion post-money valuation and $270 million total raised. |
| SP003 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | |
| SP004 | Humanoid | Humanoid and Schaeffler Announce Strategic Technology Partnership | Initial deployments are scheduled to begin with beta-stage robots in 2026-2027. |
| SP005 | Humanoids Daily | Humanoid Founder Artem Sokolov on Pragmatic Path to Commercial Robotics, Prioritizing Market-Ready Solutions | Humanoid is initially focusing on a wheeled version to tackle manipulation challenges more quickly. |
| SP006 | Figure AI | Figure | |
| SP007 | Figure AI | Figure Exceeds $1B in Series C Funding at $39B Post-Money Valuation | Figure exceeded more than $1 billion in committed capital through its Series C financing round at a $39 billion post-money valuation. |
| SP008 | Figure AI | F.02 Contributed to the Production of 30,000 Cars at BMW | Figure 02 contributed to the production of 30,000+ X3 vehicles and recorded 1,250+ operational hours. |
| SP009 | Figure AI | F.03 Arrives at BMW | |
| SP010 | Figure AI | Scaling Helix: a New State of the Art in Humanoid Logistics | Helix average handling time per package dropped from roughly 5.0 seconds to 4.31 seconds. |
| SP011 | Figure AI | Figure Announces Strategic Partnership with Brookfield | |
| SP012 | Figure AI | Figure Signs Agreement with Catalyst Brands to Scale Humanoid Operations | |
| SP013 | Apptronik | Apptronik - Home | |
| SP014 | Apptronik | Apptronik Closes Over $935 Million Series A | Apptronik announced a $520 million Series A-X funding round and over $935 million Series A total. |
| SP015 | Forbes | Apptronik Scores $935 Million, Hits Top 3 For Humanoid Robotics Funding | |
| SP016 | TechCrunch | Humanoid robot startup Apptronik has now raised $935M at a $5B+ valuation | TechCrunch learned that Apptronik’s post-money valuation is now about $5.3 billion. |
| SP017 | CNBC | Apptronik raises $520 million to beat Chinese humanoids, Tesla Optimus to market | |
| SP018 | Forbes | Apptronik Announces Robot Park, A 90,000 Square Foot Humanoid Data Factory, Teases New Robot | Apptronik CEO Jeff Cardenas said the industry today is essentially all prototypes. |
| SP019 | Business Insider | This $5.5 billion robotics startup built a school for humanoids | Most of the time, the robots are controlled remotely at Apptronik Robot Park. |
| SP020 | Agility Robotics | Humanoid Solutions | Agility | |
| SP021 | Agility Robotics | Digit Deployed at GXO in Historic Humanoid RAAS Agreement | |
| SP022 | Agility Robotics | GXO Signs Industry-First Multi-Year Agreement with Agility Robotics | GXO is deploying Digit robots and Agility Arc as part of a Robots-as-a-Service agreement. |
| SP023 | Agility Robotics | Agility Robotics Broadens Relationship with Amazon | |
| SP024 | Agility Robotics | Agility Robotics Announces Commercial Agreement with Toyota Motor Manufacturing Canada | |
| SP025 | Agility Robotics | Agility Robotics and Churchill Capital Corp XI Announce Confidential Submission of Draft Registration Statement | |
| SP026 | Agility Robotics | Digit Moves Over 100,000 Totes in Commercial Deployment | |
| SP027 | 1X Technologies | 1X | Home Robots | |
| SP028 | 1X Technologies | NEO Home Robot | |
| SP029 | 1X Technologies | Discover | 1X | |
| SP030 | Humanoids Daily | 1X Opens Pre-Orders for NEO Humanoid Robot at $20,000 | The journey there is not going to be perfect, and we are not going to pretend it is going to be. |
| SP031 | Humanoids Daily | 1X Partners with EQT to Deploy Up to 10,000 Humanoid Robots Across Global Industry | The announcement suggests this is an enabling agreement rather than a guaranteed purchase order. |
| SP032 | Boston Dynamics | Atlas Humanoid Robot | Boston Dynamics | |
| SP033 | Boston Dynamics | An Electric New Era for Atlas | |
| SP034 | Unitree Robotics | Humanoid robot G1 | Unitree Robotics | Unitree G1 price starts from US $13.5K and offers 23 to 43 joint motors. |
| SP035 | Unitree Robotics | Universal humanoid robot H1 | Unitree Robotics | |
| SP036 | Sanctuary AI | Embodied AI Robotics Systems for Industry | |
| SP037 | UBTECH | UBTECH: Humanoid Robot | AI Education Robot | Commercial Robot Solutions | |
| SP038 | IEEE Spectrum | Humanoid Robots Are Getting to Work | The answer seems to be 2024, when a handful of well-funded companies will be deploying their robots in commercial pilot projects. |
| SP039 | Humanoids Daily | A Formula One Pace: Tesla Optimus Director Teases Gen 3 Prototype | |
| SP040 | Humanoids Daily | A Humanoid Robot in Aisle 5? Unitree G1 Appears on Walmart Website for $21,600 | |
| SP041 | Humanoids Daily | NEURA Robotics Secures Record $1.4 Billion Series C, Led by Tether | |
| SP042 | Humanoids Daily | A Brain in the Alps: Flexion Raises $50M to Build the Android of Humanoids | |
| SP043 | Humanoids Daily | Fourier Intelligence Launches N1 Humanoid Robot with Open-Source Strategy | |
| SP044 | Humanoids Daily | UBTECH Targets 10-Fold Output Jump to 5,000 Humanoids in 2026 | |
| SP045 | Humanoids Daily | Tesla Optimus Production Delayed by China Rare Earth Export Controls | |
| SP046 | Humanoids Daily | The Humanoid Question: Ex-Tesla Optimus Lead Voices Factory Skepticism | |
| SI001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | |
| SI002 | Humanoid | Humanoid and Schaeffler Announce Strategic Technology Partnership | |
| SI003 | Humanoid | Humanoid Secures Partnership with Bosch Following a Successful POC | The agreement follows the successful completion of a joint proof of concept (POC) in March 2026. |
| SI004 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | Humanoid ... announced a $152 million Series A financing at a $1.35 billion post-money valuation. |
| SI005 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | |
| SI006 | Silicon Republic | UK-based Humanoid secures $152m Series A funding | |
| SI007 | SiliconANGLE | Humanoid raises $152M at $1.35B valuation to bring human-like robots into factories | |
| SI008 | Forbes | A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million | |
| SI009 | TechStartups | Humanoid raises $152M at $1.35B valuation to build AI-powered industrial humanoid robots | The London startup has now raised $270 million, secured 34,000 robot pre-orders worth $2.4 billion. |
| SI010 | The AI Insider | UK's Humanoid Raises $152M in Series A Funding at $1.35B Valuation for Humanoid Robots | |
| SI011 | Phemex | Humanoid Robotics Secures $152M Series A, Valued at $1.35B | |
| SI012 | Traders Union | Humanoid secures $152 million Series A as robotics investment accelerates | |
| SI013 | EU-Startups | New Unicorn! Humanoid secures €133 million at €1.1 billion valuation to scale industrial robotics and physical AI | |
| SI014 | The Economic Times | Robotics startup Humanoid raises $152 million Series A round at $1.35 billion valuation | |
| SI015 | Yahoo Finance | Robotics startup Humanoid raises $152 million Series A round at $1.35 billion valuation | |
| SI016 | Unite.AI | Humanoid Raises $152 Million to Scale Industrial Humanoid Robots | |
| SI017 | The Next Web | London robotics startup Humanoid raises $152M Series A, with Bosch set to manufacture its wheeled robots at scale | |
| SI018 | Robonomy | Humanoid and Schaeffler Sign 1,000-Robot Deal — A Path to 100,000 by 2031 | Most humanoid partnerships are press releases. This one is a purchase order plus an actuator supply contract. |
| SI019 | Forbes | Humanoid’s 1,000+ Robot Deal with Schaeffler Hints At 100,000 Units By 2031 | Over 1,000 humanoid robots by Humanoid ... will now be working for Schaeffler. |
| SI020 | John Koetsier | Humanoid’s 1,000+ Robot Deal with Schaeffler Hints At 100,000 Units By 2031 | |
| SI021 | Forbes | Humanoid’s New Deal: Bosch Will Build Its Robots With Schaeffler Parts | |
| SI022 | Tech My Money | Humanoid HMND 01: Bosch Leads Production | |
| SI023 | Global Banking & Finance Review | Humanoid Targets U.S. IPO by 2030 With $2.4B Pre-Order Revenue | |
| SI024 | Embodied Global | UK Startup Humanoid Secures 34,000 Pre-Orders, Considers IPO in 2029-2030 | |
| SI025 | 36Kr Europe | After Two-Year Entrepreneurship, Robotics Company Achieves $27 Billion in Revenue and Strikes Gold | 200 employees, two years of establishment, 34,000 orders, IPO - Is Humanoid a genius or a fraud? |
| SI026 | Companies House | All search results: HUMANOID | |
| SI027 | Companies House | HUMANOID LTD overview | Accounts for HUMANOID LTD show last accounts made up to 1 May 2025 and next accounts due by 31 January 2027. |
| SI028 | Companies House | HUMANOID LTD filing history | |
| SI029 | PitchBook | Humanoid AI company profile | |
| SI030 | Dealroom | Humanoid AI company information, funding & investors | |
| SI031 | Crunchbase | Humanoid AI organization profile | |
| SI032 | Tracxn | Humanoid AI profile lookup | |
| SI033 | Sifted | Humanoid robotics unicorn Series A lookup | |
| SI034 | Reuters | Robotics startup Humanoid raises $152 million Series A round lookup | |
| SI035 | Bloomberg | Humanoid raises $152 million at $1.35 billion valuation lookup | |
| SI036 | Financial Times | Humanoid robotics AI valuation lookup | |
| SI037 | Roland Berger | Humanoid robots 2026: The convergence moment | |
| SI038 | Roland Berger | Humanoid robots 2026 study PDF | |
| SI039 | K2 Capital | Humanoid Actuators Thematic Research | Why the component accounting for 30–50% of humanoid robot cost is the highest-conviction investment. |
| SI040 | Goldman Sachs | The global market for humanoid robots could reach $38 billion by 2035 | |
| SI041 | McKinsey & Company | Turning humanoid supply chain constraints into billion-dollar wins | |
| SI042 | Robotics & Automation News | Humanoid robots show clearer ROI, but commercial success depends on effective output | Commercial success depends on effective output. |
| SI043 | RoboChronicle | Humanoid Robot Cost in 2026: Price, Components, and What You’re Really Paying For | |
| SI044 | There Is A Robot For That | Humanoid Robot Price 2026: Best Cost & ROI Breakdown | |
| SI045 | TechCrunch | Humanoid robot startup Apptronik has now raised $935M at a $5B+ valuation | |
| SE001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | KinetIQ and KinetIQ Ascend form the foundation of our AI framework for end-to-end orchestration of humanoid robot fleets. |
| SE002 | Humanoid | Product - Humanoid | HMND 01’s hardware and software are designed using modular concept. |
| SE003 | Humanoid | HMND 01 ALPHA WHEELED - Humanoid | Built in just 7 months, HMND 01 Alpha Wheeled is the fastest-developed humanoid, delivering efficient performance across industrial use cases. |
| SE004 | Humanoid | HMND 01 ALPHA BIPEDAL - Humanoid | Developed in just five months and walking stably after only 48 hours of training. |
| SE005 | Humanoid | Technology - Humanoid | |
| SE006 | Humanoid | KinetIQ: AI Stack Overview - Humanoid | KinetIQ system architecture consists of 4 distinct layers covering different timescales and orchestration scopes. |
| SE007 | Humanoid | KinetIQ Ascend: Towards 100% Manipulation Reliability and Superhuman Speed - Humanoid | RL raised throughput by 42%... item handover throughput rose by 85% and success climbed from 80% to 98%. |
| SE008 | Humanoid | HMND 01 Alpha goes to work: Humanoid completes automotive manufacturing logistics POC with SAP and Martur Fompak | The POC focused on a real-world logistics picking workflow. |
| SE009 | Humanoid | Siemens and Humanoid bring Physical AI to the factory floor: deploying humanoids in industrial operations with NVIDIA | HMND 01 Alpha is purpose-built for industrial environments with advanced manipulation capabilities. |
| SE010 | Humanoid | Humanoid Secures Partnership with Bosch Following a Successful POC | Bosch will serve as Humanoid’s contract manufacturing partner. |
| SE011 | Humanoid | Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots | Schaeffler becomes preferred actuator supplier. |
| SE012 | Humanoid | Humanoid Deploys NVIDIA Technologies to Power Advanced Humanoid Robots | Humanoid is accelerating development by integrating NVIDIA Jetson Thor, NVIDIA Isaac Sim, and NVIDIA Isaac Lab. |
| SE013 | Humanoid | Deployment-Ready RL: Pitfalls, Lessons, and Best Practices | To be able to do this well, you do need a very strong mix of learning and control theory. |
| SE014 | Humanoid | Staff Control Software Engineer - Humanoid | Design and maintain modular, real-time control architectures supporting locomotion, manipulation, and teleoperation. |
| SE015 | Humanoid | Reinforcement Learning (RL) Engineer, Manipulation - Humanoid | Partner with teleoperations to collect trajectories in simulation for behavior cloning. |
| SE016 | ChipSilicon | HMND 01 | Humanoid Robot | ChipSilicon | 360° RGB cameras, depth sensors, 6D F/T sensors, haptic feedback, 6-mic array, and head display. |
| SE017 | Aparobot | HMND 01 - Robot Details, Use Case and Specifications | HMND 01 by Humanoid is a humanoid robot with 29 DoF, 15 kg payload, 2 m/s speed, and a 4 hours run time. |
| SE018 | RoboZaps | Hmnd 01 - Specifications & Deployment Status | Degrees of freedom 29 (excl. end-effectors); Runtime 3 h average run time; Max speed 1.5 m/s. |
| SE019 | Humanoid.guide | Humanoid.ai HMND 01 Specs & Price | HMND 01 is listed with 175 cm height, 70 kg weight, and 41 degrees of freedom, conflicting with newer disclosures. |
| SE020 | SAP | Martur Fompak International - SAP Innovation Awards 2026 | Martur Fompak International’s SAP Intelligent Humanoid transforms material replenishment from manual workflows to autonomous execution. |
| SE021 | SAP News Center | Martur Fompak Boosts Throughput & Efficiency with Intelligent Robotics Enabled by Joule and Embodied AI | Early results show increased throughput, fewer errors and a scalable, AI-driven intralogistics model. |
| SE022 | Siemens | Siemens and Humanoid bring Physical AI to the factory floor: deploying humanoids in industrial operations with NVIDIA | Humanoid’s HMND 01 wheeled Alpha humanoid robot, built using the NVIDIA physical AI stack, has been successfully tested in operations at Siemens’ electronics factory in Erlangen. |
| SE023 | The Robot Report | Humanoid completes live HMND PoC with SAP and Martur Fompak | The project tested the HMND 01 Alpha Wheeled robot in a working production facility and set clear benchmarks for future deployment. |
| SE024 | Humanoids Daily | From ERP to the Edge: Humanoid and SAP Bridge the Gap in Martur Fompak Logistics Trial | The POC established several key performance benchmarks for the Alpha hardware: Payload: the robot operated with an 8 kg dual-arm payload limit. |
| SE025 | Logistics Manager | Humanoid, SAP and Martur Fompak complete logistics robot test | HMND 01 robots operated as a flexible fleet, directly tasked through SAP Business AI as autonomous agents. |
| SE026 | NVIDIA | Jetson AGX Orin for Next-Gen Robotics | Jetson AGX Orin is positioned by NVIDIA for next-gen robotics and autonomous machines. |
| SE027 | NVIDIA | Jetson Thor | Advanced AI for Physical Robotics | Jetson Thor is described by NVIDIA as advanced AI for physical robotics. |
| SE028 | NVIDIA Developer | Isaac GR00T - Generalist Robot 00 Technology | Isaac GR00T comprises open data, a robot foundation model, simulation frameworks, middleware, CUDA-X runtime libraries, and Jetson Thor. |
| SE029 | NVIDIA Developer | Isaac Sim - Robotics Simulation and Synthetic Data Generation | NVIDIA Isaac Sim is an open source reference framework for robotics simulation, testing, and synthetic data generation. |
| SE030 | IEEE Spectrum | Humanoid Robots: The Scaling Challenge | Market requirements for humanoid robots include battery life, reliability, and safety; downtime and safety standards remain critical hurdles. |
| SE031 | Robotics & Automation News | The state of humanoid robotics: assessing the capabilities, limitations, and commercial potential of leading platforms | Platforms are impressively capable in demos and narrow tasks, but far from general-purpose autonomy. |
| SE032 | The Robot Report | IEEE survey sheds light on how AI and humanoids will affect robotics in 2026 | |
| SE033 | Unite.AI | Artem Sokolov, Founder of Humanoid – Interview Series | One key question is the balance between autonomy and control. How independent should robots be? |
| SE034 | Humanoids Daily | Humanoid Founder Artem Sokolov on Pragmatic Path to Commercial Robotics, Prioritizing Market-Ready Solutions | The company is adopting a staged approach to robot autonomy: Shared Autonomy; Increased Independence (2026-2027); Full Autonomy (Late 2027). |
| SE035 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | The company will focus on Beta version robots in Q4 2026, mass manufacturing of wheel-based humanoid robots, and AI/software development including KinetIQ. |
| SU001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | Homepage lists SAP, NVIDIA, Schaeffler, Siemens, Bosch and AWS logos and links to Schaeffler, Siemens and SAP/Martur Fompak case studies. |
| SU002 | Humanoid | Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots | Humanoid states it signed a binding, phased deployment and supply agreement with Schaeffler and that first systems go live in Germany before the end of 2026. |
| SU003 | Humanoid | Humanoid and Schaeffler Announce Strategic Technology Partnership | The January partnership described hundreds of Humanoid robots entering Schaeffler production facilities over five years and joint development of actuator technologies. |
| SU004 | Unite.AI | Humanoid Secures Major Manufacturing Deal With Schaeffler to Deploy Thousands of Robots | Unite.AI reports the Schaeffler agreement moves beyond pilot demonstrations into long-term industrial deployment beginning in late 2026. |
| SU005 | Robotics & Automation News | Humanoid and Schaeffler strike deal to deploy thousands of humanoid robots in factories | Robotics & Automation News repeats that the agreement is binding and phased and covers integration into live manufacturing operations. |
| SU006 | The AI Insider | Humanoid to Deploy Thousands of Humanoid Robots at Schaeffler Plants | The AI Insider reports a four-digit deployment by 2032, starting at Herzogenaurach and Schweinfurt in late 2026. |
| SU007 | Humanoid Robotics Technology | Humanoid to Deploy Thousands of Humanoid Robots on Schaeffler Factory Floors | Humanoid Robotics Technology describes one of the largest disclosed humanoid robot rollouts ever announced. |
| SU008 | Humanoid | Humanoid Secures Partnership with Bosch Following a Successful POC | Humanoid says its Bosch agreement follows a March 2026 joint proof of concept in a complex industrial workflow and marks transition to scaled production. |
| SU009 | Robotics & Automation News | Humanoid secures partnership with manufacturing giant Bosch following a successful proof of concept | Robotics & Automation News reports the Bosch POC focused on transferring boxes between a conveyor and trolley in intralogistics. |
| SU010 | Tech My Money | Humanoid HMND 01: Bosch Leads Production | Tech My Money reports Bosch will manufacture HMND 01 for Europe and Schaeffler will supply more than half of actuator demand through 2031. |
| SU011 | Humanoid Robotics Technology | Humanoid Announces Partnership with Bosch | Humanoid Robotics Technology says the Bosch agreement marks transition to scaled production after the POC. |
| SU012 | Forbes | Humanoid’s New Deal: Bosch Will Build Its Robots With Schaeffler Parts | Forbes characterizes Humanoid as taking a partner-led path: Bosch will build the robots and Schaeffler will supply components. |
| SU013 | Humanoid Guide | Humanoid HMND 01 Bosch manufacturing deal in Europe | Humanoid Guide says the Bosch and Schaeffler deals together outline a supply-chain and production path for industrial scale. |
| SU014 | PR Newswire | Siemens and Humanoid bring Physical AI to the factory floor: deploying humanoids in industrial operations with NVIDIA | The Siemens announcement says HMND 01 Alpha was tested at Siemens Erlangen and met 60 tote moves per hour, more than eight hours uptime, and above 90 percent pick-and-place success. |
| SU015 | Robotics & Automation News | Siemens, Nvidia and Humanoid partner to bring physical AI into factory operations | Robotics & Automation News reports HMND 01 Alpha was successfully tested in operations at Siemens electronics factory in Erlangen. |
| SU016 | NVIDIA | Siemens Accelerates Innovation With NVIDIA Omniverse and AI | NVIDIA positions Siemens as building an industrial AI operating system using NVIDIA Omniverse and AI for manufacturing workflows. |
| SU017 | SAP News Center | How SAP Brings Business AI into the Physical World: New Robotics Pilots Across Warehouse, Manufacturing, and Inspection Use Cases | SAP describes embodied AI pilots that connect robots to SAP business context for warehouse, manufacturing and inspection use cases. |
| SU018 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | Tech.eu reports Humanoid has partnerships with SAP, NVIDIA, Bosch and Siemens and signed the industry’s largest publicly announced commercial agreement with Schaeffler. |
| SU019 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | Forbes reports Humanoid raised $152 million at a $1.35 billion valuation and is Europe’s newest robot unicorn. |
| SU020 | Silicon Republic | UK-based Humanoid secures $152m Series A funding | Silicon Republic reports ongoing partnerships with SAP, Nvidia, Bosch and Siemens and Schaeffler large-scale deployment. |
| SU021 | SiliconANGLE | Humanoid raises $152M at $1.35B valuation to bring human-like robots into factories | SiliconANGLE reports the Series A proceeds are intended to bring human-like robots into factories. |
| SU022 | VaaSBlock | Humanoid Robotics 2026: Figure, Optimus, 1X Reality | The pilots are real, but the gap between current capability and the autonomous, general-purpose humanoid worker that the marketing narrative implies remains significant. |
| SU023 | Physical AI Guide | Humanoid Robots in 2026: What Is Actually Available, Piloting, or Deployed? | Physical AI Guide distinguishes orderable or announced robots from broadly deployed products and says announcements or feasibility plans are not active pilots. |
| SU024 | IoT Digital Twin PLM | Humanoid Robots Hit the Factory Floor: 2026 Analysis | IoT Digital Twin PLM cautions readers to treat every figure as reported by vendors or press, not independently audited. |
| SU025 | Humanoids Daily | Humanoid Founder Artem Sokolov on Pragmatic Path to Commercial Robotics, Prioritizing Market-Ready Solutions | Humanoids Daily reports Sokolov prioritized proof-of-concept deployments to lay groundwork for customer applications. |
| SU026 | Humanoid | HMND 01 Alpha goes to work: Humanoid completes automotive manufacturing logistics POC with SAP and Martur Fompak | Humanoid says SAP and Martur Fompak completed a live production logistics proof of concept using HMND 01 Alpha Wheeled. |
| SU027 | Humanoid | Siemens and Humanoid bring Physical AI to the factory floor: deploying humanoids in industrial operations with NVIDIA | Humanoid says HMND 01 Alpha completed tote-handling tasks at Siemens Erlangen with 60 moves per hour and above 90 percent pick-and-place success. |
| SU028 | Forbes | Humanoid’s 1,000+ Robot Deal with Schaeffler Hints At 100,000 Units By 2031 | Forbes reports phase one runs from December 2026 through June 2027 across two Schaeffler German sites and describes the deal as Robot-as-a-Service. |
| SU029 | The Robot Report | Schaeffler to deploy hundreds of Humanoid robots in its factories | The Robot Report covered the January Schaeffler-Humanoid plan to deploy hundreds of robots in factories. |
| SU030 | The Next Web | London robotics startup Humanoid raises $152M Series A, with Bosch set to manufacture its wheeled robots at scale | The Next Web says the claimed 34,000 robot pre-orders worth nearly two and a half billion dollars have not been independently verified. |
| SR001 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | Up to four hours of uninterrupted operation to minimize downtime. |
| SR002 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | $152 million Series A financing at a $1.35 billion post-money valuation. |
| SR003 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | |
| SR004 | The Next Web | London robotics startup Humanoid raises $152M Series A, with Bosch set to manufacture its wheeled robots at scale | The company claims 34,000 robot pre-orders worth nearly two and a half billion dollars, a figure that has not been independently verified. |
| SR005 | Silicon Republic | UK-based Humanoid secures $152m Series A funding | |
| SR006 | SiliconANGLE | Humanoid raises $152M at $1.35B valuation to bring human-like robots into factories | |
| SR007 | Unite.AI | Humanoid Raises $152 Million to Scale Industrial Humanoid Robots | 34,000 robot pre-orders, representing a commercial pipeline valued at approximately $2.4 billion. |
| SR008 | Humanoid | Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots | Humanoid ... has signed a binding, phased deployment and supply agreement with Schaeffler. |
| SR009 | Humanoid | Humanoid Secures Partnership with Bosch Following a Successful POC | The agreement follows the successful completion of a joint proof of concept (POC) in March 2026. |
| SR010 | Forbes | Humanoid’s New Deal: Bosch Will Build Its Robots With Schaeffler Parts | |
| SR011 | Forbes | A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million | A Former Russian Jewelry Magnate’s Humanoid Robot Startup Is In Talks To Raise $200 Million. |
| SR012 | Humanoid | Team - Humanoid | |
| SR013 | International Organization for Standardization | ISO 10218-1:2025 | ISO 10218-1 provides the safety requirements for the robot as a machine itself. |
| SR014 | International Organization for Standardization | ISO/TS 15066:2016 | Technical Specification: Robots and robotic devices — Collaborative robots. |
| SR015 | Occupational Safety and Health Administration | Robotics - Overview | Provides references that aid in recognizing hazards related to robotics in the workplace. |
| SR016 | European Commission | CE marking | Manufacturers must carry out conformity assessment, set up the technical file, issue the EU declaration of conformity and affix the CE marking. |
| SR017 | U.S. Treasury Office of Foreign Assets Control | Russian Harmful Foreign Activities Sanctions | OFAC-BIS Alert: Impact of Sanctions and Export Controls on Russia’s Military-Industrial Complex. |
| SR018 | UK Government | Export controls: military goods, software and technology | An export licence is needed before the export of controlled military goods, software and technology and items on the UK dual-use list. |
| SR019 | IEEE Spectrum | That Awesome Robot Demo Could Have a Human in the Loop | That Awesome Robot Demo Could Have a Human in the Loop. |
| SR020 | IEEE Spectrum | Humanoid Robots Hit a Turning Point as Their Brains Catch Up | |
| SR021 | IEEE Spectrum | Humanoid Robots Are Getting to Work | A team of experienced engineers controlled each robot, and overhead safety tethers kept them from falling over. |
| SR022 | IEEE Spectrum | Sanctuary’s Humanoid Robot Is for General-Purpose Autonomy | |
| SR023 | Ars Technica | The skeptic’s guide to humanoid robots going viral on the Internet | The skeptic’s guide to humanoid robots going viral on the Internet. |
| SR024 | Vox | Are humanoid robots all hype? | Humanoid robots have been hyped as the future of everything, from completing household chores to caring for elders to doing the dirty work on the factory floor. |
| SR025 | Los Angeles Times | Commentary: A leading roboticist punctures the hype about self-driving cars, AI chatbots and humanoid robots | Having ideas is easy. Turning them into reality is hard. |
| SR026 | Compare Robots | Humanoid robots in 2026: hype vs. reality | Many demos are teleoperated (a human is puppeteering). |
| SR027 | RoboHorizon | Forget Humanoid Hype: ICRA 2026 Reveals the Real Global Robot Wars | Spectacular prototypes are nice, but the real prize is economic sovereignty. |
| SR028 | Unitree Robotics | Humanoid robot G1_Humanoid Robot Functions_Humanoid Robot Price | This product is a civilian robot. We kindly request that all users refrain from making any dangerous modifications. |
| SR029 | NVIDIA | NVIDIA Jetson Thor | Jetson Thor gives you the performance you need for real-time control and multi-sensor processing. |
| SR030 | NVIDIA Developer | NVIDIA Jetson AGX Orin | |
| SR031 | EUR-Lex | Regulation - 2023/1230 - EN | Regulation (EU) 2023/1230 ... on machinery and repealing Directive 2006/42/EC. |
| SR032 | Bureau of Industry and Security | Deemed exports | A deemed export refers to the sharing or release of controlled technology or source code to a foreign person within the United States. |
| SR033 | Bureau of Industry and Security | Export Compliance Programs (ECPs) | ECPs are a series of procedures and tools that facilitate compliance with export controls to mitigate the risk of export violations. |
| SR034 | Bureau of Industry and Security | Reexports and exports from abroad | A U.S.-origin item subject to the EAR, remains subject to the EAR wherever located. |
| SV001 | Humanoid | Humanoid Raises $152 Million at $1.35 Billion Post-Money Valuation, Becoming Europe's First Pure-Play Humanoid Robotics Unicorn | Humanoid announced a $152 million Series A financing at a $1.35 billion post-money valuation and said total raised to date was $270 million. |
| SV002 | Humanoid | Humanoid: Pioneering Humanoid Robots for the Future | |
| SV003 | Humanoid | Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots | The agreement targets deployment of a four-digit number of wheeled units across Schaeffler facilities by 2032 under a Robot-as-a-Service model. |
| SV004 | Humanoid | Humanoid and Schaeffler Announce Strategic Technology Partnership | |
| SV005 | Humanoid | Humanoid Secures Partnership with Bosch Following a Successful POC | |
| SV006 | Humanoid | Siemens and Humanoid bring Physical AI to the factory floor | |
| SV007 | Tech.eu | UK robotics startup Humanoid hits $1.35B valuation with $152M Series A | |
| SV008 | Forbes | Humanoid Raises $152 Million At $1.35 Billion Valuation: Europe’s Newest Robot Unicorn | |
| SV009 | Silicon Republic | UK-based Humanoid secures $152m Series A funding | |
| SV010 | SiliconANGLE | Humanoid raises $152M at $1.35B valuation to bring human-like robots into factories | |
| SV011 | EU-Startups | New Unicorn! Humanoid secures €133 million at €1.1 billion valuation | |
| SV012 | Reuters | Robotics startup Humanoid raises $152 million Series A round at $1.35 billion valuation | |
| SV013 | Bloomberg | Humanoid Raises $152 Million at $1.35 Billion Valuation | |
| SV014 | PitchBook | Humanoid company profile | |
| SV015 | Dealroom | Humanoid AI company information, funding & investors | |
| SV016 | Crunchbase | Humanoid AI organization profile | |
| SV017 | CB Insights | Humanoid AI - Products, Competitors, Financials, Employees, Headquarters Locations | |
| SV018 | Companies House | HUMANOID LTD overview - Find and update company information | |
| SV019 | Companies House | HUMANOID LTD filing history - Find and update company information | |
| SV020 | Figure AI | Figure Exceeds $1B in Series C Funding at $39B Post-Money Valuation | |
| SV021 | Figure AI | F.02 Contributed to the Production of 30,000 Cars at BMW | |
| SV022 | Figure AI | Scaling Helix: a New State of the Art in Humanoid Logistics | |
| SV023 | Apptronik | Apptronik Closes Over $935 Million Series A with New $520 Million Extension Round | |
| SV024 | TechCrunch | Humanoid robot startup Apptronik has now raised $935M at a $5B+ valuation | |
| SV025 | CNBC | Apptronik raises $520 million to beat Chinese humanoids, Tesla Optimus to market | |
| SV026 | Agility Robotics | GXO Signs Industry-First Multi-Year Agreement with Agility Robotics | |
| SV027 | Agility Robotics | Agility Robotics and Churchill Capital Corp XI Announce Confidential Submission of Draft Registration Statement on Form S-4 | |
| SV028 | 1X | NEO Home Robot | |
| SV029 | Humanoids Daily | 1X Opens Pre-Orders for NEO Humanoid Robot at $20,000 | |
| SV030 | Sanctuary AI | Embodied AI Robotics Systems for Industry | |
| SV031 | Unitree | Humanoid robot G1 functions and price | Unitree lists the G1 humanoid robot with a price starting at US $13.5K, a low hardware anchor for premium humanoid ASP assumptions. |
| SV032 | Humanoids Daily | A Humanoid Robot in Aisle 5? Unitree G1 Appears on Walmart Website for $21,600 | The listing price of $21,600 makes advanced humanoid hardware accessible through a mainstream consumer channel and pressures premium pricing assumptions. |
| SV033 | U.S. Securities and Exchange Commission | Tesla, Inc. Form 10-K for fiscal year 2024 | |
| SV034 | Goldman Sachs | The global market for humanoid robots could reach $38 billion by 2035 | |
| SV035 | Morgan Stanley | Future State: 1 Billion Humanoids | |
| SV036 | MarketsandMarkets | Humanoid Robot Market Size, Share, Latest Trends & Growth Analysis, 2026 - 2035 | |
| SV037 | McKinsey & Company | Turning humanoid supply chain constraints into billion-dollar wins | Humanoid robots are approaching an inflection point, but supply-chain constraints and scaling requirements remain central commercialization risks. |
| SV038 | Forbes | Humanoid’s 1,000+ Robot Deal with Schaeffler Hints At 100,000 Units By 2031 | |
| SV039 | John Koetsier | Humanoid’s 1,000+ Robot Deal with Schaeffler Hints At 100,000 Units By 2031 | |
| SV040 | IEEE Spectrum | Humanoid Robots Are Getting to Work | The article recalls advanced humanoid robots struggling through DARPA Robotics Challenge tasks, underscoring commercialization difficulty. |
| SV041 | Figure AI | Company | Figure | |
| SV042 | Figure AI | Careers | Figure | |
| SV043 | Apptronik | Apptronik - Home | |
| SV044 | 1X | About | |
| SV045 | Unitree Robotics | Unitree Robotics | Robot Dog_Quadruped_Humanoid Robotics Company | Unitree positions itself as a mass-market robotics company, reinforcing competition from lower-cost China-linked hardware players. |
| SV046 | UBTECH Robotics | UBTECH: Humanoid Robot | AI Education Robot | Commercial Robot Solutions | UBTECH publicly markets humanoid and commercial robot solutions, adding another China-linked competitor to the valuation risk set. |
| SV047 | Figure AI | News | Figure |