Robot Era
一家融资速度很快的中国人形机器人公司,物流落地信号可信,但收入转化、经营经济性和轮次结构仍披露不足。
Robot Era 的公开部署和融资证据强于多数人形机器人创业公司,但商业化质量和经济披露仍落后于估值叙事,当前价格只有在继续尽调后才具备投资性。
封面要素
公司概况
Robot Era 是一家总部位于北京的具身智能与人形机器人公司,由清华教授 Jianyu Chen 于 2023 年 8 月创立。公开资料支撑其模块化产品栈:L7 全尺寸人形机器人、M7 上半身平台、Q5 轮式人形机器人,以及 XHAND 灵巧手系列;近期商业化最清晰的场景是结构化物流和工业流程。所提供信源包还显示,公司融资速度异常快,2026 年 3 月估值标记超过 RMB 10 billion,累计订单超过 RMB 500 million,已落地 10 多个物流中心,并有国际需求信号。核心尽调限制不是缺少动能,而是财务透明度不足:公开证据仍未清楚披露已确认收入、毛利率、现金跑道、客户集中度或最终轮次经济条款。
- 成立时间
- 2023-08-01
- 创始人
- Jianyu Chen
- 创立地点
- Beijing, China
- 总部
- Beijing, China
- 产品
- 人形机器人与具身智能技术栈,围绕 L7 全尺寸人形机器人、M7 上半身机器人、Q5 轮式人形机器人、XHAND 灵巧手,以及用于仓储和工业任务执行的 ERA-42 VLA 软件闭环展开。
- 客户
- 面向物流、制造业和部分商业服务环境中的企业运营方;这些客户需要在为人类设计的空间里,用灵活的抓取、分拣、搬运和自动化流程解决问题。
- 商业模式
- 向企业自动化用例销售或部署人形机器人、灵巧手模块和具身 AI 工作流系统,变现很可能来自硬件销售、系统部署,以及相关软件或集成服务。
- 阶段
- Series A+
- 融资情况
- 公开来源显示,Robot Era 融资路径很快:从 2025 年 7 月 Series A,到 2025 年 11 月 Series A+,再到 2026 年 3 月估值超过 RMB 10 billion 的 RMB 1 billion 战略轮,随后 2026 年又完成一轮超过 $200 million 的融资;但确切顺序和稀释时间线在公开资料中仍只部分对齐。
执行摘要
主要优势
- 强势产业投资人和合作伙伴阵容,帮助 Robot Era 快速融资,并在物流和制造场景拿到早期企业准入。
- 对一家年轻人形机器人公司而言,公开商业化证据罕见具体,包括累计订单超过 RMB 500 million、落地 10 多个物流中心,以及千台级交付说法。
- 模块化 L7、M7、Q5 和 XHAND 产品栈,加上 ERA-42 工作流叙事,显示公司不只是做单台 demo 机器人,而是在搭更宽的平台。
- 中国的政策支持、自动化深度和供应链优势,让公司的本土市场成为最可信的早期人形机器人部署环境之一。
主要风险
- 公开材料仍未披露确认收入、毛利率、现金跑道、客户集中度或股权结构条款,投资人无法清晰判断当前估值。
- 最强商业证据仍集中在结构化物流环境;Robot Era 能否在更大范围持续交付安全、可靠、高 ROI 部署,仍是开放问题。
- 跨境扩张会暴露在中美技术限制、采购摩擦,以及碎片化全球安全与合规体系之下。
- 中国和全球竞争都很激烈,Unitree、UBTECH、AgiBot、Figure 等对手在价格、规模或披露上提供了更强参照。
未决问题
- 按收入流拆分的经审计收入、backlog 转化、实际 ASP、毛利率、现金消耗和现金跑道。
- 2026 年 3 月战略轮及后续 2026 年融资的确切顺序、定价、稀释和投资人保护条款。
- 客户集中度、续约行为、合同期限,以及生产部署与研究或开发者交付之间的拆分。
- 面向跨国企业部署的 Robot Era 特定 uptime、事故、认证、网络安全和机群质量数据。
目录
01公司概况
1.1 身份、总部与公司定位
Robot Era 把自己定位为扎根北京清华生态的人形机器人和具身智能公司,而不是狭义零部件供应商。公司官网和应用资料包是最强的运营基地证据,里面出现了清华科技园地址、营销邮箱和公开电话。这一点重要,因为它把公司锚定在真实商业足迹上,而不是纯学术项目。独立创始人访谈包补充了成立时间:KrASIA 称 Robot Era 创立于 2023 年 8 月,这与一家快速扩张、异常迅速抵达大额融资节点的年轻公司叙事相符。同样关键的是,叙事并非「只做硬件」。Chen 描述的是脑与身体结合的策略:模块化、类似 Lego 的硬件,加上一层 ERA-42 VLA 软件;CES 2026 资料也把公司放在更广泛的具身智能产品家族里。综合来看,已审阅资料支持这样一个身份:一家总部位于北京、源于研究体系的商业机器人公司,有模块化平台野心,也有足够公开运营细节,可与隐身实验室项目区分开。[CO001, CO002, CO003, CO007, CO008, CO011]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 注意事项 |
|---|---|---|---|---|
| 成立 | August 2023 | 2023-08-01 | 中 | 由 KrASIA 访谈报道支撑,而不是由所给资料包中的官方注册文件支撑 |
| 总部 / 运营基地 | 北京海淀区清华科技园 VIA Building 7F | 当前官网资料包 | 高 | 运营地址公开;法律实体注册细节未公开 |
| 联系渠道 | MKT@robotera.com;电话 400 900 8798 | 当前官网资料包 | 中 | 仅有公开联系方式;未披露部门名册 |
| 创始人 | Jianyu Chen | 当前公开记录 | 中 | 公开资料包围绕一位点名创始人展开;更广泛管理梯队不清晰 |
| 2025 Series A | 近 RMB500 million | July 2025 | 中 | 基于创始人访谈;完整条款清单和投资团未公开 |
| 2025 Series A+ | 近 RMB1 billion(约 $140 million) | November 2025 | 高 | 官方和独立报道在规模及领投车企上相互吻合 |
| 战略轮估值标记 | > RMB10 billion | March 2026 | 高 | 公开公告数字;股权结构表和证券条款仍为私有 |
| 2026 年新融资轮 | > $200 million | April-May 2026 | 高 | 公开记录没有完全调和其与 2026 年 3 月战略轮之间的先后关系 |
| 累计订单 | > RMB500 million | 2025-2026 公开记录 | 高 | 公司和媒体来源一致,但待转化订单向收入转化未披露 |
| 海外订单占比 | 50% | 2025-2026 公开记录 | 高 | 占比出现在公司背书和媒体报道中,但国家组合未披露 |
| 自研部件比例 | > 95% | 2026 年融资报道 | 高 | 公司说法;所给资料包中没有独立拆解证据 |
| 交付规模目标 | 2026 年第二季度千台级交付 | Q2 2026 | 中 | 前瞻性运营说法,而不是已完全验证的交付台数 |
本表有意混合官方公告和独立报道,因为 Robot Era 的公开披露分散在访谈、新闻稿和媒体报道中。融资行保留 2026 年 3 月战略轮与随后报道的 >$200 million 融资之间尚未解决的时间线问题。
[CO001, CO002, CO003, CO004, CO011, CO012]1.2 创始人画像、领导层可见度与治理不透明
所提供资料中,公开领导层可见度主要集中在 Jianyu Chen。独立报道将他识别为 Robot Era 创始人,清华官方 IIIS 页面则确立了这个角色背后的技术履历:助理教授身份、ISR Lab 关联、清华本科,以及 2020 年在 UC Berkeley 师从 Masayoshi Tomizuka 获得博士学位。这是有意义的创始人-市场匹配信号,因为它把公司连接到顶尖控制、机器人和具身 AI 研究,而不是纯财务包装周期。反面是,公开记录在 Chen 之外仍很薄。The Wire China 补充了年龄背景并重复了一项重要客户说法,但所提供信源包没有呈现更完整的高管梯队、现任董事会名单或详细治理图。尽调上,这意味着本章可以有力支持创始人可信度,但还无法支持组织深度或决策权分布。在更完整的高管和治理资料私下提供之前,关键人集中度看起来偏高。[CO004, CO005, CO006, CO009, CO010, CO028]
| 领导者 | 公开支撑的角色 | 胜任证据 | 重要性 | 当前可见缺口 |
|---|---|---|---|---|
| Jianyu Chen | 创始人;Robot Era 的公开门面 | Tsinghua IIIS 助理教授、ISR Lab 负责人、Tsinghua 2015 年本科、UC Berkeley 2020 年博士,导师 Masayoshi Tomizuka | 在机器人、控制和具身 AI 上具备很强的技术创始人-市场匹配 | 公开记录仍未显示其完整内部管理跨度、董事会角色或持股比例 |
| 更广泛高管 / 董事会梯队 | 所给公开资料包未清晰披露 | 已审阅来源中没有扎实浮现其他现任高管或董事 | 指向围绕 Chen 的高关键人物集中度,以及有限的公开治理可见度 | 需要私有组织架构图、董事会名册和股权结构治理文件 |
所给资料包强力支持 Jianyu Chen 的背景,但不支持完整领导层名册。第二行是有意设置的披露状态行,不是点名高管传记。
[CO004, CO005, CO006, CO009, CO010]1.3 融资节奏、估值信号与利益相关方图谱
资料包里最清晰的公司层面信号是融资速度。KrASIA 报道,2025 年 7 月 Series A 近 RMB500 million,由 CDH VGC 和 Haier 领投;2025 年 11 月 A+ 轮近 RMB1 billion,由 Geely 领投、BAIC 参与。Yicai 补充称,当时已披露资本达到 CNY1.8 billion,订单已超过 CNY500 million。2026 年记录进一步加速:PRNASIA 和 Humanoids Daily 称,2026 年 3 月一轮 RMB1 billion 战略融资把估值推至 RMB10 billion 以上;另一个 2026 年 4-5 月信源群则报道,SF Group 领投、HSG 和 IDG Capital 参与的新一轮融资超过 $200 million。这条序列显示,汽车、物流和头部创投都给了强力产业与投资支持,但也带来对账问题。公开资料没有充分解释 2026 年 3 月战略轮、2026 年 5 月轮次以及外部独角兽榜单标记,如何放进一条清晰的稀释时间线。结论是:获得资本和股东方质量为正面信号,但融资结构仍不完整。[CO012, CO013, CO014, CO015, CO016, CO017]
| 利益相关方 | 在故事中的角色 | 证据支撑的重要性 | 尽调问题 |
|---|---|---|---|
| CDH VGC | 2025 年 7 月 Series A 领投方 | 所给资料包中最早点名的规模化资本提供方 | 确认出资金额、治理权利,以及后续是否继续参与 2026 年融资轮 |
| Haier Capital / Haier | Series A 联合领投方和战略支持方 | 产业验证,说明公司早期吸引力不只限于纯财务 VC | 澄清商业联系、试点范围,以及任何制造协作条款 |
| Geely | 2025 年 11 月 Series A+ 领投方 | 车企对具身智能商业化的强背书 | 索取董事会权利、战略权利,以及采购或部署承诺 |
| BAIC | 2025 年 11 月 Series A+ 参投方 | 为股权结构叙事增加第二家车企,并增强产业可信度 | 确认轮次规模、持股,以及任何联合开发安排 |
| SF Group | 2026 年 >$200 million 融资轮领投方,并且是物流中心部署中的企业伙伴 | 把融资与真实商业 / 物流用例连接起来 | 将投资理由与运营收入贡献和部署经济性拆开 |
| China Post | 物流中心点名部署交易对手 | 表明采用不止一个私人物流客户 | 索取场站数量、合同金额、续约条款和部署生产率指标 |
| UNIDO | 国际机构合作伙伴 | 将公司故事从中国商业化扩展到全球工业转型信号 | 澄清合作是象征性、试点阶段,还是能产生收入 |
本地图优先列出与融资、产业验证和商业化最相关的公开可见投资人和战略交易对手。它不是完整股东名册或客户名单。
[CO012, CO013, CO016, CO019, CO020, CO030]快速指标把融资规模、估值、商业化信号和未解决的执行问题放在一起。
融资和运营数字混合了官方稿与独立报道;未解决的时间线和缺失的经济性,应视为尽调提示,而不是已定指标。
[CO014, CO015, CO017, CO019, CO021, CO022]1.4 产品栈、垂直整合与商业化姿态
到 CES 2026,Robot Era 展示的不是一台机器人,而是一个家族:L7 人形机器人、Q5 平台和 XHAND1 灵巧手。公开技术细节中支撑最充分的是:L7 身高 171 centimeters,55 个自由度,重量大约 70 kilograms,双臂负载 20 kilograms;XHAND1 则有 12 个自由度、80 newtons 抓握力和 25 kilograms 提举能力。CES 报道中,Q5 是一个更轻的 44-DoF 伴随型号。这些规格的重要性不在于炫技指标,而在于说明 Robot Era 正试图把一个可复用平台商业化,覆盖完整人形机器人和操作器。2026 年融资稿进一步强化了这一判断:公司声称超过 95% 部件自研、2025 年 XHAND 出货超过 1,000 台、2026 年 Q2 启动千台级交付。结合 Chen 的「脑加身体」表述,已审阅证据支持一个商业化逻辑:模块化硬件、软件集成,以及对一家极年轻公司而言异常深的垂直控制。[CO007, CO021, CO022, CO023, CO024, CO025]
Robot Era 把顶尖学术机器人背景、模块化具身 AI 硬件、企业部署和激进融资连在一起,但板块成熟度仍是外部约束。
[CO005, CO006, CO007, CO021, CO023, CO026]1.5 里程碑、国际合作与行业反向背景
Robot Era 的 2025-2026 年里程碑显示,公司正试图把人形机器人热潮转成可重复的企业部署。公开来源把公司与大额订单、具名产业伙伴、China Post 和 SF Group 的物流中心建设、CES 2026 展示,以及与 UNIDO 在 Riyadh 签署、聚焦制造、物流和商业服务的合作协议联系起来。这些信号说明,公司目标是真实运营环境,而不是停留在 demo 循环。但公司外部的反向背景同样重要。CKGSB 认为,中国人形机器人行业在自主性和真正商业成熟度上仍落后美国领导者;TNW 提到买方满意度偏弱,以及电池续航和商业化问题持续存在;The Robot Report 则描述了一个拥挤赛道,2026 年将面临压力并可能洗牌。这些行业警告不能推翻 Robot Era 的牵引力主张,但能正确框定承销难题:Robot Era 也许扩张很快,却身处一个客户满意度、自主性和可持续部署经济性仍未稳定的市场。[CO020, CO030, CO031, CO032, CO033, CO034]
| 日期 | 事件 | 类型 | 金额 / 范围 | 参与方 | 重要性 |
|---|---|---|---|---|---|
| 2023-08-01 | Robot Era 成立 | 成立 | 公司于 2023 年 8 月成立 | Jianyu Chen | 标志着一个异常快速资本形成故事的起点 |
| 2025-07-01 | Series A 被报道 | 融资 | 近 RMB500 million | CDH VGC、Haier、Robot Era(相关方) | 所给资料包中首次披露的大额融资 |
| 2025-11-01 | Series A+ 公布 | 融资 | 近 RMB1 billion(约 $140 million) | Geely, BAIC, Robot Era | 验证投资人快速升温的兴趣和车企关注 |
| 2025-11-24 | UNIDO 合作协议在 Riyadh 签署 | 合作 | 制造、物流和商业服务范围 | Robot Era, UNIDO | 将叙事从国内机器人延伸到国际工业用例 |
| 2026-01-01 | CES 2026 产品线公开亮相 | 产品 | L7、Q5 和 XHAND1 展示 | Robot Era | 展示多产品商业化推进 |
| 2026-03-01 | 战略轮公开披露 | 融资 | RMB1 billion;估值 > RMB10 billion | Robot Era 和战略支持方 | 在公司背书材料中确立独角兽级估值 |
| 2026-04-27 | CnTechPost 报道新融资轮 | 融资 | $200 million 新融资轮 | SF Group 和其他投资人 | 暗示战略轮之后融资势头仍在延续 |
| 2026-05-08 | 官方公布 >$200 million 融资轮和物流部署说法 | 商业化 | 10+ 物流中心;2026 年第二季度千台级交付;>300% 增长 | SF Group、HSG、IDG Capital、China Post、Robot Era(相关方) | 将融资与企业部署和规模说法连接起来 |
当所给事实模式未注明具体日历日期时,月份级里程碑使用该月第一天。2026 年融资顺序按报道呈现,留待尽调调和。
[CO003, CO012, CO013, CO017, CO019, CO020]Robot Era 的公开时间线从 2023 年 8 月创立延伸到 2026 年融资、CES、物流和国际合作里程碑,同时板块怀疑声也在升温。
来源提供精确日期时使用精确日期;只有月份的里程碑使用该月第一天,以保留顺序,但不暗示已知具体日期。
[CO003, CO012, CO013, CO017, CO019, CO020]1.6 展示材料
02市场分析
2.1 市场边界、纳入支出与估算区间
不能用全球所有机器人、所有自动化或所有具身 AI 支出来给 Robot Era 定价。最可辩护的边界是:在结构化企业流程中商业部署的人形机器人,且人形形态能在已经按人类设计的环境里替代劳动力。BCC 的范围有用,因为它明确聚焦汽车、物流和仓储、政府和公共服务、医疗、零售或商业服务中的商业部署,同时排除试点、demo 和原型。IFR 和多份行业评论也指向仓储和制造业,认为它们是最早严肃验证的场地,而不是整个服务业宇宙。因此,纳入支出应是面向真实物流、制造和面客运营的硬件、具身 AI 软件、集成和部署服务。排除支出则是更大的传统工业机器人、AMR、仓储软件、消费电子和研究阶段系统;这些还不代表真实的人形机器人购买行为。公开估算仍不一致:BCC 给出 2025 年 $1.9 billion 市场,到 2030 年增长至 $11 billion;CNBC 对 Barclays 的摘要则称当前市场只有 $2 billion 到 $3 billion,但长期期权价值大得多。这个矛盾不是缺陷,而是正确警告:今天已经货币化的市场,相对于附着其上的未来叙事仍然很小。[CM001, CM002, CM003, CM004, CM005, CM027]
| 细分 / 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 与 Robot Era 的相关性 |
|---|---|---|---|---|
| 面向物流和仓储的商用人形机器人 | 用于拣选、分拣、导入和搬运的人形硬件、具身 AI 软件、末端执行器、集成和部署支持 | 通用仓库软件、没有人形操作能力的 AMR、机器人替代范围之外的临时人工支出 | 仓库运营方、3PL、快递公司或物流中心自动化所有者 | 近期相关性最高,因为这是 Robot Era 最清晰的公开部署切口 |
| 面向制造业的人形机器人 | 用于零部件拣选、装配、检测和其他结构化工业任务的工厂人形机器人 | 广义工业机器人市场、机床,以及不需要人形形态因素的非人形固定自动化 | 工厂运营、工程和质量负责人 | 相关性高,因为中国工厂密度和政策支持提高部署概率 |
| 商业服务人形机器人 | 零售、前台、导览和类似面向客户的具身 AI 工作流 | 消费电子小工具、娱乐机器人和通用软件助手 | 门店经营者、商场管理方、医院或服务场所运营方 | 相关,但不如物流或制造得到验证 |
| 中国具身 AI 场景市场 | 政策支持试点、标准工作、工业部署支持和企业商业化项目 | 中国全部 AI 市场和全部机器人品类 | 市政府、产业园区和企业采用方 | 最好的公开 SAM 地理范围,因为政策和应用集中在那里 |
| 全球商用人形机器人市场 | 跨企业垂直行业商业部署的人形平台收入 | 研究原型、演示和愿景型家用机器人叙事 | 全球企业买家和集成商 | 有用的 TAM 锚点,但过宽,不能作为 Robot Era 近期可服务市场 |
| 现状替代品 | 固定工业机器人、AMR、协作机器人、工作流重设计和人力 | N/A | 与上述相同的买家 | 这些替代品争夺同一预算;如果 ROI 更弱,会拖慢人形机器人转化 |
关键边界选择是把 Robot Era 放在结构化企业任务中的商用部署人形机器人里衡量,而不是拿全部自动化或全部具身 AI 话术来衡量。
[CM001, CM002, CM015, CM020, CM027, CM032]对全球商业人形机器人当前市场给出低位、中位和高位视角,保留已发布口径与投资人式口径之间的巨大差距。
前两个区间描述当下市场规模,第三个保留长期上行叙事;放在一起是为了凸显,已变现市场相对乐观预测仍然很小。
[CM003, CM004, CM028]2.2 中国政策与自动化底座
对本章而言,中国比泛泛的全球平均值更重要,因为政策和已安装自动化产能都异常集中在中国。官方和国家关联来源显示,2026 年中国推出首个人形机器人与具身智能国家标准框架;上海则另行设定 2027 年计划,目标是 100 家领军企业、100 个应用场景、100 个产品,以及超过 RMB50 billion 的产出,并用补贴支持商业化。China Daily 还把具身 AI 直接连接到 2026 年政府工作报告,并引用 Development Research Center 预测:中国国内具身 AI 市场到 2030 年达到 RMB400 billion,到 2035 年超过 RMB1 trillion。已安装自动化基础同样重要。IFR 称,中国在 2024 年占全球工业机器人部署量的 54%,运营存量超过 2 million 台,并且到 2028 年仍有约 10% 年增长空间。换句话说,Robot Era 正在世界最深的工厂自动化市场里扩张,而政策制定者也正在试图标准化并加速具身 AI 部署。这是强劲的 SAM 顺风,但不能消除技术碎片化:多个中国来源仍强调硬件质量不均、数据孤岛、灵巧手瓶颈,以及规模化常态化之前需要通用评估标准。[CM006, CM007, CM008, CM009, CM010, CM011]
2.3 目标垂直:物流、制造与商业服务
最强的买方证据不是泛化家庭人形机器人,而是已经大笔投入自动化、并感受到劳动力或吞吐压力的企业运营方。物流是最清晰切口,因为 IFR 已把运输和物流记录为最大的专业服务机器人应用,仓储自动化分析师预计 2026 和 2027 年采纳会扩大,Robot Era 的公开部署也集中在物流中心。制造业是第二大切口,因为 China Daily、IFR 和公司关联报道都指向结构化工厂任务,如零部件拣选、装配、分拣和检测;在这些任务里,可靠性比社交互动更重要。商业服务是第三个、但标准化程度更低的切口:零售、前厅、导览等任务适合人形形态,但证据比仓库或工厂更薄、更像试点。买方图谱因此分成三类。物流中,运营或自动化负责人掌握预算,仓库团队是用户。制造中,工厂、工程和质量负责人既是赞助者也是采纳者。商业服务中,付款方可能是零售运营商、商场、医院或家电合作伙伴,但采纳仍取决于机器人能否胜过更便宜的固定功能替代方案。因此,Robot Era 的市场重点不是消费者新奇感,而是哪些企业流程能足够快地奖励人类兼容形态,从而证明部署合理。[CM012, CM013, CM014, CM015, CM016, CM020]
| 细分 | 买方 | 用户 | 付款方 | 工作流 | 预算所有者 | 采用触发因素 |
|---|---|---|---|---|---|---|
| 物流中心 / 包裹作业 | 3PL、快递、仓库运营商,或大型发货方 | 现场运营、自动化工程师和一线主管 | 运营预算或已批准的资本开支 / RaaS 支出 | 包裹导入、分拣、拣选、周转箱搬运、移动操作 | COO、物流自动化负责人,或仓库总经理 | 24/7 吞吐需求、劳动力短缺,以及高差异物品处理 |
| 制造工厂 | 汽车、电子、电器,或精密制造运营商 | 产线操作员、工业工程师和质量团队 | 工厂资本开支和自动化预算 | 零部件拣选、检测、装配支持、物料搬运 | 工厂经理、制造副总裁,或质量负责人 | 需要在人类设计的产线里部署灵活自动化 |
| 商业服务场所 | 零售连锁、商场、医院,或电器 / 服务合作方 | 门店员工、前台团队,或服务主管 | 门店运营或数字化转型预算 | 导览、零售协助、配送,或简单交互任务 | 总经理、创新负责人,或服务运营负责人 | 当人形形态重要时,提升客服效率并替代部分劳动力 |
| 政策支持试点 / 产业园 | 市政府、开发区,或国资关联平台 | 园区运营方、合作企业、标准团队 | 公共补贴加企业共同投入 | 试点验证、标准制定和场景商业化 | 产业园发起方或市级未来产业办公室 | 推动具身 AI 加速落地的战略诉求 |
Robot Era 的买方地图由企业牵引:物流和制造是近期最强切入口,商业服务和政策支持试点则拓展边缘场景。
[CM012, CM015, CM016, CM020, CM021, CM022]矩阵展示 Robot Era 目标垂直领域中,买方归属和近期准备度在哪里最清晰。
[CM008, CM015, CM016, CM020, CM022, CM023]2.4 受证据约束的 Robot Era TAM、SAM 与 SOM
只有把每一层都定义得足够窄,才能搭出干净的 TAM-SAM-SOM。TAM 应是商业人形机器人市场,而不是所有机器人;这里支撑最充分的公开口径是 BCC 的 2025 年 $1.9 billion,到 2030 年增长至 $11 billion。第二个更偏宣传的口径来自 CNBC 转述的 Barclays,称当前市场为 $2 billion 到 $3 billion,到 2035 年达到 $200 billion。它适合做上行情景,不适合做基准情景。SAM 还应更窄:以中国为中心、落在物流、制造和商业服务中的具身 AI 部署;这些场景已经有政策支持、工厂密度、物流自动化需求和服务机器人采购。公开证据在定性上支持这个中国中心的 SAM,但仍不能转化为 Robot Era 单一、干净的收入池。SOM 是最受约束的一层。公开来源支持 2025 年商业订单超过 RMB500 million、以 SF Group 和 China Post 为代表的物流切口、10 多个物流中心,以及 2026 年 Q2 启动千台级交付。这些都是强商业化标记,但仍未披露已确认收入、复购率、按机器人类型定价、续约经济性或毛利率。因此,正确承销结论是:Robot Era 在中国企业部署中已经有可见的早期 SOM 落脚点,但公开数据不足以精确计算市场份额。[CM003, CM009, CM020, CM021, CM024, CM025]
| 发布方 / 视角 | 年份 | 地理范围 | 数值 | CAGR / 增长信号 | 方法论 | 置信度 | 关键限制 |
|---|---|---|---|---|---|---|---|
| BCC 商用人形机器人市场 | 2025-2030 展望 | 全球 | 2025 年 $1.9B;2030 年 $11B | 42.8% CAGR(2025-2030) | 按垂直行业和地区统计的商用部署人形机器人;排除试点和原型 | 中 | 有用的狭义 TAM,但仍是第三方市场模型 |
| Barclays / CNBC 当前市场视角 | 2026 评论 | 全球 | 当前 $2B-$3B;2035 年 $200B | 长期 100x 式扩张叙事 | 投资人展望框架,对比当前基数和长期上行 | 中 | 适合做区间框定,不适合作为近期承销基准 |
| 中国具身 AI 政策视角 | 2030 / 2035 预测 | 中国 | 2030 年 RMB400B;2035 年 >RMB1T | 政策主导的未来产业增长信号 | China Daily 引用的国家相关预测 | 中 | 具身 AI 市场比 Robot Era 可变现的人形机器人收入更宽 |
| IFR 工业机器人底座 | 2024 实际 / 2028 展望 | 中国 / 全球 | 中国安装 295k;占全球部署 54%;存量 >2M | 中国到 2028 年约 10% 平均年增长潜力 | 官方机器人安装统计 | 高 | 自动化底座,不是仅限人形机器人收入 |
| IFR + Interact 物流自动化视角 | 2024 实际 / 2027 预测 | 全球 | 2024 年售出 102.9k 台物流服务机器人;2027 年 26% 仓库实现自动化 | 仓库自动化向 2026 年继续扩展 | 官方服务机器人数据加分析师仓库渗透率预测 | 高 | 并非所有自动化仓库或物流机器人都会使用人形机器人 |
| Robot Era 证据约束的 SOM 下限 | 2025-2026 公开记录 | 中国物流 / 工业部署 | 2025 年订单 >RMB500M;>10 个物流中心;2026 年第二季度开始千台级交付 | 2026 年 300% 增长说法 | 公司相关和独立商业化标记 | 中 | 未披露确认收入、利润率或复购率 |
本表有意混合市场规模、政策和部署视角,因为没有单一公开数据集能干净衡量 Robot Era 的可寻址人形机器人市场。
[CM003, CM004, CM009, CM010, CM013, CM021]证据受限的金字塔:从广义自动化底座,逐层收窄到 Robot Era 更窄的可服务和可获取楔形市场。
这个金字塔有意叠放相邻但不可相加的视角:广义自动化价值、狭义人形机器人 TAM、中国中心的 SAM 语境,以及 Robot Era 有证据支撑的商业化下限。
[CM003, CM009, CM011, CM021, CM024, CM027]2.5 增长驱动、约束与承销含义
采纳逻辑真实存在,但仍有条件。正面看,劳动力短缺、24/7 吞吐需求、仓储自动化投资和快速下降的硬件成本,都支持人形机器人在结构化企业环境中采纳。IFR 的工业和服务机器人数据说明,买方已经愿意为工厂和物流自动化付费。Interact 和 McKinsey 显示这些预算仍在扩张;BusinessWire 和 CNBC 的成本压缩信号则说明,中国可能把人形机器人价格压低到许多西方投资者预期之外。可是,反向证据同样重要。Humanoid.guide、RobotToday、China Daily 和更广泛行业报道都反复指向同一批阻碍:灵巧手、训练数据、安全认证、功耗和散热限制,以及相对更便宜固定功能系统证明 ROI 的难题。这意味着投资争论不是市场是否存在,而是结构化试点能以多快速度变成可重复、毛利为正的推广。对 Robot Era 来说,最强的市场解读是:未来几年物流和工厂部署能支撑真实增长;主流服务采纳大概率滞后,家庭市场叙事仍然过早。公开数字足以支持正面市场逻辑,但不足以跳过对经济性、留存和客户集中度的尽调。[CM013, CM014, CM015, CM017, CM018, CM019]
| 驱动 / 约束 | 方向 | 时间 | 含义 | 尽调问题 |
|---|---|---|---|---|
| 劳动力缺口与 24/7 吞吐需求 | 驱动 | 当前 / 近期 | 仓储和工厂已经有预算理由,去测试能在人类空间工作的仿人机器人 | 要求提供客户 ROI 案例,展示人力节省、正常运行时间和错误率 |
| 中国政策、标准与补贴 | 驱动 | 当前 / 2027 | 国家和上海政策支持可能缩短在中国从试点走向部署的路径 | 区分象征性政策支持与已签采购承诺 |
| 仓库自动化扩张 | 驱动 | 当前 / 2027 | 仿人机器人进入的是自动化渗透率仍在上升、而非已经饱和的物流市场 | 验证仿人机器人能否从 AMR、固定系统和协作机器人手里赢得份额 |
| 成本快速压缩与生产放量 | 驱动 | 当前 / 2028 | 只要可靠性守得住,价格下行就能打开更广泛采用 | 检查 BOM 走势、服务负担,以及降价后能否保住利润率 |
| 灵巧手、数据与安全瓶颈 | 约束 | 当前 | 技术不成熟可能拖慢从结构化任务和小型试点走向规模化的速度 | 要求提供任务成功率、介入率、安全事故和认证路线图 |
| 公开 SOM 可见度弱 | 约束 | 当前 | 订单和交付计划看起来强劲,但公开来源没有披露已确认收入、留存或单位经济性 | 要求按垂直行业提供客户 cohort 数据、已实现收入、续约行为和毛利率 |
承销问题在于时间:真实需求存在,但商业化速度取决于成本下行能否跑赢长期存在的可靠性、安全和 ROI 瓶颈。
[CM013, CM014, CM017, CM018, CM019, CM029]采用漏斗展示广泛市场兴趣如何收窄为可重复的 Robot Era 部署。
指数值是根据已审阅证据推导的示意性阶段权重,不是公司报告的转化率。
[CM015, CM024, CM031, CM032, CM033, CM034]03竞争格局
3.1 竞争版图与当下最重要的对手
不应只拿 Robot Era 对标一两个人形机器人知名品牌。真实买方选择集覆盖几类玩家:中国直接人形机器人 OEM 正在价格、规模和工业交付上竞速;全球工业人形机器人在汽车或物流流程中有更强公开证明;软件栈玩家不卖整机,却抬高自主性门槛;更低成本的替代方案则用移动操作解决部分任务,而不是完整人形身体。按 Robot Era 自己的记录,最强公开优势是物流部署、高自研部件占比,以及把 L7 人形机器人与 XHAND1 灵巧操作结合起来的产品家族。最重的压力点因对手而异。Unitree 重置了入门价格和出货量预期。UBTECH 有上市公司披露和汽车工厂证据。AgiBot 与 Galbot 比 Robot Era 更用力推动中国规模叙事。Figure、Boston Dynamics 和 Agility 展示了同一挑战的西方版本:公开企业部署、明确集成和更会讲的软件栈。因此,正确承销框架不是一场赢家通吃竞赛,而是一组加权威胁,从不同角度攻击 Robot Era。[CP004, CP009, CP014, CP021, CP025, CP027]
| 竞争对手 | 类别 | 规模 / 融资信号 | 目标细分市场 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Robot Era | 中国直接仿人机器人现有玩家 | > $200M 最新一轮融资;>10 个物流中心;称 2026 年 Q2 交付千台级 | 物流、制造、商业服务、科研 | 声称 95%+ 核心部件自研;L7 加 XHAND1 栈;物流 PMF | 没有公开标价;公开汽车工厂证明少于部分对手 |
| AgiBot | 中国直接仿人机器人及平台同业 | 声称第 10,000 台机器人下线;Expedition 出货 2,000+ 台;2025 年收入 RMB1B | 工业、物流、服务、零售、安防、教育 | 产品组合更广,AIMA 生态加上大规模部署叙事 | 公开标价仍不透明 |
| Unitree | 中国直接仿人机器人价格领导者 | IPO 申报指向 RMB1.71B 的 2025 年收入,并声称出货领先 | 科研、教育、早期商业和工业用途 | 公开 $13.5K 的 G1 价格,且积极压缩 ASP | 基础 G1 负载相对重型工业仿人机器人偏小 |
| Fourier | 中国灵巧度和开发者导向仿人机器人同业 | 官方将 GR-1 定位为量产;GR-2 增加更丰富的手部工具 | 科研、医疗、工业试点、开发者生态 | 12-DoF 触觉手,并明确支持 SDK | 公开的工厂级客户数量较少 |
| UBTECH | 中国上市工业仿人机器人同业 | 2025 年仿人机器人收入 RMB820.6M;数百台 Walker S2 批量订单 | 汽车、智能工厂、物流、工业运营 | 公开申报透明度高,工厂部署姿态强 | 价格仍不透明,定位偏工业优先而非广泛消费 / 开发者 |
| Galbot | 中国移动操作机器人与具身 AI 同业 | 累计融资 $800M;估值 $3B;声称千台级订单 | 制造、零售、移动操作、具身 AI 部署 | 融资规模很大,工业大客户标识亮眼 | 官方产品细节弱于融资或部署头条 |
| Figure | 全球仿人机器人与自主栈同业 | > $1B Series C,估值 $39B;BMW 部署协议 | 制造、物流、家庭 | Helix 软件品牌加公开负载 / 续航披露 | 商业收入质量仍未公开详述 |
| Physical Intelligence | 自主栈相邻竞争者 | 知名投资方阵容;VLA 研究节奏快 | 覆盖多种形态的机器人 OEM 与自主合作伙伴 | 卖点是泛化和软件深度,而不是销售机器人本体 | 没有披露直接硬件或部署经济性 |
| 1X | 家庭优先的仿人机器人相邻同业 | NEO 早期访问需支付 $200 定金;有限家庭测试 | 家务协助和消费 / 家庭机器人 | 家庭定位清晰,设计强调安全 | 距离规模化商业化仍很远 |
| Boston Dynamics | 全球工业现有玩家基准 | 提到 Hyundai 商业化路径和企业 Atlas 规格表 | 汽车、物料搬运、企业自动化 | 本组中披露的负载和系统集成最强 | 这里没有清晰的公开价格或大批量交付数字 |
| Agility Robotics | 全球物流和汽车部署同业 | 付费 GXO 与 Toyota RaaS 交易,加上 Amazon 或 Schaeffler 参考 | 仓储、物流、制造 | 有真实付费部署证据和工作流集成栈 | 公开数量仍小,标价未披露 |
| AgileX | 相邻移动操作替代方案 | 官网聚焦底盘、机械臂、ALOHA 套件和数据采集 | 科研、教育、移动机器人、部分自动化 | 对部分具身智能任务,是更便宜或更简单的替代方案 | 不是直接的全尺寸仿人机器人基准 |
这些行有意偏向真实 Robot Era 买方决策中最可能重要的玩家,而不是市场评论里提到的每一个仿人机器人品牌。
[CP004, CP008, CP014, CP017, CP021, CP025]按与 Robot Era 当前论点的重合度和公开商业化证据,对最相关直接同业及相邻公司做序数定位。
坐标轴是基于已审阅证据质量和竞争重合度打出的 1–10 序数分,不是已发布市占率或收入排名。
[CP009, CP014, CP021, CP025, CP027, CP032]3.2 中国直接对手在价格、规模和披露上给 Robot Era 施压
最近的竞争压力仍来自中国。AgiBot 在产品广度和生态野心上看起来最宽:公开资料已覆盖人形机器人、单臂移动操作、灵巧手、四足机器人、数据采集系统,以及不断扩展的 AIMA 开发栈。Unitree 是资料包中最具冲击力的公开价格信号。$13.5K 的 G1 不能与 Robot Era 的 L7 做一一替代,但它建立了一个市场预期,而许多具身 AI 供应商仍用不透明定价回避这个预期。Unitree 的 IPO 披露也让其收入增长、出货结构和主动 ASP 压缩异常可见。Fourier 在出货规模上没那么重要,重要的是灵巧性和开发者友好度:GR-2 的 12-DoF 双手、触觉传感器和明确 SDK 支持,形成了比许多同行更清晰的研究与操作基准。UBTECH 又不一样。其上市公司文件和 Walker S2 大规模交付表述,提供了中国最强公开证据:人形机器人收入已经成为一条与工厂使用绑定的真实收入项。Galbot 补上最后一个中国威胁:即便产品细节更薄,巨额融资、移动操作器定位和汽车 logo 也能迅速塑造买方与投资人认知。面对这个赛场,Robot Era 在结构化物流和垂直硬件控制上仍显得最强,但它不是这个类别唯一快速前进的中国竞争者。[CP001, CP002, CP003, CP005, CP006, CP007]
| 公司 | 公开价格信号 | 操作 / 负载信号 | 工业或物流证明 | 软件 / 自主信号 | 最适配细分市场 | 关键未知 |
|---|---|---|---|---|---|---|
| Robot Era | 未找到公开标价 | L7 加 XHAND1;直驱手;聚焦物流操作 | >10 个物流中心,并声称千台级交付 | 提到 ERA-42,但公开文档少于 Figure 或 PI | 中国物流和工业工作流 | 已实现 ASP、续约率和服务经济性 |
| AgiBot | 已获取来源中未知 / 未公开 | 仿人机器人、灵巧手和移动操作机器人产品线较广 | Expedition 出货 2,000+ 台,2026 年 Q2 产能售罄 | AIMA 栈和 AGIBOT World 数据集叙事 | 中国多场景具身 AI 部署 | 实际企业定价和客户集中度 |
| Unitree | 强:G1 标价 $13.5K | 23-43 DoF G1;基础臂负载低于更重的工业同业 | 招股书指向出货领先和较高仿人机器人收入占比 | 入选 Nvidia 平台增强生态信号 | 价格敏感的科研、教育和早期商业规模化 | 高端工业用途与更重型同业相比表现如何 |
| Fourier | 已获取来源中未知 / 未公开 | 强:GR-2 12-DoF 触觉手和 380 N.m 峰值扭矩 | 官方页面强调实际应用,但没有大型具名车队 | 明确支持 SDK 和开发者工具 | 重灵巧度的科研、医疗和工业试点 | 具名客户规模和已实现定价 |
| UBTECH | 公开来源显示未知 / 基于合同 | 41 关节工业仿人机器人,带工厂集成 | 强:数百台 Walker S2 批量交付和重要仿人机器人收入线 | 基于 LLM 的决策,以及 BrainNet / Co-Agent 表述 | 汽车、智能工厂、物流 | 利润率结构和实际利用率 |
| Figure | 未知 / 企业合同 | 强:公开页面显示 20 kg 负载和 5 小时续航 | BMW 制造证明和物流定位 | 本组中公开 Helix 品牌最强 | 制造、物流,最终进入家庭 | $39B 叙事中有多少已经由经常性部署支撑 |
| Physical Intelligence | 不适用 | 模型层面,而非机器人本体规格 | 此处没有直接部署清单;由合作伙伴牵引 | 很强:任何机器人、任何任务的野心,加上开放世界泛化工作 | 面向多种机器人形态的自主栈 | 收入模式和合作伙伴集中度 |
| Agility Robotics | 未知 / RaaS | 中等负载;面向重复性工业劳动设计 | 强:GXO、Toyota、Amazon 和 Schaeffler 参考 | 通过 Arc 和服务支持实现较强工作流集成 | 仓储和制造 | 规模经济和长期硬件 ASP |
| Boston Dynamics | 未知 / 企业合同 | 很强:50 kg 瞬时负载、30 kg 持续负载、56 DoF | Hyundai 商业化路径和 Georgia 工厂演示 | 通过 Orbit 和企业车队工具实现强集成 | 重型企业物料搬运和汽车 | 产量爬坡和标价 |
该矩阵强调已获取来源真正支持的内容。未知不代表弱;它表示留存的公开证据还不足以清晰评估该属性。
[CP004, CP009, CP014, CP017, CP021, CP028]3.3 全球对手与软件优先玩家抬高自主性和企业门槛
全球赛场重要,因为它凸显 Robot Era 公开记录仍较薄的地方。Figure 把资料包中最清晰的软件品牌,与有意义的硬件披露结合起来:Figure 03 已公开 20 kg 负载数字,Helix 被定位为面向物流和家庭任务的机载 VLA,BMW 是具名制造业合作方,公司融资规模也已按数百亿估值衡量,而不是普通创投轮次。Physical Intelligence 不是直接硬件同行,但战略上重要,因为其公开文章解释了为什么大多数机器人仍首先在高度受控环境中成功,以及为什么开放世界泛化是另一个竞争前沿。这正是买方越来越会拿来比较 Robot Era 和更高调自主性技术栈的缺口。1X 作为近期工厂对手较弱,因为它仍优先家庭场景,距离规模化商业化很远,但它会争夺家庭人形机器人的心智。Boston Dynamics 在已披露负载、续航和企业集成上仍是最具技术威慑力的基准;Agility 则可能拥有最清楚的运营故事,证明企业买方确实会通过可重复的 RaaS 式物流和汽车部署为人形机器人付费。这些同行不能否定 Robot Era 的中国切口,但确实提高了全球可持续护城河的标准。[CP026, CP027, CP028, CP029, CP030, CP031]
| 公司 / 型号 | 公开价格或合同信号 | 融资 / 财务披露 | 商业部署信号 | 含义 |
|---|---|---|---|---|
| Robot Era | 已获取来源中没有公开标价 | > $200M 最新一轮融资;私有资本结构 | >10 个物流中心;2026 年 Q2 启动千台级交付;2025 年已发货 200 台 | 商业化叙事强,但投资者仍需要已实现 ASP 和复购数据 |
| AgiBot | 已获取来源中没有公开标价 | 报道 2025 年收入 10 亿元;声称第 10,000 台机器人下线 | Expedition 出货 2,000+ 台;2026 年 Q2 产能售罄 | 即使定价不透明,规模和速度也可能压缩品类预期 |
| Unitree G1 | G1 标价 US$13.5K;EDU 版本面向开发加价 | IPO 路径显示 2025 年收入 17.1 亿元,并披露 ASP 压缩 | 声称出货领先,并有 Nvidia 生态信号 | 公开价格锚是中国最清晰的商品化威胁 |
| UBTECH Walker S2 | 无公开标价;工业解决方案销售 | 上市公司申报显示仿人机器人收入 RMB820.6M | 数百台交付;2026 年年产能目标 5,000 台 | 披露质量和工业可信度超过大多数非上市同业 |
| Galbot | 无公开标价;订单牵引的企业销售 | 累计融资 $800M;估值 $3B;2026 年国资支持轮 | 声称千台级订单,并讲移动操作机器人放量故事 | 在买方感知里,资本实力可能比价格透明度更重要 |
| Figure 03 | 无公开标价;企业或试点合同 | > $1B Series C,估值 $39B | BMW 工厂部署和 Helix 物流卖点 | 即便公开经济性尚不清晰,融资规模也支撑软件和车队建设 |
| 1X NEO | $200 定金和早期访问家庭机器人漏斗 | 留存来源中没有重大财务披露 | 有限家庭测试;明确距离规模化很远 | 更像在未来消费叙事上竞争,而不是当前企业采购 |
| Agility Digit | RaaS / 合同模式,而非公开标价 | 私营公司经济性仍未披露 | 正式付费 GXO 和 Toyota 部署,加上 Amazon 或 Schaeffler 参考 | 相比资本开支较重的买方,服务模式可能降低采用摩擦 |
| Boston Dynamics Atlas | 无公开标价;企业产品销售 | 此处没有公开的 Atlas 专属收入线 | Hyundai 商业化路径和 Georgia 工厂证明 | Robot Era 在企业集成上面临的基准比标价更难 |
最有信息量的比较不只是标价,而是竞争对手是否披露标价、融资或申报是否验证规模,以及公司是否已经从试点叙事跨入付费或重复部署。
[CP002, CP003, CP008, CP011, CP012, CP019]紧凑展示哪些竞争者公开披露了与 Robot Era 相关的价格、操作、部署和软件优势。
单元格概括保留来源能够清楚支撑的内容;低或未知代表公开证据薄弱,不一定代表底层能力弱。
[CP004, CP016, CP021, CP028, CP034, CP039]3.4 定价不透明、工作流集成与护城河耐久性
最重要的承销问题不是 Robot Era 是否有真实竞争者,而是买方决策中哪些部分可能保持粘性。公开定价仍是最大盲区之一。在这组证据里,Unitree 是唯一有清晰公开标价锚点的主要同行;Robot Era、UBTECH、Figure、Boston Dynamics 和 Agility 大多围绕合同、服务或企业部署来描述销售。因此,切换成本不只看标价,更要看客户买的是机器人本体、灵巧手、工作流软件、安全验证、服务支持,还是完整运营方案。Agility 的 Arc 技术栈、Boston Dynamics 的 Orbit 和工作流集成、UBTECH 的工厂系统连接,都说明企业竞争者如何试图越过硬件,进入流程锁定。Physical Intelligence 和 Figure 则从自主性一侧攻击同一问题,比 Robot Era 在公开叙事中更明确地宣传泛化和推理。Robot Era 在垂直整合和物流中心部署上仍有真实护城河主张,但弱点清楚:价格不透明,汽车案例研究比 UBTECH、Agility 或 Boston Dynamics 薄,公开软件叙事也不如 Figure 的 Helix 或 Physical Intelligence 的 VLA 发表成熟。因此,投资人应把 Robot Era 视为有竞争力且有前景,但还不能认为它已经被隔离保护。[CP004, CP014, CP017, CP021, CP027, CP032]
| 护城河主张 | 威胁 | 严重性 | 有证据支撑的理由 | 缓释 / 尽调问题 |
|---|---|---|---|---|
| 垂直整合与自研部件 | Unitree 价格压缩和 AgiBot 规模 | 高 | Robot Era 声称 >95% 自研确有价值,但 Unitree 和 AgiBot 仍可能靠销量和更低入门价压缩市场预期。 | 要求 Robot Era 提供 BOM、毛利率桥接,并证明垂直整合带来成本或可靠性优势,而不只是增加复杂度 |
| 物流 PMF 和部署牵引 | Agility 付费物流部署和 UBTECH 工厂扩张 | 高 | Robot Era 拥有强物流中心证明,但 Agility 已宣传付费 GXO 部署,UBTECH 也在扩大工业交付。 | 要求 Robot Era 提供与 Agility、UBTECH 及仓库工作流中固定功能替代方案对比的赢单 / 输单数据 |
| 灵巧手差异化 | Fourier、Figure 和 Physical Intelligence 的软件或手部栈进展 | 中 | Robot Era 的 XHAND1 可信,但 Fourier 公开披露更丰富的手部规格,Figure 或 PI 的自主或操作叙事更强。 | 验证抓取成功率、循环寿命,以及相对具名同业的基准化操作表现 |
| 中国融资势头 | Galbot 和 AgiBot 的融资或规模叙事 | 中 | 即便底层经济性不同,Galbot 的融资头条和 AgiBot 的 10,000 台机器人主张也可能盖过 Robot Era。 | 区分品牌势能与已实现部署、客户留存和交付质量 |
| 软件与自主栈 | Figure Helix 和 PI 泛化研究 | 高 | Robot Era 提到 ERA-42,但公开技术深度弱于 Figure 的 Helix 和 PI 的 VLA 论文。 | 要求提供技术基准、泛化测试,以及证明自主能力超出结构化演示的客户证据 |
| 人形形态战略 | AgileX 式局部自动化,或更便宜的移动操作替代方案 | 中 | 不是每个买家都需要完整人形机身;一些任务用移动底盘、机械臂或数据套件工作流就可能解决。 | 梳理 Robot Era 哪些场景必须靠人形形态取胜,哪些场景低成本替代方案也能满足预算 |
严重程度是基于截至 2026-06-10 保留来源作出的定性承销判断,不是市场份额预测。
[CP004, CP014, CP021, CP027, CP032, CP035]五个精简指标概括 Robot Era 看起来最强的地方,以及公开记录中对手目前更强的地方。
[CP004, CP014, CP021, CP032, CP049, CP050]3.5 展示材料
04财务情况
4.1 收入模型、积压订单信号,以及为什么订单标题不等于收入
公开记录支持 Robot Era 有真实商业引擎,但不支持一个已确认收入模型。公司关联与独立来源持续指向:订单超过 RMB 500 million,单个物流订单超过 RMB 50 million,在 10 多个物流中心部署,2026 年 Q2 启动千台级交付,2025 年出货 200 台人形机器人,100 多个订单推进中,2025 年出货超过 1,000 台 XHAND。这些是有意义的商业化标记,尤其是复购订单和大型物流合同出现在同一组信源中。但它们仍是预订、出货或部署信号,不是收入确认披露。即便最接近结构数据的信息,也仍是公司口径:Yicai 报道海外业务占订单数字的一半,KrASIA 则称海外收入约占总收入一半,但没有披露分母。这在方向上正面,却不足以搭出干净的收入桥。投资人因此应把 Robot Era 视为一家有可信积压订单和出货牵引力的公司;在管理层披露分业务已入账收入、积压订单转化和客户集中度之前,收入质量仍未验证。[CI002, CI003, CI004, CI005, CI007, CI008]
| 收入流 | 机制 | 单位 | 当前价值 / 状态 | 质量 | 尽调要求 |
|---|---|---|---|---|---|
| 企业人形系统销售 | 机器人本体,加上在物流、制造或服务工作流中的部署 | 订单价值 / 交付台数 | 已披露订单 >RMB 500M;2025 年出货 200 台机器人;称 2026 年 Q2 已开始千台级交付 | 需求为中,收入确认为低 | 按机器人 SKU、地区和季度提供已确认收入。 |
| 大型物流解决方案合同 | 针对海关查验、分拣或拣选部署等具体场景的合同 | 合同金额 | 已披露最大单笔订单 >RMB 50M;称与中国邮政、顺丰集团覆盖 >10 个物流中心 | 积压订单信号为中,利润率信号为低 | 披露合同范围、验收标准、计费里程碑和积压订单账龄。 |
| 独立灵巧手销售 | XHAND 作为单独模块销售,而不只嵌入完整人形机器人 | 出货台数 | 根据公司 PR,2025 年 XHAND 出货 >1,000 台 | 中 | 提供 XHAND 的实际 ASP、毛利率和复购客户结构。 |
| 国际商业活动 | 机器人或模块出口及海外部署 | 订单结构 / 收入结构 | 第一财经称海外业务占订单 50%;KrASIA 称海外收入约占总收入一半,但没有披露分母 | 低到中 | 按地区对海外订单、收入和出货台数做勾稽。 |
| 支持、软件和集成 | 叠加在部署之上的工作流集成、支持或软件附加 | 未公开披露 | 部署表述有所暗示,但没有公开的独立收入线或定价 | 低 | 将经常性支持和软件收入从硬件中单独拆出。 |
各行把需求信号与已确认收入质量分开。除非来源明确说是收入,否则订单、出货和部署都不按收入处理。
[CI002, CI003, CI005, CI007, CI008, CI009]展示 Robot Era 可能如何从工业需求走向订单、交付、验收,最终确认收入;关键会计步骤仍未公开披露。
这是一座结构性桥梁,不是已发布的会计政策。它基于已披露订单和部署信号,以及收入确认披露明确缺失这一事实。
[CI002, CI004, CI007, CI008, CI009, CI015]4.2 定价信号、成本代理与公开单位经济性的边界
Robot Era 最大的财务盲区仍是定价。保留的 Robot Era 来源没有一个公布 L7、M7 或 XHAND 在商业部署中的标价,因此本章只能依赖明确的同行锚点和市场定价阈值。Unitree 给出最干净的直接锚点:其官方 G1 页面列出 USD 13,500 基础价,第三方采购报道则把更可编程的 EDU 配置放在 USD 43,900 到 USD 73,900 之间。CnTechPost 对 Unitree IPO 流程的报道又补充了一个产业代理:2025 年人形机器人 ASP 为 167,600 yuan。另外,SCIO 和 OFWeek 更直接地框定企业 ROI 问题:许多人形机器人售价仍在数十万元人民币,企业可能把接近 300,000 yuan 的价格视为合理 ROI 阈值。这些信号重要,因为它们同时显示价格压缩,以及 demo 硬件与量产级经济性之间的差距。但它们仍不能解决 Robot Era 自己的 ASP 或成本栈。任何把公司 >RMB 500 million 积压订单换算成台数的尝试,都只是一个粗略、由同行锚定的情景,而不是收入模型。正确结论是:公开需求存在,但 Robot Era 已实现单位经济性仍不透明。[CI017, CI018, CI019, CI020, CI021, CI022]
| 产品 / 信号 | 公开价格或单位 | 标价与实际 | 解读 | 来源 |
|---|---|---|---|---|
| Robot Era L7 / M7 / XHAND 商业部署 | 保留来源中没有公开标价 | Robot Era 仍然价格不透明,因此积压订单无法换算成实际 ASP 或利润率。 | Robot Era 保留来源 | |
| Unitree G1 基础版 | $13,500,另加运费 | 官方标价式价格 | 这是入门级人形硬件最直接的参照,但不包含二次开发,也不是 Robot Era 价格。 | Unitree 官方产品页 |
| Unitree G1 EDU 区间 | $43,900 至 $73,900 | 第三方采购口径的市场价格汇总 | SDK 访问、更好算力和灵巧手纳入后,价格会明显抬升。 | BotInfo.ai |
| Unitree 人形机器人 ASP | 2025 年前 9 个月 167,600 元 | 招股书报道中的已实现 ASP | 可作为中国工业价格压缩的有用代理,但不是 Robot Era 的直接数据。 | CnTechPost |
| 企业 ROI 门槛 | 约 300,000 元 | 政策 / 行业评论门槛 | 暗示买家可能在这一价格点判断人形机器人 ROI 是否可接受。 | SCIO |
| 美国机构分销商路径 | 询价式 / 分销商采购 | 不是清晰标价 | 说明机构采购路径会增加摩擦,且经常让买家偏离制造商裸机标价。 | RoboStore |
只有一行直接覆盖 Robot Era,且因没有保留公开标价而有意留空。其他各行都是外部锚点或采购语境。
[CI017, CI018, CI019, CI020, CI021, CI026]| 指标 | 数值 / 状态 | 置信度 | 重要性 | 尽调要求 |
|---|---|---|---|---|
| 已确认收入 | 低 | 没有已确认收入,增长质量和经营杠杆无法承销。 | 提供 2025-2026 年按月、按收入流拆分的已确认收入。 | |
| 毛利率 | 低 | 毛利率决定 Robot Era 更像可扩展的机器人供应商,还是资本密集型集成商。 | 提供产品和服务毛利率,并附 BOM 与人工明细。 | |
| 商业订单簿 | >RMB 500M | 中 | 积压订单说明需求,但不能证明收入转化或现金回收。 | 提供积压订单账龄、取消条款和收入转化情况。 |
| 已披露最大单笔订单 | >RMB 50M | 中 | 大合同能验证企业相关性,也会暴露客户集中风险。 | 披露客户、范围、付款里程碑和利润率结构。 |
| 2025 年出货信号 | 出货 200 台人形机器人;>100 个订单推进中 | 中 | 出货台数是最清晰的运营牵引力代理,但仍不能证明已入账收入。 | 勾稽出货台数、验收台数和开票台数。 |
| 独立模块出货 | 2025 年 XHAND >1,000 台 | 中 | 模块销量可能比全系统部署带来更好利润率,但定价未披露。 | 提供 XHAND ASP、利润率,以及相对独立销售的附加率。 |
| 示意性台数代理 | 用 Unitree ASP 代理测算,>RMB 500M 订单约对应 2,980 台 | 低 | 只是展示规模量级的粗略情景,不是 Robot Era 报告数字。 | 用按 SKU 拆分的实际交付台数和入账台数替代代理测算。 |
| 现金消耗 / 跑道 | 低 | 现金和跑道是关键前瞻变量,但公开来源仍缺失。 | 提供账面现金、月度消耗和基准 / 下行情景跑道。 |
公开证据无法支撑指标时,空值是有意保留。唯一的数字化台数估算是同业锚定代理,不应当作 Robot Era 披露统计。
[CI015, CI016, CI027, CI028, CI033, CI036]从制造和部署映射到毛利的成本与价值逻辑,同时显示 Robot Era 的关键数字输入仍未披露。
该图为定性图,因为 Robot Era 没有披露实际 ASP、收入成本、服务负担或现金转换。同业文件只能提供背景,不能替代模型。
[CI016, CI017, CI018, CI019, CI020, CI026]展示有来源支撑的轮次规模、商业订单信号和相邻人形机器人价格锚点,这些边界塑造了 Robot Era 的承销框架。
2026 年项目混合了约人民币 10 亿元轮次折算出的美元等值下限,以及一轮超过 USD 200 million 的融资。商业订单项目混合了单笔已披露订单和累计订单簿数字,因此是背景区间,不是时间序列。
[CI005, CI010, CI011, CI018, CI019, CI029]4.3 资本充足性、融资节奏,以及同行文件对烧钱的启示
公开融资能力不是 Robot Era 的主要弱点。可支撑的时间线显示,截至 2025 年 11 月三轮融资合计 CNY 1.8 billion,之后 2026 年 3 月完成 RMB 1 billion 战略轮,2026 年又完成一轮超过 USD 200 million 的融资。这个节奏对一家私有人形机器人创业公司来说异常强,并由一组与潜在渠道和客户生态重叠的战略投资人加强。但同样的节奏也意味着资本密集。短时间连续大额融资,通常说明管理层在为制造、部署和营运资金需求提前融资,而不是业务已经自我供血。公开同行能解释原因。Serve Robotics 有 SEC 支撑的业绩披露显示,收入有限,同时经营和投资现金流出很大。Symbotic 即便有数十亿美元收入,仍强调流动性、利润率和自由现金流。iRobot 的 10-K 记录了硬件机器人经济性在收入不及预期时能多快恶化。UBTECH 文件证明,人形机器人收入和公司毛利率可以在规模上披露,也让 Robot Era 对现金、债务和资产负债表细节的沉默更显眼。务实解读是:Robot Era 很可能有足够外部资本在近期继续扩张,但管理层之外没人能仅凭公开记录负责任地算出精确跑道。[CI001, CI006, CI010, CI011, CI029, CI030]
| 项目 | 公开数值 / 状态 | 置信度 | 重要性 | 尽调要求 |
|---|---|---|---|---|
| 2025 年 A+ 轮 | 2025 年 11 月由吉利资本领投,接近 CNY 1B | 中 | 显示 2026 年加速阶段前已有强工业资本背书。 | 确认准确交割日期、估值、优先权结构和董事会权利。 |
| 截至 2025 年 11 月融资 | 一年多内三轮合计 CNY 1.8B | 中 | 确定进入 2026 年时的资本基础。 | 提供完整逐轮 cap table 和投资人持股。 |
| 2026 年 3 月战略轮 | RMB 1B;估值 >RMB 10B | 中 | 新资金和估值重置表明赞助方兴趣强。 | 确认投后估值和扣费后净融资额。 |
| 2026 年后续融资轮 | >USD 200M,由顺丰集团领投,HSG 和 IDG 参与 | 中 | 战略轮后不久又补充了可观总资本。 | 披露准确交割日期、金额、估值和资金用途。 |
| 账面现金 / 月度消耗 / 跑道 | 低 | 这是决定前瞻偿付能力的变量,公开仍未披露。 | 提供当前现金、月度消耗,以及基准和下行情景跑道。 | |
| 债务 / 担保 / 项目融资义务 | 低 | 隐性义务可能大幅改变稀释和流动性风险。 | 披露债务额度、担保、供应商融资和或有负债。 | |
| 同业消耗代理 | Serve 披露规模化前有重大经营和投资现金流出;公开同业仍突出流动性和利润率 | 低到中 | 同业披露说明,反复融资可能反映硬件真实资本密集度,而不只是可选性。 | 将 Robot Era 与内部消耗、资本开支和营运资金计划对标。 |
纳入历史融资只为框定前瞻资本充足性。公开现金、消耗和债务仍缺失,因此仅靠公告无法精确建模偿付能力。
[CI001, CI006, CI010, CI011, CI029, CI030]说明在收入质量可见之前,私募股权轮次可能如何为制造、部署和营运资本提供资金。
这张地图刻意保持结构性。保留的公开证据足以描述可能的现金用途,但不足以精确测量当前现金或 runway。
[CI029, CI030, CI032, CI033, CI039, CI043]4.4 披露质量、反向证据与仍最重要的尽调阻塞项
公开档案强在商业化叙事,弱在财务卫生。Bain、McKinsey、Interact、SCIO、OFWeek 和 KrASIA 的反向报道都指向同一方向:早期人形机器人需求存在,但商业化仍集中在结构化环境,ROI 仍脆弱,电池和灵巧性仍是关口,投资人热情可能跑在证据前面。Robot Era 自己的披露也符合这个模式。公司清楚谈订单、部署和产业投资人,但保留来源没有披露已确认收入、已实现 ASP、毛利率、现金、债务、积压订单账龄、质保准备或计费机制。投资人面对的是一家商业上看起来真实、财务上披露不足的公司。因此,尽调阻塞项很具体,而不是抽象:按收入流的收入确认、按产品和服务的毛利率、现金和月度烧钱、大型部署的营运资金条款,以及已报告订单簿的真实转化质量。在这些交付之前,本章最有支撑的判断是:需求形成和融资能力为正面,经济性和跑道仍需谨慎。[CI015, CI016, CI023, CI024, CI025, CI038]
| 缺失的私有指标 | 对分析的影响 | 当前公开状态 | 精确尽调路径 |
|---|---|---|---|
| 按收入流拆分的已确认收入 | 阻碍承销增长质量、结构,以及积压订单向收入的转化。 | 保留的 Robot Era 来源中未公开披露。 | 要求按月、按季度提供机器人、模块、软件、集成和支持收入。 |
| 实际 ASP 与折扣 | 阻断任何扎实的单位经济性或定价权分析。 | 没有保留公开的 Robot Era 标价或已实现 ASP。 | 按 SKU、地区和客户分部取得发票级 ASP。 |
| 毛利率和成本栈 | 阻断利润率路径和任何可扩展经济性的判断。 | 未公开披露。 | 要求产品和服务毛利率,以及 BOM、人工和保修明细。 |
| 现金余额、月度消耗和跑道 | 融资依赖无法用数字量化。 | 未公开披露。 | 取得 24 个月现金桥、当前流动性报表和下行跑道情景。 |
| 积压订单账龄和合同转化 | 阻碍判断订单质量和收入时点。 | 公开只有总订单信号和出货轶事。 | 审查头部客户合同、取消条款和积压订单转化历史。 |
| 营运资金与计费机制 | 即便积压订单看起来很大,也可能实质性缩短跑道。 | 应收、库存、里程碑或准备金细节均未公开。 | 要求 AR 账龄、库存周转、计费安排和保修准备金政策。 |
| 债务、担保和或有负债 | 未披露义务可能改变跑道和下行空间。 | 没有保留公开债务或担保披露。 | 审查债务计划、契约、担保和供应商融资安排。 |
每一行都是具体尽调交付物,而非泛泛请求,因为公开记录商业化信号丰富,但财务机制很薄。
[CI015, CI016, CI038, CI040, CI043, CI044]05产品与技术
5.1 产品组合与当前公开规格底线
对 Robot Era 最有支撑的公开认知是一个模块化产品家族,而不是单一旗舰机器人。保留的直接来源一致支持四个当前锚点:L7 全尺寸双足人形机器人、M7 上半身人形配置、Q5 轮式人形机器人,以及 XHAND1/XHAND1 Lite 灵巧手系列。官方网站资料包和 CES 2026 发布让层级异常清晰:L7 承载全尺寸、55-DoF 人形叙事;M7 是上半身部署变体;Q5 是轮式服务和研究平台;XHAND1 是贯穿技术栈的精密操作模块。这种模块化框架重要,因为它说明 Robot Era 正试图用不同实体形态解决多种近期部署约束,而不是强迫所有工作负载都通过同一种人形形态。已披露规格底线也相当扎实。L7 反复被描述为 171 centimeters、55 个自由度、20 kilograms 双臂负载;Q5 反复被描述为 44 个自由度、约 165 centimeters。XHAND1 明确营销为 12 个主动 DoF、全直驱手,M7 则被营销为源自同一架构的上半身平台。主要缺口是范围纪律:保留的当前直接来源不能给出可支撑的公开 XBot-L 产品表面,因此本章不把 XBot-L 视为已验证的在售模块。[CE001, CE002, CE003, CE006, CE007, CE009]
| 模块 / 资产 | 主要用户 / 工作流 | 当前公开规格底线 | 成熟度 / 状态 | 差异化信号 | 尽调缺口 |
|---|---|---|---|---|---|
| ERA-42 具身模型 + 物流解决方案 | 仓库运营商集成拣选 / 扫描 / 装箱工作流 | 端到端 VLA 框架;官方工作流覆盖从订单释放到出库 | 公开叙事已上线;真实部署架构仍部分不透明 | 工作流集成语言比泛泛的人形 AI 宣传更具体 | 没有公开模型卡、延迟基准或安全边界 |
| L7 全尺寸人形机器人 | 物流、制造、研究和高曝光演示部署 | 171 cm;55 DoF;20 kg 双臂载荷;2.1 m 臂展;L7/L7-M7 共享架构 | 支撑最充分的旗舰产品;多家外部来源覆盖 | 结合全尺寸运动、直驱手集成和企业工作流叙事 | 公开重量、续航和认证披露仍不一致或不完整 |
| M7 上半身人形机器人 | 腿部价值有限的拣选、扫描、装配、零售和研究场景 | 43 DoF(含手);1660 mm 臂展;2.1 m 作业范围 | 已公开列出,概念清晰,但独立文档少于 L7 | 复用核心臂 / 手栈,同时降低部署和集成负担 | 少有独立来源描述生产成熟度或客户结构 |
| Q5 轮式人形机器人 | 接待、服务、医疗、研究、遥操作和室内交互 | 44 DoF;约 165 cm;轮式底盘;11-DoF XHAND Lite;保留来源中作业时间 >4 h / 240 min | 官方和独立规格数据库均有支撑 | 比完整双足更适合室内服务和数据采集部署 | 公开定价、现场可靠性和准确软件边界仍薄 |
| XHAND1 灵巧手 | 研究实验室、人形机器人 OEM,以及 Robot Era 自有平台 | 12 主动 DoF;全直驱;80 N 抓握力声明;25 kg 提升力声明;触觉指尖;支持 ROS/C++/Python | 栈中公开文档最充分的模块 | 最清晰证明公司有自研部件野心,也具备跨平台复用性 | 企业认证、校准漂移和使用寿命数据仍稀疏 |
| XHAND1 Lite | Q5,以及更轻量的研究或遥操作配置 | 面向 Q5 的来源显示 11 DoF;730 g;官方套件称 100 万次空载抓取循环 | 公开可见的配件 / 模块,但文档完整度低于 XHAND1 | 更轻的手拓宽了轮式服务机器人的部署菜单 | 公开力值、负载下耐久性和客户部署细节仍有限 |
该矩阵只纳入本轮保留直接 URL 可支撑的模块。XBot-L 被排除,因为保留的当前公开记录没有提供稳定、可支撑的直接产品页面。
[CE001, CE002, CE003, CE006, CE009, CE010]Robot Era 主要模块的公开成熟度视图,区分公司已有具体部署证明的部分和披露仍以叙事为主的部分。
[CE001, CE018, CE019, CE026, CE033, CE034]5.2 ERA-42 架构表述与软硬件闭环
Robot Era 最清晰的技术故事不是完整模型卡,而是一个运行闭环描述。在官方解决方案页面,ERA-42 被描述为端到端 VLA 模型,把具身大脑连接到自动化系统,并覆盖仓库从订单下发到出库、拣选、扫码和装箱的序列。官方文本还描述了双向闭环:业务系统下发任务,模型回传进度和状态,标准化接口层应减少定制再训练。这比泛泛的「AI 驱动人形机器人」更具体,因为它把软件锚定在真实企业流程和集成要求上。独立报道补充了更多色彩,但没有给出完整透明度。Robotics & Automation News 称,Robot Era 把 ERA-42 描述为统一的视觉-语言-触觉-姿态模型,与 XHAND1 匹配,能够快速学习任务,并由世界模型概念扩展;KrASIA 又转述了快慢分层架构和 Ctrl-World 合作。GitHub 证据还显示,公司发布 ROS 2 和 Python 控制界面、轨迹和 MPC 示例,以及手部控制 API。因此,承销结论是正面但有边界:公开来源支持真实的软硬件闭环,但没有发布投资人通常想要的核心内部信息,如参数量、基准方法、安全边界,或负载下实测延迟。[CE020, CE021, CE022, CE023, CE024, CE025]
| 层 / 组件 | 公开披露角色 | 依赖 / 接口 | 公开证明质量 | 主要技术风险 |
|---|---|---|---|---|
| 全身 / 上半身 / 轮式平台 | L7、M7 和 Q5 为不同工作流提供形态选择 | 在部署变体之间共享通用手部、感知和控制概念 | 中到高 | 公开形态策略比真实制造复杂度或现场服务负担更清晰 |
| XHAND 末端执行器层 | 直驱五指操作、触觉感知和可用工具的手部控制 | 依赖关节模块、遥操作和机器人侧控制栈 | 高 | 内部实验室指标强于独立耐久性证明 |
| 感知 + 机载计算 | LiDAR、深度视觉、麦克风、x86 + Orin 算力和实时控制输入 | 依赖传感器融合、机载推理和机身控制回路 | 中 | 公开来源更容易披露组件,而不是延迟、冗余或故障处理 |
| ERA-42 具身模型 | 面向感知到行动的端到端 VLA / 世界模型 / 快慢控制叙事 | 依赖训练数据、仿真、手 / 身执行和工作流集成 | 中 | 公开架构叙事超过已发布基准和模型卡细节 |
| 工作流集成层 | 连接 WMS 或客户业务系统,把业务任务转成机器人任务,并回传状态更新 | 依赖标准接口、状态码和客户侧流程映射 | 中 | 公开资料没有披露具体 API 规格、部署手册或错误预算 |
| 数据 / 训练 / 数字孪生闭环 | 遥操作数据采集、处理、模型训练、仿真检查和闭环验证 | 依赖真机数据、开源预训练和操作员工具 | 中 | 公开披露没有量化数据规模、标注质量或 sim-to-real 命中率 |
| 开发者 / 控制界面 | 通过公开代码仓库提供 ROS2 services/actions、轨迹控制、MPC 和手部 / 底盘 API | 依赖 ROS2、CycloneDDS、teleop_client 以及内部开发环境假设 | 中 | 公开 SDK 确实存在,但仍不能完整代表企业集成文档 |
公开证据质量只反映本次保留资料。“高”不等于完全披露;它表示该层至少有一个直接技术界面,并有旁证支持。
[CE004, CE005, CE014, CE015, CE017, CE019]Robot Era 公开披露的技术栈:从开发者 / 控制界面和数据闭环,经 ERA-42,延伸到机器人本体和工作流集成。
[CE019, CE020, CE021, CE022, CE023, CE024]Robot Era 最具体的公开运营流程是仓储单元:ERA-42 连接 WMS 派单、人形机器人拣选、扫码和状态反馈。
[CE020, CE021, CE026, CE027, CE028, CE029]5.3 模块化、自研部件深度与公开开发者界面
保留记录中最可辩护的产品差异化,是由异常激进的垂直整合主张支撑的模块化。官方来源显示,L7 和 M7 可在全身与上半身部署之间重组,Q5 和 XHAND 系列则把技术栈拓展到服务机器人和独立操作。KrASIA 与 PRNASIA 随后推动更深一层逻辑:据称超过 95% 部件自研,包括关节、灵巧手和电机。XHAND1 是这个主张中暴露最充分的证明点,因为官方资料包不止给出营销形容词,还列出触觉覆盖、控制频率、电气范围、通信接口、ROS 支持和远程操作路径。这让手部成为比更大人形身体更强的尽调锚点;后者公开规格表面仍较薄。开发者信号也真实存在,但仍有限。Robot Era 有公开 GitHub 组织、官方 ROS 2 示例仓库,以及一个引用 Q5、L3 和 L7 模型控制、XHand/XHand Lite API 的 Python SDK。与此同时,官网文档中心有表单门槛,公开仓库还不能替代完整现场文档、集成手册或透明可靠性仪表盘。换句话说,Robot Era 看起来比纯封闭 demo 硬件厂商开放得多,但还没开放到足以让外部人完整承销部署复杂度。[CE002, CE012, CE014, CE015, CE016, CE017]
| 日期 / 阶段 | 产品 / 里程碑 | 当前状态 | 产品技术含义 | 来源 |
|---|---|---|---|---|
| 2024-12 公开模型发布 | 独立报道中对 ERA-42 的公开定位 | 已发布 / 已公开描述 | 在 2025-2026 年更重的商业化推进前,先建立 VLA 叙事 | Robotics & Automation News |
| 2025-07 产品发布 | L7 全尺寸人形机器人公开亮相 | 已发布 | 标志着公司从早期机器人平台转向全尺寸人形产品界面 | China Daily |
| 2025 物流解决方案披露 | CeMAT 期间围绕 L7 + ERA-42 的仓库拣选工作流 | 已公开展示 / 试点阶段 | 在保留记录中提供最清晰的真实工作流切入口 | Humanoids Daily / RoboHorizon |
| 2025 模块化家族扩展 | XHAND1、XHAND1 Lite、Q5、L7 和 M7 均出现在模块化组合报道中 | 已发布 / 产品组合公开 | 显示 Robot Era 在搭建可复用模块栈,而不只是做单一机身演示 | 官方资料包 / KrASIA |
| 2026-01 CES 展示 | Hexa-Core 产品线:L7、Q5、XHAND1 | 已发布 / 已展示 | 在全球展会上更新公开规格底线和采用证据 | PRNewswire |
| 2026 公开代码界面 | GitHub 组织、ROS2 SDK 和 Python SDK 仍然公开 | 在线 | 表明除静态营销页面外,公司仍在支持开发者和集成 | GitHub |
| 2026 Q2 交付爬坡 | 千台级交付,以及 300%+ 销售增长主张 | 公司称正在推进 | 释放规模野心和部署相关性,但仍需要经审计的现场质量证据 | PRNASIA / Automated Warehouse / CnTechPost |
该表是产品与开发时间线,不是融资时间线。重点放在会改变技术尽调判断的发布、公开工作流披露、公开代码界面和扩产主张。
[CE024, CE025, CE026, CE029, CE033, CE034]Robot Era 的产品论点与其说依赖某一具机器人本体,不如说依赖把数据、控制软件、自研机电和客户系统连成稳定运营闭环。
[CE019, CE020, CE021, CE023, CE028, CE032]5.4 各实体形态的部署相关性与产品成熟度
当 Robot Era 把问题收窄到结构化流程,尤其是物流拣选、工业搬运和受控服务环境时,部署逻辑最强。Humanoids Daily 和 RoboHorizon 都描述了一个具体仓库单元:货到人系统把料箱送到可触达范围后,L7 接管柔性料箱拣选、条码扫描、装箱和异常处理。相比开放式家庭机器人叙事,这个切口可信得多,因为它只在刚性自动化仍失效的地方使用人形操作。Q5 和 M7 也因同样原因重要。Q5 的轮式底盘和高 DoF 上半身,使它在研究、酒店、接待和导览服务部署中比拥挤室内空间里的全双足更可信;M7 则为拣选、扫码和装配式工作降低了实体负担,因为这些场景里腿不是主要价值驱动。公开采纳证明也优于年轻硬件公司平均水平:PRNewswire 称累计交付超过 600 台,China Daily 提到 2025 年出货 200 台人形机器人且 100 多个订单推进中,PRNewswire 和 KrASIA 也指向 XHAND1 周边外部研究采纳和开发者生态。不过,成熟度图景仍混合。公开记录支持部署相关性和产品广度,但还不能完整证明可重复正常运行时间、安全性、客户 ROI 或广泛开放世界自主性。[CE008, CE010, CE011, CE013, CE018, CE026]
| 用户任务 / 环境 | 当前工作流问题 | Robot Era 方案 | 公开可衡量收益 / 证明点 | 主要限制 |
|---|---|---|---|---|
| 货到人仓库拣选 | AGV 送来料箱,但混合 SKU 拣选和扫码处理仍需要人类灵巧性 | L7 + XHAND1 + ERA-42 + WMS 联动工作流 | 公开来源称机器人覆盖最高 2.1 m 作业空间,并处理拣选 - 扫描 - 装箱 / 回收回退步骤 | 没有公开客户级生产率、拣选速率或扫码准确率数据集 |
| 工厂搬运 / 装配 / 检验 | 灵活工业任务仅靠刚性单用途机械臂更难自动化 | L7 处理全身任务;上半身可达性和灵巧性比移动更重要时用 M7 | 公开来源反复点名拧螺丝、装配、拣选和检验任务 | 公开证明停留在场景层面,不是经审计的吞吐或在线率 |
| 室内服务 / 接待 / 医疗 / 旅游 | 人机互动、导航和物品递交需要比固定服务亭更安全、更有表现力的形态 | Q5 轮式人形机器人,配高 DoF 机械臂、XHAND Lite、视觉和遥操作支持 | 公开来源提到导览、接待、配送、医疗辅助和零售互动 | 公开定价和服务维护经济性仍薄 |
| 研究 / 遥操作 / 数据采集 | 实验室需要可控平台,用于模仿学习、遥操作和闭环验证 | Q5、M7、XHAND1,以及公开 ROS2/Python 控制接口 | 官方和 GitHub 来源展示了遥操作、MPC、轨迹和手部控制路径 | 公开文档仍不完整,部分官方手册仍设门槛 |
| 独立灵巧操作 | 许多实验室和 OEM 需要可复用的手,而不是完整人形机身 | XHAND1,配直驱、触觉指尖,并支持 ROS/C++/Python | 公开来源提到外部采用者,包括 Skild AI、Extend Robotics、UC Berkeley、MIT 和 ByteDance Seed | 公开校准、故障率和量产服务数据仍未披露 |
可衡量收益只限保留来源直接陈述的内容,例如工作空间覆盖、已发布载荷或已命名工作流步骤。若公开只有定性收益,限制栏记录缺失的分母。
[CE011, CE018, CE019, CE026, CE027, CE028]5.5 信任、质量控制与主要技术限制
Robot Era 的公开产品技术叙事强在能力框架,弱在质量披露。官网有真实隐私和采购界面,保留来源也清楚显示,可下载材料、API、遥操作路径和 SDK 都存在。但同一批保留来源没有提供企业买方和投资人在承销车队规模前会期待的更深控制:保留的当前公开来源没有给出正式安全认证、防护等级、现场故障率、维护间隔、质保条款、网络安全架构,或 ERA-42 的经审计基准包。独立技术数据库明确把 L7 部分字段留空,外部行业分析师也反复回到同一批关口:先从结构化环境起步,灵巧性仍难,电池续航仍短,狭窄高价值流程之外经济性仍不清楚。KrASIA 的反向行业文章尤其有用,因为它指出部分汽车部署可能仍是战略绑定,而不是规模化销售。正确尽调解读应保持平衡。Robot Era 似乎拥有比纯炒作项目更可信的具身 AI 技术栈,但公司仍需要证明其技术主张能在规模化安全、可靠性和经济性审查中站住。[CE032, CE045, CE046, CE047, CE048, CE049]
| 控制 / 披露领域 | 公开状态 | 范围 | 重要性 | 缺口 / 下一步尽调问题 |
|---|---|---|---|---|
| 隐私 / 数据同意界面 | 官方已提供 | 官网采购与下载表单 | 说明公司已为企业入站互动搭起基础隐私和同意层 | 审查完整的数据留存、删除和访问控制政策,范围应覆盖运营机器人数据,而不只是网页表单数据 |
| 公开文档 / 手册 / SDK 参考 | 部分提供 | 需填表获取的文档中心,以及公开 GitHub 代码仓库 | 表明产品和开发者材料不止停留在营销主页 | 获取完整手册、版本历史和 API 兼容矩阵 |
| 安全认证 / 标准披露 | 保留资料未公开披露 | 当前保留资料没有出现 L7/Q5/M7 的 CE、ISO 或同等级机器人安全认证 | 企业部署通常依赖正式安全评审和集成签核 | 索取全部安全认证、风险评估,以及急停 / 失效安全文档 |
| 防护 / 环境等级 | 未公开披露 | 独立数据库中 L7 的 IP / 防护字段为空 | 环境边界会影响仓库和工厂的在线时间 | 按 SKU 索取 IP 等级、温度范围、防尘能力和清洁要求 |
| 可靠性 / 生命周期指标 | 稀薄且不均衡 | XHAND1 和 XHAND1 Lite 披露了部分循环寿命和温度数据;机队 MTBF 和机器人在线时间未公开 | 模块级耐久度不等于系统级可靠性 | 索取 MTBF、服务间隔、备件计划和保修退货历史 |
| 安全 / OTA / 漏洞处置姿态 | 未公开披露 | 保留公开资料没有描述软件更新控制、认证模型或漏洞披露机制 | 接入企业系统的联网机器人会带来网络安全和运营风险 | 索取安全架构、OTA 流程、SBOM 和披露政策 |
| 保修 / 支持 SLA | 未公开披露 | GitHub 仓库有支持邮箱;公开服务承诺没有披露 | 买方信任取决于现场支持和更换响应时间 | 索取 SLA 条款、支持团队配置、地域覆盖和备件物流 |
| 性能基准 / 经审计 ROI | 未公开披露 | 公开资料描述了工作流和试点,但没有经审计的生产率、错误率或 ROI 材料包 | 没有基准纪律,能力主张很难跨供应商比较 | 索取客户 KPI 材料包、验收测试和部署前后指标 |
保留公开资料不足以支持正式披露时,表中有意保留 null 式缺口。该表区分网页 / 隐私控制、开发者界面和缺失的运营质量披露。
[CE019, CE037, CE045, CE046, CE047, CE048]5.6 展示材料
06客户情况
6.1 客户图谱:物流是主要切口,但公开 logo 集覆盖多类买方
不应把 Robot Era 的公开客户证据看成一张同等生产客户的扁平名单。最强买方类别是物流运营商;Robotera 反复描述与 SF 关联实体和 China Post 相关的包裹分拣、小件包裹导入和跨境查验流程。第二类是工业制造:Geely 是最清晰的具名例子,公开来源也把 Renault、TCL 和 Lenovo 放进更广泛产业伙伴集合。第三类是商业服务或零售运营,Haier 是最具体的证明,因为独立报道称共同开发的机器人已经进入零售门店。第四类是开发者生态,ByteDance 机器人实验室和 Skild AI 显示的是平台或研究采纳,而不是工厂或仓库经常性收入。最后,公司声称拥有更广泛的国际顶级科技客户足迹,但多数名称仍未披露。因此,正确承销视角应分层。物流客户显示最佳生产意图。物流之外的工业 logo 有帮助,但往往记录不足。开发者实验室买方有战略验证价值,但不应被算作与规模化仓库或工厂部署同等质量的证明。[CU005, CU009, CU011, CU014, CU026, CU027]
| 分层 | 具名界面 | 买方 / 用户 / 付款方 | 公开用例 | 当前判断 | 主要缺口 |
|---|---|---|---|---|---|
| 物流运营商 | SF Group / SF Express;China Post(客户 / 合作方) | 物流运营商或海关流程负责人付款;仓库员工和包裹网络是用户 | 包裹导入、分拣、扫描、跨境查验 | 客户证据中确定性最高的一类 | 仍缺客户自述证据和收入集中度 |
| 汽车与工业制造 | Geely;Renault | 工业运营商付款;生产团队是用户 | 零部件拣选、装配、质量检查 | 可信,但细节只披露了一部分 | 工厂数量、吞吐量和合同金额未公开 |
| 消费电子制造 | TCL | 工厂运营商付款;产线工人和工程师是用户 | 精密制造 / 兼容洁净室的自动化 | 已具名,但披露很薄 | 没有公开的站点级案例研究 |
| 商业服务 / 零售 | Haier 零售合作 | 运营商付款;门店员工和访客是用户 | 店内清洁和导览功能;零售门店机器人 | 比多数非物流 logo 更具体 | 规模和经济性仍未披露 |
| 企业开发者 / 研究实验室 | ByteDance robotics lab;Skild AI;未具名头部科技客户 | 实验室或开发者买方付款;机器人研究人员是用户 | 研发、技术转移、具身 AI 实验 | 具备战略验证价值,但不等同于仓库收入 | 名称和收入贡献大多未披露 |
| 更广泛的国际客户 | 前十大科技公司中九家的主张;海外地区已具名 | 海外市场买方未披露 | 中国以外的硬件平台和具身 AI 采用 | 支撑全球需求叙事 | 客户名称、订单结构和续约未公开 |
各行把生产运营商与开发实验室买方分开,避免把每个 logo 都算作同等质量的复购客户证据。
[CU002, CU005, CU011, CU012, CU013, CU014]Robot Era 的公开客户路径最常从高价值物流或工业痛点开始,之后才变成复购或多元化收入证明。
该地图概括公开来源暗示的推进路径;它不是已披露的内部销售漏斗。
[CU004, CU006, CU017, CU020, CU021, CU030]6.2 规模证明:物流是最清晰的试点到生产路径,海外需求看起来真实但仍定义松散
本章最强的采用证据不在泛泛的品牌客户名单里,而在物流场景。Robotera 2026 年 3 月和 5 月的自有公告称,公司正在 10 多个物流中心落地,并从 2026 年第二季度开始交付千台级设备。独立报道大多复述了同一模式,而不是推翻它:Humanoids Daily、RoboticsTomorrow、Automated Warehouse 和 Mike Kalil 都提到,公司正与 China Post 及 SF 相关实体在物流中心部署,并附带 24/7 运行、效率约达人类 85% 等运营细节。产品侧叙事也能对上。CeMAT ASIA 2025 上,公司把 L7 定位为解决仓库柔性拣选缺口的方案,包含 WMS 集成、条码扫描和异常路由,比通用跳舞演示更接近真实运营模块。国际需求也有分量。公开资料称,海外业务约占收入或订单额一半,并点名北美、欧洲、中东、日本和韩国。不过,分母口径并未持续披露,因此更稳妥的解读是:国际需求可信,但还没有经过干净审计。[CU001, CU002, CU004, CU006, CU007, CU012]
| 信号 | 公开细节 | 日期 / 阶段 | 来源依据 | 含义 | 缺失分母 |
|---|---|---|---|---|---|
| 累计订单 | Robotera 称累计订单超过人民币 5 亿元,且部分客户最多复购六次 | 2026 披露状态 | PR Newswire / PRNasia | 需求已越过概念阶段 | 未按客户或分层拆分 |
| 已披露最大物流订单 | 最大单笔物流订单约人民币 5,000 万元;与 SF 相关的海关解决方案已获官方部署 | 2025-2026 公开报道 | PR Newswire + KrASIA + Humanoids Daily(来源组合) | 一个物流客户看起来具有经济意义 | 该订单是否复购或高度集中仍不清楚 |
| 物流中心覆盖 | 与 China Post 和 SF 相关主体合作超过 10 个物流中心 | 2026 当前状态 | PR Newswire + Humanoids Daily + Automated Warehouse(来源组合) | 暗示部署模板可复制 | 没有客户侧中心清单 |
| 千台级交付 | 2026 Q2 千台级交付,增长超过 300% | 2026 Q2 | PR Newswire + 独立行业媒体 | 表明从试点走向机队部署 | 未按客户拆分出货量 |
| 运营效率 | 某些物流环境中效率最高可达人工水平的 85%,并可 24/7 运行 | 2026 当前状态 | Humanoids Daily + Mike Kalil | 至少有一个公开性能指标 | 没有吞吐量、ROI 或在线时间 SLA |
| 2025 出货 | 2025 年出货 200 台人形机器人,另有 100 多个订单在推进 | 2025 当前状态 | China Daily + EqualOcean | 单篇新闻稿之外也存在更广泛的商业活动 | 没有客户级出货映射 |
| 国际结构 | 约一半订单或收入来自海外,北美、欧洲、中东、日本和韩国已具名 | 2025-2026 当前状态 | KrASIA + Yicai + Humanoids Daily | 国际需求看起来真实存在 | 各来源使用的分母不同 |
该表记录已披露商业信号,并明确标出分母或经常性收入语境仍不可得之处。
[CU001, CU002, CU004, CU006, CU007, CU010]公开证据从大量具名 logo 到极少数已披露耐久指标,迅速收窄。
计数只概括已审阅的公开材料,并刻意区分 logo 覆盖面和已披露耐久指标。
[CU001, CU006, CU007, CU017, CU025, CU035]6.3 按客户点名验证:SF 相关物流证据强,Geely 和 Haier 中等,其他若干 logo 偏薄
逐个客户看,证据质量差异很大。SF 相关物流是置信度最高的一行,因为证据包同时包含点名客户路径、海关部署表述、大额披露订单规模和千台级交付说法。China Post 也反复出现在物流中心叙事里,但公开证据仍主要来自公司侧和媒体侧,而不是 China Post 自身,因此这行真实但仍不完整。Geely 是下一个最强的点名 logo,因为它反复被界定为战略投资方和产业伙伴,且资料描述了相关制造场景。Haier 比多数剩余工业客户更具体,因为独立报道称双方共研的机器人已经进入零售门店。Renault、TCL 和 Lenovo 的意义在于把 logo 组合扩展到汽车、消费电子和企业科技,但每一家仍缺少特定场站的公开案例研究。ByteDance 和 Skild AI 提供了另一类证明:它们说明 Robotera 正在把系统卖入或投放到先进机器人研发环境,但应计作开发者或实验室采用,而不是与规模化物流合同同等质量的证明。[CU004, CU005, CU009, CU014, CU015, CU019]
| 客户 / 采用方 | 分层 | 部署 / 用例 | 生产 vs 试点 | 结果 / 证据 | 局限 |
|---|---|---|---|---|---|
| SF Group / SF Express | 物流 | 跨境查验、包裹导入和分拣工作流 | 已披露最强的从试点到规模化路径 | 海关部署表述、超过人民币 5,000 万元的单笔订单,以及千台级交付叙事 | 没有客户自述 KPI 或续约披露 |
| China Post | 物流 | 在 10 多个物流中心网络中运行人形机器人 | 有生产规模主张,但客户侧证据仍薄 | 官方与行业报道反复把它列为物流中心部署对象 | 本资料包中没有 China Post 直接点名 Robotera 的声明 |
| Geely | 汽车 / 工业制造 | 零部件拣选、装配和质量检查场景 | 可能是生产实验,尚未充分量化 | 反复被列为契合工业用例的战略伙伴 | 没有公开站点数量或合同规模 |
| Haier | 商业服务 / 零售 | 共同开发的机器人已进入零售门店 | 早期商业部署 | 具体零售门店用法超过泛泛 logo 提及 | 门店数量和经济性未披露 |
| Renault | 汽车制造 | 制造栈中的工业伙伴 | 具名关系,证据质量中等 | 多个来源把 Renault 列为工业伙伴 | 没有公开部署 KPI 或站点细节 |
| TCL | 消费电子制造 | 精密制造 / 兼容洁净室的自动化 | 具名关系,证据质量中等 | 独立报道把 TCL 与制造自动化用例放在一起 | 没有公开铺开数量或客户引述 |
| Lenovo | 企业技术 / 工业生态 | 工业伙伴和投资方相关生态客户 | 具名关系,证据质量中低 | 在伙伴集合中反复出现 | 没有公开部署细节 |
| ByteDance 机器人实验室 / Skild AI | 开发者生态 | 对 Robotera 系统的研发和技术转移采用 | 真实开发者采用,不是生产物流 | 具名实验室用户显示该平台与先进机器人团队相关 | 不应计作稳定仓库收入 |
枚举范围覆盖本章本地资料包中保留的全部具名外部客户、伙伴或采用方界面;未具名的“头部科技公司”主张不纳入,因为名称未公开。
[CU004, CU005, CU006, CU009, CU014, CU017]证明质量在顺丰相关物流上最强,在多数具名 logo 的收入耐久性上最弱。
评级反映公开来源的证明质量,而不是客户质量;收入可见度低,意思是缺少披露,不代表该客户价值低。
[CU017, CU025, CU030, CU031, CU032, CU033]6.4 耐久性:logo 组合有希望,但公开披露离可审计客户经济性还差得很远
本章最大的判断点不是 Robot Era 有没有真实客户,而是披露证据是否足以推断出持久、多元的生产收入。这个问题的答案仍是否定的。作为早期人形机器人公司,Robot Era 确实给出了少见的具体部署说法,尤其在物流场景里,甚至有一些头部复购表述。但证据包仍缺少客户数、合同期限、NRR、GRR、流失、续约率和头部客户占比。它也没有把工业生产收入与开发者实验室销售或研究平台采用清楚拆开。反向行业资料让这一区分更重要。IEEE Spectrum、McKinsey、Robotics & Automation News 和 Forbes 都提醒,人形机器人 logo、试点和展示视频可能跑在运营经济性前面。这些资料没有推翻 Robot Era 的部署,但确实反驳了投资人把强 logo 页直接等同于持久收入结论的冲动。务实综合看法是建设性但克制:Robot Era 似乎拥有人形机器人领域较好的公开客户证明包之一,但已披露证据最强的是部署活动,最弱的是集中度、留存和经审计生产收入质量。[CU001, CU017, CU029, CU034, CU035, CU036]
| 指标 / 代理指标 | 数值 | 分层 | 置信度 | 尽调问题 |
|---|---|---|---|---|
| 正式 NRR / GRR / 流失 | 全部分层 | 其未披露这一点置信度高 | 索取按分层划分的客户留存 cohort 和 logo 流失 | |
| 正式客户数量 | 全部分层 | 其未披露这一点置信度高 | 索取活跃客户数、前十大账户和区域结构 | |
| 复购信号 | 据称部分客户最多复购六次 | 国际 / 头部科技结构 | 中 | 索取哪些客户复购,以及它们占 bookings 的比例 |
| SF 相关规模耐久性 | 已有具名订单和部署主张,但续约节奏未公开 | 物流 | 中 | 索取合同期限、扩张路径和已实现 ROI |
| China Post 耐久性 | 物流 | 缺少直接客户侧证据这一点置信度高 | 索取 China Post 背调电话或签署版部署声明 | |
| 开发实验室耐久性 | ByteDance 和 Skild AI 等实验室显示平台有拉力,但不代表经常性工业收入 | 开发者生态 | 高 | 索取开发套件与生产客户之间的收入拆分 |
null 表示审查过的公开资料包没有披露该指标;因此,在可用处只能用复购订单和后续部署表述作为复购代理。
[CU001, CU017, CU025, CU026, CU027, CU035]| 扩张驱动 / 风险 | 重要性 | 当前证据 | 若属实的影响 | 尽调路径 |
|---|---|---|---|---|
| SF 相关物流集中度 | 最清晰的生产路径也可能是最集中的收入流 | 具名海关部署、大额披露订单和物流中心规模 | 可能带来强劲早期增长,但客户集中度风险高 | 索取头部客户占比和合同续约节奏 |
| China Post 证据缺口 | 重大客户标识出现在材料包中,但仍缺少客户亲自确认 | Robotera 侧与媒体侧引用占主导 | 最强的客户广度叙事可能没有看起来那么扎实 | 索取客户侧案例研究或采购记录 |
| 国际业务口径不清 | 海外需求看起来真实,但 50% 这一数字在部分来源中指收入,在另一些来源中指订单 | KrASIA、第一财经和 Humanoids Daily 用词不同 | 可能扭曲对全球牵引力和收入质量的估值 | 对齐同一期间的订单、收入和交付台数 |
| 工业客户标识过度解读 | Geely、Renault、TCL 和 Lenovo 扩大了客户标识集合,但站点级证据稀疏 | 名称被反复提及,运营 KPI 不多 | 可能让客户基础显得比实际更成熟 | 索取逐客户部署地图和 KPI |
| 开发者 / 实验室口径混杂 | ByteDance 和头部科技客户说法有战略价值,但不等同于仓储场景的经常性收入 | 研发采纳是真实的,但经济属性不同 | 如果混入同一个客户数,可能夸大耐久性 | 索取实验室、开发者和工业客户的收入拆分 |
| 行业炒作风险 | 人形机器人行业评论提示,客户标识和演示可能跑在商业价值前面 | IEEE、McKinsey、Forbes 以及 CB Insights 相关报道偏审慎 | 如果经济性不及预期,估值倍数可能收缩,薄弱部署也会暴露 | 尽调应锚定在线率、ROI 和续约证据,而不是客户标识数量 |
风险表把集中度和披露问题分开处理,避免把丰富的客户标识集合误读为多元且经审计的客户质量。
[CU004, CU017, CU030, CU033, CU035, CU036]| 已点名主体 | 材料包中是否有客户亲自发布的来源? | 最佳可用来源 | 能证明什么 | 仍不能证明什么 |
|---|---|---|---|---|
| SF / SF Express | 否 | Robotera 官方发布加独立媒体 | 已点名的部署路径和披露的大额订单规模 | 客户侧 KPI、续约或满意度证据 |
| China Post | 无直接 Robotera 提及 | Robotera 官方发布加 China Post 自动化文章 | China Post 是现实的自动化采购环境,并出现在 Robotera 部署说法中 | China Post 本身已公开点名确认 Robotera |
| Geely | 否 | 独立媒体和 Robotera 侧合作伙伴名单 | 制造业用例匹配,且合作伙伴名称被反复提及 | 工厂数量、吞吐量或合同金额 |
| Haier | 否 | 独立媒体 | 零售门店用例细节,不只是泛泛的客户标识 | 门店数量,能否转化为规模化经常性收入 |
| Renault / TCL / Lenovo | 否 | Robotera 侧与独立合作伙伴名单 | 这些名称把客户标识扩展到多个工业垂直领域 | 任何具体站点的部署结果或客户引述 |
本表单独标出哪些地方只有公司侧证据,帮助区分真正的点名证据和客户亲自发布的验证。
[CU004, CU014, CU031, CU033, CU034, CU039]6.5 展示材料
07风险
7.1 地缘政治、出口管制与监管暴露
Robot Era 在地缘政治层面有两重暴露:一是训练和部署具身 AI 系统所需的算力栈,二是中国制造机器人进入市场时附着的政治问题。公司自己的 L7 手册描述了 x86 加 Orin 的算力栈,以及摄像头、LiDAR 和麦克风硬件,这正是如今处在更对抗性中美政策框架里的先进、传感器密集型具身系统。BIS、CRS 和 Sidley 都显示,美国管制已经从泛泛的半导体话术,推进到对先进计算、Entity List 指定乃至某些 AI 模型权重的实际限制。FDD 和 IEEE Spectrum 补上了市场后果:美国政策制定者现在愿意讨论中国机器人采购禁令和聚焦传感器的限制,而供应链仍相互缠绕,并不能干净切开。对 Robot Era 而言,出口管制风险不只是训练算力问题。它还可能缩窄客户集合、拖慢海外渠道扩张,并提高跨国买家的尽调摩擦。监管碎片化又加重了问题。OSHA 仍没有人形机器人专门标准,EU AI Act 和 Machinery Regulation 提出了正式的健康、安全和权利义务,McKinsey 也认为自动人形机器人的标准栈仍不完整。中国在人形机器人专项标准制定上走得更快,但这并不会简化跨境合规。实际风险因此是中高发生概率、关键影响:Robot Era 可以继续卖进以中国为中心的结构化工作流,但任何假设全球无摩擦扩张的投资逻辑,都缺少当前公开记录支撑。[CR001, CR003, CR004, CR005, CR006, CR007]
| 风险 / 问题 | 司法辖区 / 暴露面 | 状态 | 可能性 | 严重性 | 缓释措施 | 剩余暴露 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| 算力和 AI 出口管制冲击 | 美国 BIS / 跨境算力采购 | 2025-2026 年仍在生效并收紧 | 中高 | 关键 | 优先中国部署、替代供应商、模型效率优化、库存规划 | 高 — 规则正向上游模型权重移动,而 Robot Era 的具身智能栈仍依赖先进算力和传感器可得性 | 获取当前算力 BOM、出口分类备忘录、供应商地图,以及美国来源算力或传感器受限时的应急计划 |
| 美国对中国机器人采购和市场准入限制 | 联邦采购以及可能向下游商业市场外溢 | 政策风险活跃,但尚未形成普遍商业禁令 | 中 | 高 | 避开敏感采购渠道,并按司法辖区本地化 go-to-market | 中高 — 正式禁令到来前,公共政策动能就可能让买方降温 | 询问管理层当前管线中的地域、客户类型和公共部门渠道,以及联邦采购禁令下暴露如何变化 |
| 美国、欧盟和中国之间安全与合规制度割裂 | OSHA 指引、EU AI Act、EU Machinery Regulation、中国人形机器人标准 | 框架已存在,但人形机器人专属规则仍不完整 | 高 | 高 | 采用结构化环境部署、认证前测试和逐市场合规打包 | 高 — 今天没有留存的 Robot Era 专属认证证据能补上缺口 | 在承保海外扩张前,索取 CE/ISO/UL 或等效资料包、风险评估、技术文件和分国家合规计划 |
| 传感器丰富机器人带来的隐私和工作场所数据暴露 | 欧盟 / 美国隐私、AI 与工作场所数据处理 | 只要机器人采集视觉、音频或运营数据,风险就存在 | 中 | 高 | 数据最小化、留存期限、客户 DPA 模板,并尽可能采用端侧处理 | 中高 — 法律审查面在扩大,Robot Era 公开姿态仍薄 | 索取留存计划、子处理方名单、训练数据政策、客户 DPA,以及机器人生成数据的事件响应流程 |
| 产品责任和事故披露风险 | 工厂、仓库和公共空间部署 | 品类风险真实存在;公司专属公开事故历史仍未披露 | 中 | 高 | 受控部署、有防护的工作流、故障安全程序和事故记录 | 中高 — 行业已有真实事故先例,本轮资料中 Robot Era 没有公开安全档案 | 按客户垂直领域审查事故日志、险情报告、急停设计、保险覆盖和合同赔偿条款 |
严重性按对市场准入、部署速度和尽调摩擦的可能影响排序,而不是按 Robot Era 已知公开执法排序。
[CR001, CR002, CR003, CR004, CR005, CR006]政策、技术与商业化风险如何传导到 Robot Era 的收入质量、融资能力和估值。
[CR003, CR006, CR008, CR010, CR012, CR017]7.2 技术、安全、质量与数据治理风险
核心技术承销问题在于,前沿 VLA 叙事仍领先于经现场验证的机器人运营。DeepMind 发布 Gemini Robotics 时仍把它表述为迈向通用机器人技术的一步,并称有用的具身系统需要通用性、交互性、灵巧性和安全行动。相关论文说得更清楚:把大型多模态模型转化为物理智能体仍是重大挑战,更难的灵巧任务或新形态任务仍需要额外微调。Bain、McKinsey 和 Interact 都给出同一方向的反向证据。早期部署仍大多在结构化环境中,电池续航和灵巧性仍是门槛,大规模采用需要跨过安全、正常运行时间、移动能力和成本这些桥,而行业尚未跨过去。这与 Robot Era 直接相关,因为它最好的公开商业证明是物流中心,不是开放世界自主。聚焦结构化物流是理性的缓释手段,但也说明公司还没有摆脱该品类更广泛的技术边界条件。安全和数据治理风险同样真实。OSHA 框架是通用的,不是人形机器人专用;中国标准仍在形成;本次保留的公开资料也没有提供 Robot Era 专属认证包、事故日志、网络安全态势、OTA 控制、MTBF 或正常运行时间指标。与此同时,L7 手册中的摄像头、LiDAR 和麦克风栈意味着每次部署都可能在工作场所和环境中采集数据,而美国法律环境里围绕 AI 和追踪的隐私主张已经在扩张。风险结论是:Robot Era 可能足以胜任狭窄工作流,但公开证据仍没有证明其车队能在跨国规模上安全、可靠、合规运行。[CR002, CR014, CR015, CR016, CR017, CR018]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 剩余暴露 | 未解决缺口 |
|---|---|---|---|---|---|
| VLA 泛化能力在结构化物流单元之外失效 | 高 | 关键 | 低至中 — Robot Era 明智地聚焦结构化工作流,但品类内领先实验室仍把具身泛化视为未完成课题 | 高 | 没有公开 Robot Era 基准资料包展示开放环境中的性能漂移、错误处理或安全恢复 |
| 车队在线率、维护或现场质量在快速交付爬坡中断裂 | 中高 | 关键 | 低 — 公开部署头条跑在公开可靠性披露前面 | 高 | Robot Era 部署没有公开 MTBF、备件、服务间隔或保修退货数据 |
| 无围栏或人机混合作业中,安全、跌倒或人机交互控制失效 | 中 | 关键 | 低至中 — 标准和最佳实践存在,但人形机器人专属合规仍不完整 | 高 | 没有留存的 Robot Era 认证、事故日志或经验证的安全边界披露 |
| 网络安全或 OTA 弱点让客户暴露在运营或数据风险下 | 中 | 高 | 低 — 留存来源未显示公开的 OTA、认证、SBOM 或漏洞披露姿态 | 高 | 本轮没有公开 Robot Era 网络控制文档 |
| 电池、灵巧操作和任务切换限制让狭窄工作流之外的 ROI 脆弱 | 高 | 高 | 中 — 结构化工作流和上半身 / 轮式变体可减轻压力,但品类障碍仍在 | 中高 | 没有公开 Robot Era 单任务成本、电池衰减或利用率数据集 |
各行把底层技术可行性风险,与车队规模部署所需的信任、质量和网络安全风险分开。
[CR002, CR014, CR015, CR016, CR018, CR019]对 Robot Era 最关键风险做剩余风险评分,重点标出高影响与公开缓释成熟度不足重叠的区域。
[CR003, CR006, CR008, CR014, CR019, CR026]7.3 竞争、商业化与资本强度风险
Robot Era 并不处在一个只靠执行就能决定结果的平静融资市场里。它参与的是一场资本竞赛,竞争对手的披露质量和规模已经在重置买家与投资人的预期。Robot Era 自身以超过 RMB 10 billion 的估值融资 RMB 1 billion,数周后又融资超过 USD 200 million,同时宣称千台级交付和 10 多个物流中心部署。这很亮眼,但同一组资料也显示基准已经变得多么苛刻。Unitree 正以约 RMB 50 billion 的估值冲刺 IPO,并称收入已超过 RMB 1 billion 且已经盈利;UBTECH 既有 8 亿元订单叙事,也有 HKEX 支撑的人形机器人收入披露;AgiBot 宣传已经推出 10,000 台机器人;Figure 仍凭 Helix 掌握十角兽资本和强公开 VLA 品牌。Humanoids Daily 把该行业描述为极其昂贵的资本竞赛,而 KrASIA 是有用的制衡:行业仍缺少清晰收入路径,一些被热捧的部署只是战略绑定,而非规模化销售。IEEE Spectrum 又从需求侧提醒,目前还没有人找到每个设施需要数千台人形机器人的用例。对 Robot Era 来说,这产生两个相连风险。第一,竞争可能压缩价格,或让客户预期移动得比 Robot Era 证明现场经济性更快。第二,公司可能不得不在自身收入质量、可重复性和利润率轮廓对外清晰之前继续融资。因此,合理的剩余判断不是“融资到位,风险解决”,而是“融资到位,烧钱和证明义务被推后”。[CR035, CR036, CR037, CR038, CR039, CR040]
| 风险 | 可监测触发器 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 出口管制和市场准入冲击 | 影响算力、传感器或中国制造机器人的政策变化 | 任何阻断 Robot Era 核心算力部件、关键传感器或目标海外采购渠道的新规则 | 在管理层拿出可信的技术和商业绕行方案前,暂停全球规模承保 |
| 技术就绪缺口 | 规模化期间现场 KPI 下滑 | 随着交付爬坡,公开或私下证据均无法证明在线率、错误恢复和安全运行指标在改善 | 从增长逻辑切换到受控试点逻辑,并因服务负担下调估值 |
| 商业化失败 | 积压订单 / 订单故事停止转化 | 旗舰物流部署没有转化为复购订单、已确认收入质量或客户多元化 | 把订单头条视为宣传而非经济证明,并收紧下行情景假设 |
| 资本竞赛劣势 | 同业市场保持火热,而 Robot Era 融资转弱 | 平轮或下轮、融资延迟,或相较 Unitree、UBTECH、Figure 和 AgiBot 等同业,投资者质量明显转弱 | 假设议价权和招聘杠杆正在侵蚀;重新评估生存期和价格压力假设 |
| 关键人物或合作伙伴冲击 | 领导层或锚定合作伙伴中断 | 创始人离任、主要客户暂停,或战略支持方退出且没有可见的梯队替代 | 升级到投资逻辑破裂复盘,因为技术可信度和商业化证明同时受损 |
止损标准按可观察的运营或政策事件来设定,让投资者能在全面减值变得明显前反应。
[CR003, CR006, CR008, CR009, CR012, CR026]7.4 关键人物依赖、伙伴集中与投资逻辑破裂触发点
Robot Era 的公开故事异常集中在一个创始人画像和一个商业切口上。Tsinghua 和 The Wire China 合起来显示,陈建宇不仅是公司创始人,还是 Tsinghua 助理教授,拥有 Berkeley 机器人学博士学位,并仍在运营一个学术机器人实验室。这对技术可信度、招聘和投资人信号是真正优势,但也带来典型关键人物暴露。保留的公开记录没有浮现出一组清晰点名、在产品、制造、安全或国际合规上同等量级的高管梯队,能够显然吸收创始人离开、分心或声誉事件的冲击。商业上,公司最强的证明集中在物流、China Post、SF Group、海关检查,以及覆盖汽车、科技、物流和半导体的广泛但偏战略投资人生态。这给 Robot Era 带来渠道和产业背书,但也意味着乐观情景的一部分依赖于一组相对集中的交易对手、支持方和用例叙事继续配合。如果这些关系减弱,或订单头条弱于收入转化,下行会很快传导到融资杠杆、估值和客户信心。因此,否决标准应当是运营性的,而不是意识形态性的:随着部署扩大仍无法发布安全和正常运行时间证据,无法把物流切口转化为可重复确认收入,在仍火热的同行市场里融资估值持平或下行,或创始人连续性丧失且没有可见高管梯队,都应被视为损害投资逻辑的事件,而不是短期噪音。[CR036, CR037, CR038, CR039, CR040, CR041]
| 依赖项 | 交易对手 | 角色 | 集中度 | 失效情景 | 严重性 | 缓释措施 | 剩余暴露 |
|---|---|---|---|---|---|---|---|
| 先进算力和 AI 工具链访问 | 美国关联芯片和工具制度 | 训练 / 推理硬件和模型开发能力 | 中高 | 出口限制或供应商替代削弱性能或推迟发布 | 关键 | 尽可能采用中国优先技术栈,储备关键部件,并优化模型效率 | 高 |
| 物流锚定客户和渠道 | China Post、SF Group、海关 / 物流项目 | 商业证明和部署切入口 | 高 | 如果某个旗舰项目暂停或收入转化不及预期,狭窄的物流集中度会削弱 | 高 | 在披露续约和扩张质量的同时,向制造业和服务业用例多元化 | 高 |
| 工业投资者生态 | 横跨汽车、科技、半导体、电信的战略投资者 | 资本、信号和客户引介层 | 中 | 在自我维持的经济性被证明前,投资者收缩或战略优先级转移会减少渠道支持 | 高 | 拓宽融资基础,并证明不依赖战略叙事也能复制客户 ROI | 中高 |
| 跨境合规和分销路径 | 中国以外的监管机构、进口商和企业买方 | 海外商业化 | 中 | 即便产品需求存在,监管摩擦也会拖慢或阻断扩张 | 高 | 按司法辖区分阶段扩张,并在规模承诺前取得合规文件 | 中高 |
| 创始人关联研究网络 | Tsinghua / 学术招聘和声誉 | 人才漏斗和技术可信度 | 中 | 创始人连续性丧失,会同时削弱招聘、融资和技术权威 | 高 | 在产品、制造和安全领导层建立可见的梯队 | 中高 |
本登记表强调那些会在显性法律事件出现前就冲击收入和融资的依赖项。
[CR001, CR003, CR006, CR007, CR008, CR009]| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 创始人 / 科学领导力 | 公开叙事、技术可信度和招聘集中在 Jianyu Chen 身上 | 中 | 关键 | 在创始人之下建立明确的继任安排和授权运营角色 | 索取组织架构、授权决策权、留任方案和董事会批准的继任计划 |
| 制造和现场服务领导力 | 快速交付说法让执行负担增长得比公开组织细节更快 | 中高 | 高 | 引入有经验的工厂、质量和服务运营负责人,并公开承担责任 | 索取制造、质量、支持和供应链具名负责人,以及 KPI 归属 |
| 安全 / 合规领导力 | 跨境部署需要产品安全和隐私运营人员,公开材料中看不到这些角色 | 中 | 高 | 在更大规模国际化前建立专门的安全与合规职能 | 索取负责认证、事故复盘、隐私和出口合规的具名责任高管 |
| 商业团队纵深 | 公开证据在关系和公告上很强,但对可复制收入治理披露很薄 | 中 | 高 | 围绕积压订单转化补强销售运营、财务和客户成功能力 | 索取提交董事会的 cohort 数据、续约负责人和客户集中度审查 |
公开资料对创始人质量异常清晰,对其余运营梯队异常稀薄,这正是本登记表具备重要性的原因。
[CR036, CR037, CR038, CR039, CR040, CR041]Robot Era 当前风险画像背后的关键外部依赖。
[CR001, CR003, CR006, CR008, CR012, CR014]7.5 展示材料
08估值
8.1 建议与价格纪律
Robot Era 确实有一个市场出清的私募估值标记,但公开记录支持的更多是战略期权价值,而不是收入支撑的公允性。2026 年 3 月资料相互印证了一轮 RMB 1 billion 战略融资,估值超过 RMB 10 billion,第三方报道把这个锚点折算到约 15 亿美元上下。在 Figure、Physical Intelligence 和其他人形机器人龙头都融到巨额资金的市场里,这个价格并不荒谬;Robot Era 异常广泛的产业投资人基础、累计订单 >RMB 500 million、10 多个物流中心部署,以及千台级交付表述,也提供了支撑。问题在分母。Robot Era 仍未披露经审计收入、待转化订单转收入情况、毛利率、现金或优先权条款。订单、出货和投资人 logo 都有意义,但它们并不等同于入账收入或清晰退出经济性。因此,价格敏感的结论是继续研究,置信度中、风险高、估值偏高,而不是干净的买入或回避。[CV001, CV002, CV003, CV004, CV005, CV006]
| 维度 | 评估 | 决策含义 |
|---|---|---|
| 建议 | 继续研究 | 继续跟踪 Robot Era,但不要只凭公开证据就把 2026 年 3 月的私募估值视为已充分承保。 |
| 信心 | 中 | 证据足以约束价格纪律,但不足以消除分母风险。 |
| 风险评级 | 高 | 商业牵引真实存在,但披露、流动性和商业化风险仍然重大。 |
| 估值立场 | 偏高 | 当前估值已经把大量收入转化提前计入,而公开证据尚未跟上。 |
| 入场纪律 | 要求收入证明或下行保护 | 相比盲目支付头条价格,结构化融资、明确优先权认知或尽调后的价格重置更容易自洽。 |
| 改变判断的因素 | 收入转化和现金披露 | 经审计收入、毛利率和 cap table 经济性,比再一个演示视频更重要。 |
本表明确对价格敏感:它判断的是当前 2026 年 3 月估值是否可投,而不是人形机器人这一品类的整体质量。
[CV001, CV002, CV003, CV004, CV005, CV029]| 论点 | 正方逻辑 | 什么会改变观点 |
|---|---|---|
| 工业赞助 | Robot Era 拥有中国人形机器人中较强的战略投资者阵容之一,应能帮助分销、制造和客户触达。 | 如果后续文件显示本轮带有异常沉重的优先权或战略约束,赞助方质量对公司的支持就没有头条显示得那么强。 |
| 商业牵引 | 累计订单超过 RMB 500 million、大型物流订单、海外占比和千台级交付表述,说明真实需求正在形成。 | 如果这些指标被证明以试点为主、可取消,或转化为确认收入很慢,当前估值就会失去核心支撑。 |
| 中国部署优势 | 中国供应链深度、政策支持和企业自动化需求,可以支撑相对普通机器人故事的溢价。 | 如果政策支持不能转化为经常性付费部署,“中国溢价”就会坍缩成又一个融资叙事。 |
| 全球可比公司折价 | Robot Era 估值低于 Figure、Physical Intelligence 等美国头部具身 AI 品牌,有助于相对定位。 | 如果收入披露仍明显弱于 Unitree 或 UBTECH,相对十角兽的折价并不够。 |
| 披露缺口 | 反方逻辑不是 Robot Era 没有需求,而是需求质量、利润率质量和投资者经济性仍披露不足。 | 经审计收入、毛利率和合同付款数据,可能把立场从偏高推向合理。 |
| 行业炒作风险 | 独立反向研究认为,整个品类仍活在结构化环境中,可能跑在真实商业化前面。 | 如果 Robot Era 发布在线率、转化和续约数据,足以推翻行业层面的怀疑,反向权重应下降。 |
反方逻辑主要关乎分母质量和流动性风险,而不是人形机器人在技术上是否有趣。
[CV002, CV003, CV004, CV005, CV006, CV007]战略融资和订单支撑了当前估值标记,但收入转化与股权结构证据缺失,结论仍只能止步于此。
[CV001, CV003, CV018, CV019, CV020, CV021]8.2 可比边界与中国相对全球的语境
可比公司同时支持溢价和折价。相较于陷入困境或成熟的上市机器人公司,Robot Era 应享有溢价,因为该品类仍按未来平台价值定价,而不是按当前 GAAP 盈利定价;中国的供应链深度和政策支持,也让本土部署故事比单看西方公开市场情绪更容易融资。但相较于最受追捧的美国具身 AI 融资,Robot Era 也应有折价。Figure 的私募估值从 2024 年的 $2.6 billion 升至 2025 年的 $39 billion;Physical Intelligence 在证明广泛商业部署之前,就凭基础模型热度达到 $2.4 billion 投后估值。Robot Era 的 >RMB 10 billion 锚点明显低于这些全球 AI 品牌峰值,但这并不自动说明它便宜。披露更充分的中国同行是更苛刻的参照。Unitree 在 IPO 材料中披露 2025 年收入增长至 17.08 亿元,并更新了 2026 年利润压力;UBTECH 披露 2025 年收入 RMB 2.001 billion、全尺寸人形机器人收入 RMB 820.6 million、毛利率 37.7%,而公开市场数据仍把其股权价值放在约 $7.22 billion。这些同行说明,中国公司可以获得真实溢价,但市场奖励的是已披露的收入转化,而不只是订单头条。[CV018, CV019, CV020, CV021, CV022, CV023]
| 可比对象 | 最新披露估值 / 状态 | 收入或订单信号 | 相关性 | 局限 |
|---|---|---|---|---|
| Robot Era(标的) | 2026 年 3 月战略轮估值超过 RMB 10B;Pandaily 称其超过 $1.37B | >RMB 500M 累计订单;订单不是经审计收入 | 直接入场价格,也是当前最强市场信号 | 没有公开经审计收入、毛利率、现金或优先权条款 |
| Figure AI | 2024 年 Series B 估值 $2.6B;2025 年融资据报估值 $39B | 商业部署叙事很强,但这里保留的公开来源没有给出经审计收入 | 最能映射顶端具身 AI 稀缺性溢价 | 美国软件品牌溢价和融资能力让它成为 Robot Era 极为宽松的可比对象 |
| Physical Intelligence | 2024 年融资投后估值 $2.4B | 保留的公开报道强调模型人才和投资人,而不是已验证收入 | 可用于映射基础模型可选性溢价 | 比 Robot Era 更偏软件优先,部署验证也更少 |
| Apptronik | 2025 年宣布 $350M Series A;保留来源未披露估值 | 试点交易和商业化推进已有讨论,但这里没有保留经审计收入 | 显示严肃的人形机器人公司在规模化前仍需要多少资本 | 融资金额不等于估值,不应过度解读 |
| Agility Robotics | 2024 年获 Schaeffler 少数股权战略投资;估值未披露 | 商业 RaaS 部署和 100 家工厂的客户意向,比披露估值更重要 | 衡量企业部署严肃度和商业化压力的好基准 | 保留来源中没有当前公开估值锚点 |
| Unitree | 2026 年上海 IPO 申报,拟募资 42 亿元 | 2025 年收入 17.08 亿元;2026 年增长下仍有利润压力;官方 G1 入门价从 US$13.5K 起,未含运费 | 中国披露增长、价格压力和 IPO 准备度的最佳基准 | 招股书动能本身并不说明长期盈利稳定 |
| UBTECH | 2026 年 6 月公开市值约 $7.22B | 2025 年收入 RMB 2.001B;全尺寸人形机器人收入 RMB 820.6M;毛利率 37.7% | 有披露支撑的中国人形机器人相邻公开可比公司中最佳 | 公开市值每日波动,且包含一条机器人产品线以外的更广业务组合 |
| Symbotic | 2026 年 6 月公开市值约 $25.66B | TTM 收入约 $2.51B;估算市销率约 10x | 对已规模化物流自动化赢家来说,是有用的公开自动化估值倍数上限 | 仓储系统比 Robot Era 成熟得多,也更有收入支撑 |
| iRobot | 2026 年 6 月公开市值约 $14.84M | 2024 年收入 $682M;隐含公开市值 / 2024 年销售额接近零 | 提醒下行情形:增长和利润率断裂后,公开机器人股权会怎样坍塌 | 消费清洁机器人不是直接人形机器人可比对象,应作为压力案例,而不是估值同业 |
| Fourier Intelligence | 2025 年 Series E 融资近 RMB 800M;这里保留的公开来源未披露估值 | 融资轮确认中国机器人公司需要大量资本,但这里没有保留经审计收入 | 可用于观察中国私募资本深度 | 在保留的公开证据中,它不是干净的估值或收入倍数可比对象 |
表内有意混合私募估值、融资轮、公开市值和已披露收入 / 订单信号。凡是估值计算需要依赖订单或未披露收入的行,均明确写出这一局限。
[CV001, CV003, CV011, CV012, CV013, CV014]这组 KPI 显示,公司融资真实、订单储备有动能,但披露质量还不足以干净支撑当前估值标记。
[CV001, CV002, CV003, CV006, CV018, CV019]8.3 情景框架与倍数纪律
约束估值的正确方式,是明确哪些是估计,哪些是披露。Robot Era 最干净的已披露商业数字,是累计订单超过 RMB 500 million。如果投资人粗糙地把它当成类收入分母,当前 >RMB 10 billion 估值意味着估计订单簿倍数超过 20x;这只是粗略校验,不是真正销售倍数,因为订单不是收入,时间区间也未标准化。公开可比公司提供第二个视角。CompaniesMarketCap 把 Symbotic 估在约 $25.66 billion,对应约 $2.51 billion 的过去 12 个月收入,约 10x 销售额;而 iRobot 2026 年 6 月市值仅约 $14.84 million,对比 2024 年 $682 million 收入,显示当增长和盈利能力断裂时,公开机器人股权会多么剧烈压缩。在这个光谱上,Robot Era 既不像公开市场便宜货,也不像全球私募市场荒谬高价;它更像一轮按成功收入转化定价的战略融资。这导向一个以略低于当前估值为中心的基准区间,只有当管理层证明订单会转化为确认收入、利润率和重复部署,而不是一次性试点时,才有上行空间。[CV003, CV004, CV029, CV030, CV031, CV032]
| 情景 | 概率信号 | 假设 | 估值逻辑 | 示意公允价值区间 |
|---|---|---|---|---|
| 牛市 | 25% | 订单转化强劲,海外占比守住,千台级交付被市场接受为真实部署,管理层披露的利润率更接近中国头部同行,而不是困境中的机器人公司。 | 投资人继续为具身 AI 可选性支付战略溢价,但仍低于 Figure 式软件品牌的极端水平。 | RMB 11B-RMB 15B |
| 基准 | 50% | 订单真实,战略投资人也重要,但到下一轮尽调周期,已确认收入披露、现金和优先权条款仍不完整。 | 当前估值大致只有在战略轮语境下才站得住,公开分母证据仍有限。 | RMB 8.5B-RMB 10.5B |
| 熊市 | 25% | 订单转化放慢,利润率低于预期,或融资条款比头条估值显示的更保护投资人,同时板块热度降温。 | 估值向公开可比公司的纪律回归,远离私募轮的稀缺性价值。 | RMB 6B-RMB 8.5B |
| 概率加权 | — | 基准情形证据占主导,因为最强的公开正面因素真实存在,但并不完整。 | 上述情景组中点的加权结果。 | ~RMB 9.5B-RMB 10B |
这些情景区间是判断区间,锚定已披露融资轮、订单代理指标、公开市场可比公司和负面商业化证据;它们不是 DCF 输出,也不应被误认为经审计的内在价值。
[CV001, CV003, CV004, CV018, CV025, CV029]粗略按订单代理口径计算,当前估值标记已默认相对上市自动化公司倍数有溢价,也远高于简单 5x 或 10x 在手订单视角。
除参考估值标记本身外,所有柱形都用已披露累计订单作为粗略代理分母。这里刻意标为订单簿估算,而不是经审计销售额测算。
[CV001, CV003, CV004, CV042, CV043, CV044]按公开证据推导的基准情景接近但略低于当前参考标记;如果没有新证据,上行空间有限。
这些是根据已披露轮次锚点、同业信号、上市可比公司倍数纪律和不利商业化证据推导的判断区间;不是贴现现金流输出。
[CV001, CV003, CV018, CV025, CV029, CV030]8.4 反向证据、尽调问题与投资逻辑破裂触发点
反向资料重要,因为这个行业仍非常早期。Bain 称部署仍大多在高度结构化环境里,抓取处理和电池续航仍卡住商业化。McKinsey 称,试点演示与商业可行规模之间的差距仍很大。IEEE Spectrum 更直白:资金充裕的人形机器人公司可以一边继续融资巨额资金,一边把未解决问题挥开。Robot Era 的公开记录符合这个提醒。公司的融资势头强于财务披露质量,融资锚点仍可能受到优先权悬挂、待转化订单滑坡,或订单转为确认收入速度慢于宣传的影响。因此,本章的投资逻辑破裂点是运营和财务性的,而不是意识形态性的。要从偏高转为合理或有吸引力,需要按收入流给出可审计收入、按产品给出毛利率、披露现金和月度烧钱速度、披露 >RMB 50 million 物流订单的合同付款条款,并证明海外订单和千台级交付正在转化为已验收、已付款部署。没有这些证据,当前估值仍只适合在继续尽调或有下行保护结构的前提下投资。[CV004, CV005, CV033, CV034, CV035, CV036]
| 触发项 | 阈值 | 对投资论点的传导 | 行动含义 |
|---|---|---|---|
| 收入转化跟不上故事 | 管理层无法拿出已确认收入,且该收入与披露的 >RMB 500M 订单基础有实质匹配。 | 当前估值不再是战略溢价,而像是在为积压订单头条买单过高。 | 将案例重置到熊市区间定价,或放弃。 |
| 现金和烧钱弱于预期 | 现金跑道较短,或烧钱速度比部署转化为付费安装的速度更快。 | 在板块情绪可能降温时,公司变得依赖融资。 | 要求更强条款、更多尽调,或更低入场价。 |
| 优先权堆栈扭曲头条估值 | 清算、棘轮或治理保护让表观估值在经济意义上误导投资人。 | 头条估值不再近似真实普通股价值。 | 暂停,直到融资文件完成审阅。 |
| 商业验证仍结构化且狭窄 | 部署仍卡在演示、海关检查或高度脚本化的物流单元里,没有更广泛的车队级验证。 | 板块负面研究会变成直接针对公司的估值逆风。 | 将公司从「research-more」推向「track only」或更低。 |
| 市场倍数压缩冲击机器人板块 | 公开自动化和机器人估值倍数压缩,私募资本也更不愿为可选性买单。 | Robot Era 失去当前支撑战略轮估值的稀缺性溢价。 | 承诺投资前,按更低公开倍数区间重新承销。 |
这些是估值叫停触发项,而非泛泛的业务风险;每一项都能直接推翻当前入场价格。
[CV004, CV029, CV030, CV031, CV032, CV033]| 主题 | 缺失证据 | 重要性 | 负责人 / 尽调路径 |
|---|---|---|---|
| 已确认收入桥接 | 2025-2026 年按产品、地域和客户类型拆分的季度收入,以及积压订单转化。 | 它能补上订单与真实销售之间最大的分母缺口。 | 财务团队、审计师底稿和董事会材料。 |
| 毛利率和服务组合 | 按产品族拆分硬件、软件、集成和支持毛利率。 | 估值取决于 Robot Era 规模化后像机器人平台,还是像重人力集成商。 | CFO 复核,加 SKU 级毛利桥接。 |
| 现金和跑道 | 2026 年 3 月融资后,账面现金、月度烧钱、资本开支需求和下行情形跑道。 | 2026 年 3 月融资后仍剩多少稀释风险,融资节奏是核心。 | 资金计划表和董事会批准的经营计划。 |
| 大订单合同经济性 | >RMB 50M 物流订单的付款里程碑、验收测试、质保和取消条款。 | 大合同可以看起来亮眼,但经济质量仍然很差。 | 客户合同、法律审阅和回款数据。 |
| 优先权和治理条款 | 最新融资轮的股权类别堆栈、清算优先权、反稀释和战略权利。 | 名义估值可能与有效入场经济性大幅背离。 | 律师审阅融资文件和附函。 |
| 部署质量指标 | 已安装基数的正常运行时间、故障率、复购数据和海外订单付费转化。 | 这是检验公司是否跑赢板块商业化缺口的最快办法。 | 运营仪表盘、客户访谈和服务日志。 |
要把判断从有边界的观点推进到按当前估值承销入场,这些是最低限度的尽调要求。
[CV003, CV005, CV006, CV007, CV008, CV009]免责声明
本尽调报告由 AI 研究代理基于截至 2026-06-10 的公开来源生成,不构成投资建议。Robot Era 是私营公司,已确认收入、利润率、现金跑道、客户集中度和详细轮次条款等重要承销输入,仍未在公开记录中披露;任何投资决定都应结合管理层材料、客户访谈和经审计财务数据验证。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Robot Era’s public website materials place the company at 7F, VIA Building, Tsinghua Science Park, Haidian District, Beijing. | 高 | SO001, SO002 |
| CO002 | The website bundle lists MKT@robotera.com and the public phone number 400 900 8798 for Robot Era. | 中 | SO002 |
| CO003 | KrASIA reported that Robot Era was founded in August 2023. | 中 | SO004 |
| CO004 | Jianyu Chen is identified in independent profiles as Robot Era’s founder. | 中 | SO004, SO005 |
| CO005 | Tsinghua IIIS says Jianyu Chen is an assistant professor at the institute and runs the ISR Lab. | 中 | SO003 |
| CO006 | Tsinghua IIIS says Jianyu Chen earned a BS from Tsinghua in 2015 and a PhD from UC Berkeley in 2020 under Masayoshi Tomizuka. | 中 | SO003 |
| CO007 | KrASIA said Chen described Robot Era as combining brain-and-body capabilities with Lego-like modular hardware and an ERA-42 VLA layer. | 中 | SO004 |
| CO008 | Official and independent materials support reading Robot Era as an embodied-intelligence platform company rather than a single-product hardware vendor. | 高 | SO001, SO004, SO012 |
| CO009 | The supplied public pack does not robustly surface a broader current executive roster or board structure beyond Jianyu Chen. | 中 | SO001, SO004, SO005 |
| CO010 | The Wire China profile says Jianyu Chen was 33 at the time of publication. | 中 | SO005 |
| CO011 | Robot Era’s strongest location signal combines an official Beijing operating address with Jianyu Chen’s Tsinghua-linked research background. | 高 | SO002, SO003, SO004 |
| CO012 | KrASIA reported that Robot Era raised nearly RMB500 million in a July 2025 Series A led by CDH VGC and Haier. | 中 | SO004 |
| CO013 | Official and independent sources say Robot Era’s November 2025 Series A+ was nearly RMB1 billion, with Geely leading and BAIC participating. | 高 | SO006, SO008, SO010 |
| CO014 | Yicai Global reported that Robot Era’s total disclosed capital had reached CNY1.8 billion by November 2025. | 中 | SO010 |
| CO015 | Company-backed and independent coverage support cumulative orders above RMB500 million by late 2025 to 2026. | 高 | SO010, SO016 |
| CO016 | Yicai Global said about half of Robot Era’s orders were overseas and named Renault, SF Holding, TCL, Lenovo, and Haier as partners. | 中 | SO010 |
| CO017 | Official and independent sources say a March 2026 strategic round of RMB1 billion pushed Robot Era’s valuation above RMB10 billion. | 高 | SO016, SO017 |
| CO018 | External unicorn trackers in March 2026 placed Robot Era around a $1.5 billion valuation after roughly $145 million of funding. | 中 | SO021, SO022 |
| CO019 | Official and multiple independent 2026 sources say Robot Era raised over $200 million in a new round led by SF Group with HSG and IDG Capital. | 高 | SO015, SO007, SO009, SO020 |
| CO020 | Official and independent 2026 sources say Robot Era had built more than 10 logistics centers with China Post and SF Group. | 高 | SO015, SO009 |
| CO021 | The 2026 funding-release cluster said Robot Era expected thousand-unit deliveries in Q2 2026 and growth above 300%. | 高 | SO015, SO009 |
| CO022 | Official and independent 2026 funding coverage said more than 95% of Robot Era’s components were developed in-house. | 高 | SO015, SO009 |
| CO023 | Official CES 2026 materials said Robot Era showcased a lineup including the L7 humanoid, the Q5 platform, and XHAND1. | 高 | SO012, SO018 |
| CO024 | Humanoid.Guide lists the L7 at 171 cm, 55 degrees of freedom, about 70 kg, and 20 kg dual-arm payload. | 中 | SO013 |
| CO025 | CES 2026 coverage described Q5 as a lighter companion model with 44 degrees of freedom. | 中 | SO012, SO018 |
| CO026 | Humanoid.Guide lists XHAND1 with 12 degrees of freedom, 80 newtons of grip force, and 25 kilograms of lift. | 中 | SO014 |
| CO027 | The March 2026 strategic-round release said Robot Era shipped more than 1,000 XHAND units in 2025, with 50% sold overseas. | 中 | SO016 |
| CO028 | The Wire China profile said Robot Era claimed 9 of the world’s 10 largest tech firms as customers. | 中 | SO005 |
| CO029 | KrASIA’s orders coverage portrayed Robot Era as moving to fulfill major orders rather than remaining at the demo stage. | 中 | SO011 |
| CO030 | PRNewswire said Robot Era and UNIDO signed a strategic cooperation agreement in Riyadh on November 24 focused on manufacturing, logistics, and commercial services. | 中 | SO026 |
| CO031 | The UNIDO agreement extends Robot Era’s story from domestic product launches toward global industrial-transformation positioning. | 中 | SO026, SO015 |
| CO032 | CKGSB said China’s humanoid-robot sector still lags U.S. leaders on autonomy and true commercial maturity. | 中 | SO023 |
| CO033 | TNW said buyer satisfaction was only 23% in one survey and that battery life and commercialization remained major gaps in humanoids. | 中 | SO024 |
| CO034 | The Robot Report said the crowded Chinese robotics market faced competitive pressure and a likely 2026 shakeout. | 中 | SO025 |
| CO035 | The rapid sequence of disclosed financing announcements from mid-2025 through mid-2026 shows strong investor appetite but leaves public round sequencing and dilution history hard to reconcile. | 高 | SO004, SO006, SO015, SO016 |
| CO036 | The combined address, founder, and academic evidence supports a Beijing-and-Tsinghua-rooted company identity for Robot Era. | 高 | SO002, SO003, SO004 |
| CO037 | Orders above RMB500 million, a 50% overseas mix, logistics-center deployments, and thousand-unit delivery claims together imply Robot Era had moved beyond pure lab demonstration by 2026. | 高 | SO009, SO010, SO015, SO016 |
| CO038 | The reviewed public pack still does not provide canonical revenue, gross margin, current headcount, or a full governance and cap-table disclosure for Robot Era. | 中 | SO001, SO004, SO015 |
| CM001 | The right market boundary for Robot Era is commercially deployed humanoid robots in enterprise workflows, not all robotics or all AI. | 高 | SM001, SM004 |
| CM002 | BCC's humanoid market scope covers commercial deployments across automotive, logistics and warehousing, government and public services, healthcare, and retail or commercial services, while excluding pilots, demos, and prototypes. | 中 | SM001 |
| CM003 | Published views of the current commercial humanoid market are still wide: BCC gives $1.9 billion in 2025 while CNBC's Barclays summary describes only a roughly $2 billion to $3 billion market today. | 高 | SM001, SM022 |
| CM004 | Barclays via CNBC presents a much larger upside case, arguing the humanoid market could reach $200 billion by 2035. | 中 | SM022 |
| CM005 | Robozaps says it tracks 26 humanoid robots across seven countries, with more than $5 billion of total industry investment and 14 commercially available or pre-order models as of March 2026. | 中 | SM003 |
| CM006 | China released its first national standard system for humanoid robotics and embodied AI in 2026, covering the industry chain and lifecycle. | 高 | SM006, SM019, SM023 |
| CM007 | China's 2026 standardization push is a response to fragmentation in data, component quality, dexterous hands, and evaluation methods. | 高 | SM002, SM006, SM019 |
| CM008 | Shanghai's embodied-AI plan targets 100 leading enterprises, 100 application scenarios, 100 competitive products, and more than RMB50 billion of output by 2027, with subsidy support for commercialization. | 中 | SM007 |
| CM009 | China Daily says the Development Research Center expects China's domestic embodied-AI market to reach RMB400 billion by 2030 and exceed RMB1 trillion by 2035. | 中 | SM018 |
| CM010 | China is the world's largest robot market: IFR says it represented 54% of global industrial-robot deployments in 2024, while CNBC says China installs around half of all industrial robots globally. | 高 | SM008, SM022 |
| CM011 | IFR says the global market value of industrial robot installations reached $16.7 billion and humanoid applications are now coming into focus in warehousing and manufacturing. | 中 | SM004 |
| CM012 | IFR says transportation and logistics led professional service robotics with 102,900 units sold in 2024, followed by hospitality with more than 42,000 and medical robots with around 16,700. | 中 | SM009 |
| CM013 | Warehouse-automation demand remains broadening into 2026 and more than 26% of warehouses are expected to be automated by 2027. | 中 | SM010, SM011 |
| CM014 | McKinsey says warehouse robot shipments could grow up to 50% annually through 2030, with warehouse automation growing more than 10% per year. | 中 | SM012 |
| CM015 | Near-term humanoid demand is strongest in structured logistics and manufacturing workflows rather than unstructured home use. | 高 | SM004, SM016, SM020, SM022 |
| CM016 | China Daily says early adoption is occurring in automotive and home-appliance manufacturing plus logistics and sorting, while safety and data remain bottlenecks. | 中 | SM020 |
| CM017 | Humanoid.guide says dexterous hands, safety-by-design, certification, and data quality are gating factors for scale beyond pilots. | 中 | SM002 |
| CM018 | RobotToday argues China currently leads physical deployment scale, shipment volume, and cost compression, while Western players still dominate more of the software and reasoning narrative. | 中 | SM017 |
| CM019 | Robotics & Automation News says 2026 is increasingly about whether humanoids can do real work in logistics, manufacturing, and services rather than perform impressive demos. | 中 | SM016 |
| CM020 | Robot Era's relevant market is concentrated in logistics, manufacturing, and commercial services rather than a generic all-humanoid category. | 高 | SM013, SM014, SM015, SM021 |
| CM021 | KrASIA reported that Robot Era had secured more than RMB500 million in commercial orders for 2025 and had named partners including Geely, Renault, SF Express, TCL, Haier, and Lenovo. | 中 | SM013 |
| CM022 | KrASIA said Robot Era's manufacturing use cases include component picking, high-precision assembly, and quality inspection, while commercial-service robots have entered retail stores via a Haier collaboration. | 中 | SM013 |
| CM023 | ROBOTERA and UNIDO say they plan embodied-intelligence deployments in manufacturing, logistics, and commercial services. | 中 | SM014 |
| CM024 | PRNASIA, Automated Warehouse, and Humanoids Daily all say Robot Era has deployments across more than 10 logistics centers with China Post and SF Group and started thousand-unit deliveries in Q2 2026. | 高 | SM015, SM021, SM025 |
| CM025 | Humanoids Daily says Robot Era reported 300% growth in 2026 as it shifted from pilots toward large-scale logistics delivery. | 中 | SM025 |
| CM026 | PRNASIA and Automated Warehouse say Robot Era claims more than 95% in-house core component development and emphasizes dexterous manipulation for logistics and industrial environments. | 高 | SM015, SM021 |
| CM027 | The best-supported TAM for Robot Era is the commercial humanoid market that BCC sizes at $1.9 billion in 2025 and $11 billion by 2030. | 中 | SM001 |
| CM028 | A broader option-value lens exists, but it should remain an upside case: CNBC's Barclays summary takes today's small market toward $200 billion by 2035. | 中 | SM003, SM022 |
| CM029 | A China-centered SAM lens is supported by policy and market substrate rather than one clean revenue total: China Daily cites a RMB400 billion 2030 embodied-AI market, IFR says China installs about half of industrial robots, and IFR service-robot data shows logistics automation already scales. | 高 | SM008, SM009, SM018 |
| CM030 | Robot Era's most defensible SAM is China-centered enterprise demand in logistics, manufacturing, and commercial services where structured tasks, policy support, and existing automation budgets intersect. | 中 | SM004, SM007, SM020, SM024 |
| CM031 | Robot Era's SOM cannot be calculated cleanly from public data because sources disclose orders, logistics centers, deliveries, and growth but not recognized revenue, renewals, or unit economics. | 高 | SM013, SM015, SM021, SM025 |
| CM032 | Warehouse-automation and service-robot evidence implies that humanoids are competing for existing automation budgets rather than creating an entirely new budget category. | 高 | SM009, SM010, SM011, SM012 |
| CM033 | Falling cost curves and aggressive production targets are a major adoption driver: BusinessWire says Unitree's $5,900 R1 shocked the market, and CNBC says Chinese robots are produced at roughly half Western cost. | 高 | SM022, SM024 |
| CM034 | Even bullish market sources preserve real bottlenecks around dexterous hands, data, safety, reliability, and proving ROI versus cheaper fixed-function automation. | 高 | SM002, SM004, SM019, SM020 |
| CM035 | Robot Era appears ahead of many peers on Chinese logistics commercialization, but the investment case still depends on whether those deployments turn into repeatable, margin-positive revenue rather than headline orders. | 高 | SM013, SM015, SM017, SM025 |
| CP001 | Automated Warehouse says Robot Era builds more than 95% of core components in-house and highlights its direct-drive dexterous hand architecture. | 中 | SP002 |
| CP002 | Automated Warehouse says Robot Era has deployments across more than 10 logistics centers with China Post and SF Group, began thousand-unit deliveries in Q2 2026, and reported growth above 300%. | 中 | SP002 |
| CP003 | China Daily says Robot Era's L7 stands 171 cm tall and that the company had shipped 200 humanoid robots in 2025 with more than 100 orders in progress. | 中 | SP004 |
| CP004 | Taken together, Robot Era's clearest public strengths are vertical integration, logistics-center traction, and the L7 plus XHAND1 product stack rather than public price transparency. | 中 | SP002, SP004 |
| CP005 | AgiBot's homepage shows a multi-product lineup spanning the A2, G1, X2, and X1 and explicitly markets commercial mass production. | 中 | SP005 |
| CP006 | AgiBot's 2026 conference release says the company launched five new platforms and expanded its AIMA stack across industrial, logistics, retail, security, and service use cases. | 中 | SP006 |
| CP007 | AgiBot's PR release says it rolled out its 10,000th robot in March 2026. | 中 | SP006 |
| CP008 | CnTechPost says AgiBot generated 1 billion yuan of revenue in 2025, targets more than 10 billion yuan in 2027, and passed 2,000 cumulative Expedition humanoid shipments with Q2 2026 capacity sold out. | 中 | SP007 |
| CP009 | AgiBot pressures Robot Era on scale, data ecosystem breadth, and visible commercialization cadence, even though public list pricing remains opaque. | 中 | SP005, SP006, SP007 |
| CP010 | Unitree's G1 page lists 23-43 degrees of freedom, about 35 kg weight, about two hours of battery life, and about 2 kg arm load for the base model. | 中 | SP008 |
| CP011 | Unitree's store page lists the G1 at US$13.5K and says secondary development requires the EDU edition. | 中 | SP009 |
| CP012 | CnTechPost's summary of Unitree's prospectus says 2025 revenue reached 1.71 billion yuan, humanoids exceeded 51% of main-business revenue, and average humanoid ASP fell to 167,600 yuan. | 中 | SP010 |
| CP013 | CNBC says Nvidia picked Unitree as its humanoid platform for researchers while the company pursued an IPO. | 中 | SP011 |
| CP014 | Unitree is the clearest public price leader in the reviewed set and likely the toughest Chinese rival for Robot Era to beat on entry price and shipment scale. | 高 | SP009, SP010, SP011 |
| CP015 | Fourier's GR-1 page markets it as a mass-produced humanoid with 165 cm height, 55 kg weight, 44 joints, and 230 N.m peak torque. | 中 | SP012 |
| CP016 | Fourier's GR-2 page says the newer model is 175 cm tall, 63 kg, has 53 joints, 12-DoF hands with tactile sensors, and peak torque above 380 N.m. | 中 | SP013 |
| CP017 | Fourier competes on dexterity, tactile hands, and developer-tooling support more than on publicly disclosed factory-scale deployments. | 中 | SP012, SP013 |
| CP018 | UBTECH's Walker S page positions the robot for multi-task industrial scenarios, with 1.7 m height, 41 servo joints, and factory-system integration. | 中 | SP014 |
| CP019 | UBTECH says Walker S2 entered mass production and delivery in several-hundred-unit batches, targeting 500 units within the year and 5,000 annual capacity in 2026. | 中 | SP015 |
| CP020 | UBTECH's 2025 HKEX results say humanoid products and services generated RMB820.6 million and became the company's largest revenue source within total revenue of RMB2.001 billion. | 中 | SP016 |
| CP021 | UBTECH has stronger public listing transparency and auto-factory commercialization proof than Robot Era. | 高 | SP014, SP015, SP016 |
| CP022 | The Robot Report framed Galbot's late-2025 financing as capital to scale mobile manipulator deployments. | 中 | SP018 |
| CP023 | CnTechPost says Galbot raised 2.5 billion yuan ($363 million) in early 2026 and reinforced its position as China's highest-valued unlisted humanoid robotics company. | 中 | SP019 |
| CP024 | Galbot's PR release says total funding reached $800 million at a $3 billion valuation and cites orders for thousands of units from CATL, Bosch, Toyota, and Hyundai. | 中 | SP020 |
| CP025 | Galbot pressures Robot Era on funding headline size and automotive-manufacturing logos, but public product detail is thinner in this evidence set than Unitree, Fourier, or UBTECH. | 中 | SP018, SP019, SP020 |
| CP026 | AgileX's official site centers on mobile bases, robot arms, ALOHA embodied-intelligence kits, and data-collection or education hardware rather than a full-size humanoid. | 中 | SP021 |
| CP027 | AgileX matters as an adjacent substitute for buyers who can automate with mobile manipulation or data-collection kits instead of a full humanoid. | 中 | SP021 |
| CP028 | Figure's Helix page says its generalist VLA runs perception, movement, and reasoning on board in real time and is already being positioned for logistics. | 中 | SP023 |
| CP029 | TechCrunch reported BMW would deploy Figure's humanoid robot at its South Carolina plant. | 中 | SP024 |
| CP030 | TechCrunch says Figure raised a Series C that exceeded $1 billion at a $39 billion valuation and nearly $2 billion total funding since 2022. | 中 | SP025 |
| CP031 | Figure's homepage and Figure 03 page position it as a home-capable general-purpose robot with 20 kg payload, 61 kg weight, 5-hour runtime, and Helix-enabled autonomy. | 中 | SP037, SP038 |
| CP032 | Figure is stronger than Robot Era on public software branding and fundraising scale, while Robot Era shows clearer China logistics-center PMF in this source set. | 高 | SP002, SP023, SP025, SP037, SP038 |
| CP033 | Physical Intelligence says it aims to build models that control any robot to do any task and lists support from Bond, Bezos, Khosla, OpenAI, Sequoia, CapitalG, and Thrive. | 中 | SP026 |
| CP034 | Physical Intelligence's pi0.5 article says most commercial robots still succeed in tightly controlled factories or warehouses, while its own goal is open-world generalization to new homes and objects. | 中 | SP027 |
| CP035 | Physical Intelligence is not a direct hardware OEM, but it raises the bar for autonomy, generalization, and software depth that humanoid hardware vendors may need to match. | 中 | SP026, SP027 |
| CP036 | 1X's NEO page markets a home robot with a $200 deposit, chore automation, and early-access autonomy rather than a near-term factory or logistics pitch. | 中 | SP029 |
| CP037 | TechCrunch says 1X's Neo Gamma remains far from commercial scaling and is still in limited in-home testing. | 中 | SP030 |
| CP038 | 1X competes more for long-run home-assistant mindshare than for Robot Era's near-term warehouse and manufacturing wedge. | 高 | SP029, SP030 |
| CP039 | Atlas' official product page says Boston Dynamics' robot has 50 kg instant payload, 30 kg sustained payload, 56 DoF, 4-hour battery life, and MES or WMS integration via Orbit. | 中 | SP031 |
| CP040 | Boston Dynamics said commercialization of electric Atlas will start with Hyundai as the first testing ground. | 中 | SP032 |
| CP041 | The Robot Report documented Atlas working at Hyundai's Georgia plant in early 2026. | 中 | SP033 |
| CP042 | Boston Dynamics has the strongest publicly disclosed payload and enterprise-integration stack in this evidence set, but not the clearest price or volume disclosure. | 高 | SP031, SP032, SP033 |
| CP043 | Agility's solutions page says Digit is the first humanoid to prove real value at scale, integrates with AMRs and management systems via Arc, and cites Amazon, GXO, and Schaeffler as customer references. | 中 | SP039 |
| CP044 | TechCrunch says Agility converted its GXO pilot into a formal paid deployment at a Georgia Spanx factory under a RaaS model. | 中 | SP035 |
| CP045 | Forbes says Toyota Canada signed a commercial RaaS agreement with Digit after pilot and that Digit also works with GXO, Schaeffler, and Amazon. | 中 | SP036 |
| CP046 | Agility's strongest pressure on Robot Era is repeatable logistics and automotive deployment under service contracts rather than splashy demo performance. | 高 | SP035, SP036, SP039 |
| CP047 | The relevant buyer choice set spans Chinese direct humanoid rivals, global industrial humanoids, autonomy-stack adjacencies, and cheaper mobile-manipulation substitutes. | 高 | SP021, SP026, SP039 |
| CP048 | Among the reviewed direct sources, Unitree is the clearest public list-price anchor, while most other rivals—including Robot Era—remain publicly price opaque or contract-based. | 中 | SP002, SP009, SP014, SP029, SP031, SP039 |
| CP049 | Robot Era's weaker publicly evidenced points versus rivals are price transparency, automotive-factory case studies, and public software-depth disclosure. | 高 | SP002, SP014, SP023, SP027, SP031, SP035, SP039 |
| CP050 | Weighted overall, Robot Era looks strongest in China logistics commercialization plus vertical integration, but Unitree on price, UBTECH and Agility or Boston or Figure on public deployment proof, and AgiBot or Galbot on scale all create real competitive pressure. | 高 | SP002, SP007, SP010, SP015, SP019, SP020, SP025, SP031, SP035, SP036 |
| CI001 | RobotEra said it closed a strategic funding round of RMB 1 billion and that the round pushed its valuation past RMB 10 billion. | 中 | SI001 |
| CI002 | RobotEra said cumulative orders had exceeded RMB 500 million. | 中 | SI001 |
| CI003 | RobotEra said its standalone XHAND product shipped more than 1,000 units in 2025, with half of those shipments going overseas markets. | 中 | SI001 |
| CI004 | RobotEra said some customers had placed repeat orders as many as six times. | 中 | SI001 |
| CI005 | RobotEra said a cross-border logistics inspection deployment with SF Express produced single orders exceeding RMB 50 million. | 中 | SI001 |
| CI006 | Official and independent coverage both reported a new RobotEra financing round of more than USD 200 million led by SF Group with HSG and IDG Capital participating. | 中 | SI002, SI003, SI008 |
| CI007 | Official and independent coverage said RobotEra began thousand-unit deliveries in Q2 2026 and paired that statement with a growth claim above 300 percent. | 中 | SI002, SI003, SI005 |
| CI008 | Independent coverage said RobotEra had deployments across more than 10 logistics centers through work with China Post and SF Group. | 中 | SI005 |
| CI009 | China Daily reported that RobotEra had shipped 200 humanoid robots in 2025 and had more than 100 orders in progress. | 中 | SI007 |
| CI010 | Yicai reported that RobotEra raised almost CNY 1 billion in a November 2025 A+ round led by Geely Capital. | 中 | SI004 |
| CI011 | Yicai reported that by November 2025 RobotEra had completed three rounds in just over a year totaling CNY 1.8 billion. | 中 | SI004 |
| CI012 | Yicai reported that RobotEra said orders had exceeded CNY 500 million and that overseas business accounted for half of that figure. | 中 | SI004 |
| CI013 | KrASIA reported that Robot Era had secured more than RMB 500 million in commercial orders for 2025 and had entered North America, Europe, Japan, and South Korea. | 中 | SI006 |
| CI014 | KrASIA reported that overseas revenue accounted for about half of Robot Era’s total revenue, but the source did not disclose an absolute revenue figure. | 低 | SI006 |
| CI015 | Across retained Robot Era sources, management and reporters disclose orders, shipments, deployments, and funding, but none of those retained sources disclose recognized revenue or revenue by stream. | 中 | SI001, SI002, SI004, SI005, SI006, SI007 |
| CI016 | Across retained Robot Era sources, no public cash balance, gross margin, debt balance, or monthly burn figure is disclosed. | 中 | SI001, SI002, SI004, SI006 |
| CI017 | No retained Robot Era source publishes a public list price for the L7, M7, or XHAND in enterprise deployments. | 中 | SI001, SI002, SI004, SI005, SI006, SI007 |
| CI018 | Unitree’s official G1 product page lists a base price of USD 13,500 before shipping and notes that the basic product does not support secondary development. | 中 | SI009 |
| CI019 | BotInfo said the programmable Unitree G1 EDU range starts at USD 43,900 and reaches USD 73,900, while direct-to-China basic pricing can be as low as USD 13,500. | 中 | SI010 |
| CI020 | CnTechPost reported that Unitree cut humanoid average selling price to 167,600 yuan in the first three quarters of 2025 to accelerate commercialization and market share. | 中 | SI011 |
| CI021 | SCIO said commercial humanoids often still sell for hundreds of thousands of yuan and cited around 300,000 yuan as a price point enterprises may view as acceptable for ROI. | 中 | SI022 |
| CI022 | OFweek said investors are increasingly evaluating humanoid companies on products, orders, cost, and reliability, and that even a 200,000 to 300,000 yuan machine remains a meaningful enterprise expense. | 中 | SI023 |
| CI023 | Interact Analysis projected global humanoid robot shipments of only just over 40,000 units and about USD 2 billion of revenue by 2032 despite large addressable-market rhetoric. | 中 | SI021 |
| CI024 | Bain said early humanoid deployments are still mostly limited to highly structured environments and many demonstrations remain dependent on human supervision. | 中 | SI019 |
| CI025 | McKinsey said humanoids must cross from pilot purgatory to repeatable ROI in semi-structured workflows before broad commercial reality emerges. | 中 | SI020 |
| CI026 | Public price anchors from Unitree show that the gap between demo hardware pricing and more capable developer or industrial configurations is wide, so Robot Era backlog cannot be translated into realized ASP from public data alone. | 中 | SI009, SI010, SI011 |
| CI027 | Using Unitree’s 167,600 yuan humanoid ASP only as a rough external anchor, Robot Era’s disclosed order book above RMB 500 million would imply roughly 2,980 robot-equivalent units, but the proxy is weak because Robot Era likely sells different hardware and solution bundles. | 低 | SI001, SI006, SI011 |
| CI028 | Using the same rough Unitree ASP anchor, a single RobotEra logistics order above RMB 50 million would imply roughly 300 G1-equivalent units, but the actual unit count could differ materially because the contract likely bundles integration and software. | 低 | SI001, SI006, SI011 |
| CI029 | The March 2026 RMB 1 billion strategic round and the later 2026 financing of more than USD 200 million show unusually strong near-term capital access for a private humanoid startup. | 中 | SI001, SI002, SI003 |
| CI030 | The locally supported 2025 to 2026 financing chronology shows at least CNY 1.8 billion raised by November 2025, another RMB 1 billion strategic round in March 2026, and then a further round of more than USD 200 million. | 中 | SI001, SI002, SI004 |
| CI031 | Robot Era’s investor list spans strategic names such as Geely, BAIC, SF, Lenovo, Haier, Renault, ICBC Capital, and China Unicom affiliates, which can reduce customer-acquisition friction but also increase concentration risk around a few industrial ecosystems. | 中 | SI001, SI004, SI006 |
| CI032 | Rapid repeat fundraising within months implies Robot Era is financing ahead of a cash-intensive scale-up phase rather than already funding growth from disclosed operating cash flow. | 中 | SI001, SI002, SI003 |
| CI033 | Serve Robotics’ 2025 SEC exhibit showed only USD 687 thousand of third-quarter revenue alongside USD 210 million of liquidity, USD 50.6 million of nine-month operating cash outflow, and USD 122.1 million of investing cash outflow, illustrating how robot scale-up can consume capital before large revenue arrives. | 中 | SI016 |
| CI034 | Symbotic’s fiscal 2025 official results showed USD 2.247 billion of revenue but still highlighted liquidity, gross margin expansion, and free cash flow as key management priorities. | 中 | SI015 |
| CI035 | iRobot’s 2024 10-K said revenue fell to USD 681.8 million and that management had largely completed an operational restructuring to align the cost structure with near-term revenue expectations. | 中 | SI014 |
| CI036 | UBTECH’s 2025 annual results filing said full-size humanoid products and services reached RMB 820.6 million of revenue in 2025 and that company gross margin rose to 37.7 percent. | 中 | SI013 |
| CI037 | UBTECH’s official PR and related filings said Walker series orders surpassed RMB 800 million and the company targeted roughly 500 unit deliveries during 2025, providing a disclosed industrial benchmark for humanoid commercialization scale. | 中 | SI012, SI013 |
| CI038 | Compared with Unitree and UBTECH, Robot Era reveals funding and orders but does not publish revenue, gross margin, or balance-sheet detail in retained sources. | 中 | SI001, SI002, SI004, SI011, SI013 |
| CI039 | Robot Era’s 2026 gross fundraising alone appears larger than Serve Robotics’ September 2025 liquidity of USD 210 million, so near-term solvency is less obviously the problem than capital efficiency and disclosure quality. | 低 | SI002, SI003, SI016 |
| CI040 | The strongest public financial verdict is that Robot Era has real demand and financing momentum, but investors still cannot underwrite revenue quality, margin path, or runway because orders and shipments are not recognized revenue and cash metrics remain undisclosed. | 中 | SI001, SI002, SI004, SI005, SI006, SI013 |
| CI041 | A3’s Q1 2026 market data showed 9,055 robot orders worth USD 543 million in North America, with non-automotive demand broadening even as order revenue softened. | 中 | SI017 |
| CI042 | ResearchAndMarkets coverage said China’s robotics sector logged 50 billion yuan of financing in the first nine months of 2025 and highlighted shock low-price moves such as Unitree’s USD 5,900 R1, signaling abundant capital and price compression around adjacent humanoid categories. | 中 | SI018 |
| CI043 | Public sources do not disclose working-capital terms, warranty reserves, receivable dynamics, or contract billing mechanics for Robot Era’s logistics deployments. | 中 | SI001, SI002, SI006, SI016 |
| CI044 | Any explicit burn or runway number for Robot Era would therefore be a low-confidence scenario rather than a reported company metric. | 中 | SI001, SI002, SI004, SI006, SI016 |
| CE001 | Retained current direct sources support a Robot Era product family centered on L7, Q5, M7, XHAND1, and XHAND1 Lite, while not providing a supportable current XBot-L surface for this run. | 中 | SE001, SE002 |
| CE002 | The official website bundle presents L7 and M7 as shared-architecture full-body versus upper-body configurations and explicitly describes zero-cost migration between the two forms for different deployment scenarios. | 中 | SE001 |
| CE003 | The official website bundle gives L7 a public spec floor of 55 degrees of freedom, 171 centimeters of height, and 20 kilograms of dual-arm payload. | 中 | SE001 |
| CE004 | The official website bundle says L7 uses 80 TOPS of x86 compute plus 275 TOPS on Orin and lists a 2.1 meter arm span. | 中 | SE001 |
| CE005 | The official website bundle describes L7 with a 3-DoF waist, 12-DoF hands, and 12-DoF cross-axis wrists as part of its manipulation stack. | 中 | SE001 |
| CE006 | The CES 2026 official release says L7 stands 171 centimeters tall, weighs 70 kilograms, has 55 degrees of freedom, and supports both full bipedal and upper-body configurations. | 中 | SE002 |
| CE007 | Independent technical databases describe L7 as roughly 63 to 65 kilograms without hands or depending on configuration, creating a public weight mismatch versus the 70 kilogram official CES figure. | 中 | SE010, SE013, SE015 |
| CE008 | China Daily and independent technical coverage both describe L7 as combining high-dynamic demo behavior with practical sorting, scanning, screw-driving, and other logistics or factory tasks. | 中 | SE007, SE010, SE013 |
| CE009 | Official and third-party sources consistently present Q5 as a 44-DoF wheeled humanoid platform. | 中 | SE001, SE002, SE011 |
| CE010 | The retained Q5 sources place the robot at about 165 centimeters and around four hours of operation per charge or 240 minutes of max operation time. | 中 | SE001, SE011, SE016 |
| CE011 | Independent Q5 sources describe 7-axis arms, XHAND Lite or Fairy Hands with 11 degrees of freedom, and about 10 kilograms of single-hand payload. | 中 | SE011, SE016, SE017 |
| CE012 | The official website bundle gives M7 43 degrees of freedom including hands, a 1660 millimeter arm span, and a 2.1 meter operating-range diameter. | 中 | SE001 |
| CE013 | The official website bundle positions M7 for logistics, manufacturing, commercial-service, and research scenarios rather than as a pure demo platform. | 中 | SE001 |
| CE014 | The official website bundle describes XHAND1 as a fully direct-drive, five-finger dexterous hand with 12 active degrees of freedom. | 中 | SE001 |
| CE015 | The official website bundle says each XHAND1 fingertip uses a 270-degree tactile array with more than 100 sensing points and 0.05 N precision. | 中 | SE001 |
| CE016 | The official website bundle says XHAND1 finger motors exceed 1 N·m and that the whole hand can lift more than 25 kilograms, while also warning that the lift figure comes from internal lab tests and can vary by environment. | 中 | SE001 |
| CE017 | The official website bundle lists XHAND1 at 1100 grams, 191 by 94 by 47 millimeters, 83 Hz whole-hand control, 24 to 72 volt operation, and EtherCAT or RS485 interfaces. | 中 | SE001 |
| CE018 | The CES 2026 official release says XHAND1 offers 12 active degrees of freedom, 80 N of maximum single-hand grip force, and the ability to lift objects weighing up to 25 kilograms. | 中 | SE002 |
| CE019 | The official website bundle says XHAND1 software support includes Ubuntu, SDKs, C++ and Python, ROS 1 and ROS 2, x86 and ARM compatibility, and multiple teleoperation schemes including VR and motion-capture gloves. | 中 | SE001 |
| CE020 | The official solution text says ERA-42 is an end-to-end VLA embodied model covering the logistics workflow from order release and outbound through picking, scanning, and boxing. | 中 | SE001 |
| CE021 | The official solution text says business systems issue tasks to the model and receive real-time progress feedback, creating a bi-directional closed loop. | 中 | SE001 |
| CE022 | The official solution text says the workflow stack exposes ten layers of standardized interfaces and claims stronger generality without separate custom training. | 中 | SE001 |
| CE023 | The official solution and Q5 surfaces describe a software-hardware loop built on open-source plus real-machine pretraining, high-frequency inference, digital-twin monitoring, automated data cleaning, and closed-loop validation. | 中 | SE001, SE011 |
| CE024 | Robotics & Automation News says Robot Era frames ERA-42 as integrating vision, language, touch, and body posture inside one model. | 中 | SE005 |
| CE025 | Robotics & Automation News says Robot Era claimed ERA-42 paired with XHAND1 can perform more than 100 intricate tasks and learn new tasks within two hours using minimal data. | 中 | SE005 |
| CE026 | Humanoids Daily says Robot Era is targeting the warehouse flexible-picking gap where AGVs and rigid automation still leave item-level picking to humans. | 中 | SE006 |
| CE027 | Humanoids Daily says L7's 3-DoF waist gives it a 2.1 meter workspace and supports target picking, barcode scanning, order-bin placement, and recycle-bin fallback when scans fail. | 中 | SE006 |
| CE028 | KrASIA says the standardized logistics handling and sorting solution is built on L7, powered by ERA-42, and integrated into logistics clients' business systems. | 中 | SE008 |
| CE029 | KrASIA says Robot Era has deployed pilot installations of its standardized handling and sorting solution with multiple global logistics companies. | 中 | SE008 |
| CE030 | KrASIA says ERA-42 was publicly described as Robot Era's first system combining a world model with a fast-slow hierarchical architecture. | 中 | SE008 |
| CE031 | KrASIA says Robot Era and Stanford's Chelsea Finn proposed Ctrl-World and claimed a 44.7 percent improvement in task completion rates in unfamiliar environments. | 中 | SE008 |
| CE032 | Retained direct sources market ERA-42 as an end-to-end VLA, world-model, and fast-slow system, but do not disclose parameter count, benchmark suite, inference latency, or the safety envelope around production deployment. | 中 | SE001, SE005, SE008 |
| CE033 | Robot Era's public GitHub organization shows a non-trivial public code surface, including Humanoid-Gym, Video Prediction Policy, and text indicating a Robotera VLA release tied to M7 workflows. | 中 | SE018 |
| CE034 | The official ros2_sdk repository provides public ROS 2 packages, custom interfaces, service and action definitions, trajectory execution, and an RL controller example. | 中 | SE019 |
| CE035 | The official xbot_sdk_api repository provides a ROS2-based Python SDK with trajectory control, MPC, XHand and XHand Lite control, base-drive commands, and model selection for Q5, L3, and L7. | 中 | SE020 |
| CE036 | The xbot_sdk_api examples and environment notes point to a real developer environment around ros2-humble, CycloneDDS, teleop_client, and hand or base control APIs rather than a demo-only showcase. | 中 | SE020 |
| CE037 | Official download-center text says Robot Era has product one-pagers, manuals, and a document center, but those assets are form-gated rather than openly published. | 中 | SE001 |
| CE038 | PRNASIA and KrASIA both report that more than 95 percent of Robot Era's components, including joints, dexterous hands, and motors, are self-developed. | 中 | SE003, SE008 |
| CE039 | KrASIA says Robot Era follows a modular, Lego-like hardware approach and specifically names XHAND1, XHAND1 Lite, Q5, L7, and M7 as released modules or platforms. | 中 | SE008 |
| CE040 | The CES 2026 official release says Robot Era has cumulatively delivered more than 600 units across exhibition, retail, and logistics environments. | 中 | SE002 |
| CE041 | PRNASIA, Automated Warehouse, and CnTechPost all say thousand-unit deliveries began in Q2 2026 and pair that with a sales-growth claim above 300 percent. | 中 | SE004, SE025, SE026 |
| CE042 | China Daily says Robot Era had shipped 200 humanoid robots in 2025 and had more than 100 orders in progress. | 中 | SE007 |
| CE043 | The CES 2026 official release says XHAND1 has already been adopted by Skild AI, Humanoid AI, Extend Robotics, UC Berkeley, MIT, and ByteDance Seed. | 中 | SE002 |
| CE044 | KrASIA says Robot Era's developer ecosystem includes ByteDance's robotics lab, Skild AI, and other institutions doing robotics R&D and technology transfer. | 中 | SE008 |
| CE045 | The official website bundle exposes a real privacy and procurement consent surface for downloads and purchase inquiries, but retained public product sources do not disclose robot safety certifications, warranty terms, or software-security architecture. | 中 | SE001 |
| CE046 | Aparobot and ChipSilicon both show formal safety rating and ingress protection fields as blank or undisclosed for L7. | 中 | SE010, SE013 |
| CE047 | Bain says early humanoid deployments are mostly limited to highly structured environments and that handling and battery life remain gating factors. | 中 | SE021 |
| CE048 | McKinsey says the first commercial value for humanoids will come from semi-structured use cases such as tote picking, palletizing, and line feeding rather than open-world general labor. | 中 | SE022 |
| CE049 | Interact Analysis says adoption will remain low in the short to medium term because of cost, dexterity, safety, and form-factor barriers. | 中 | SE023 |
| CE050 | The SCIO industrial-reality article says most humanoids still operate for only two to three hours per charge and that battery bulk, thermal management, dexterity, and cost remain major bottlenecks. | 中 | SE024 |
| CE051 | The SCIO article says enterprise ROI may require prices around 300,000 yuan and explicitly identifies manufacturing and logistics as the sensible first commercialization wedges. | 中 | SE024 |
| CE052 | KrASIA's adverse sector analysis says commercialization remains murky and that some automotive-factory deployments may be strategic tie-ups rather than repeatable sales. | 中 | SE009 |
| CE053 | The official website bundle's join surface includes campus recruitment and employee-story roles for algorithm, software, and mechanical engineers, implying continuing hiring across core technical functions. | 中 | SE001 |
| CE054 | Public adoption evidence is strongest for structured external users of the hand and the developer ecosystem, not for audited long-duration field fleets with public uptime statistics. | 中 | SE002, SE008, SE020 |
| CE055 | Across retained sources, Robot Era's current commercial relevance is concentrated in structured logistics, manufacturing, indoor service, exhibition, and research workflows rather than open-world general-purpose autonomy. | 中 | SE006, SE008, SE021, SE022, SE024 |
| CU001 | Robotera says cumulative orders have exceeded RMB 500 million and some customers have placed repeat orders up to six times. | 高 | SU001, SU022 |
| CU002 | Robotera says 50% of cumulative orders come from international clients. | 高 | SU001, SU022 |
| CU003 | Robotera says more than 1,000 XHand units shipped in 2025 and about half went to overseas markets. | 高 | SU001, SU022 |
| CU004 | Robotera says its SF Express cross-border logistics inspection solution has been officially deployed at customs and that single orders there exceed RMB 50 million. | 高 | SU001, SU022 |
| CU005 | Robotera publicly names Samsung, Geely, Renault, Lenovo, Haier, TCL, and Century Golden Resources as industrial clients. | 高 | SU001, SU022 |
| CU006 | Robotera says it is deployed across more than 10 logistics centers in collaboration with China Post and SF Group. | 高 | SU002, SU023 |
| CU007 | Robotera says it started thousand-unit deliveries in Q2 2026 and that growth exceeded 300%. | 高 | SU002, SU023 |
| CU008 | Robotera frames SF Group, Geely Capital, Lenovo, Haier, and other industrial investors as sources of real-world deployment scenarios and commercial demand. | 高 | SU002, SU023 |
| CU009 | KrASIA reports that Robot Era has more than RMB 500 million of 2025 commercial orders and partnerships with Geely, Renault, SF Express, TCL, Haier, and Lenovo. | 中 | SU003 |
| CU010 | KrASIA and Humanoids Daily both say Robotera’s largest disclosed logistics order is close to RMB 50 million. | 中 | SU003, SU005 |
| CU011 | KrASIA and Yicai describe Robotera use cases across logistics sorting and scanning, manufacturing picking and quality inspection, and commercial-service cleaning or guided tours. | 中 | SU003, SU004 |
| CU012 | KrASIA says Robot Era products have entered North America, Europe, Japan, and South Korea, with overseas revenue around half of total revenue. | 中 | SU003 |
| CU013 | Yicai says Robotera’s overseas business accounts for half of the CNY500 million-plus order figure and that the company has presence in North America, Europe, the Middle East, Japan, and South Korea. | 高 | SU004, SU003 |
| CU014 | Yicai names Geely, Haier, Renault, SF Holding, TCL, and Lenovo as industrial partners. | 高 | SU004, SU003 |
| CU015 | Humanoids Daily says Robotera had about a US$70 million 2025 order book and clients including Geely, Renault, SF Express, TCL, and Haier. | 中 | SU005 |
| CU016 | Humanoids Daily says overseas business is roughly 50% of total revenue and links that mix to a domestic-solutions-plus-overseas-developer strategy. | 中 | SU005 |
| CU017 | Humanoids Daily and Mike Kalil both say Robotera robots are working in more than 10 logistics centers via China Post and SF Group or SF Express, with 24/7 operation and up to 85% of human-level efficiency. | 中 | SU006, SU012 |
| CU018 | Humanoids Daily, RoboticsTomorrow, and Automated Warehouse all report that Robotera began thousand-unit deliveries in Q2 2026. | 中 | SU006, SU008, SU013 |
| CU019 | Humanoids Daily says Geely and BAIC provide structured automotive environments for testing, while Samsung and TCL collaborations target precision manufacturing and cleanroom-compatible automation. | 中 | SU006 |
| CU020 | Humanoids Daily’s CeMAT coverage says Robotera positioned the L7 as a solution for the warehouse flexible-picking gap rather than simple AGV transport. | 中 | SU007 |
| CU021 | Humanoids Daily’s CeMAT coverage says the L7 uses a 12-DoF dexterous hand, about 2.1 meters of vertical workspace, and direct WMS integration for barcode scanning and exception handling. | 中 | SU007 |
| CU022 | Automated Warehouse corroborates Robotera’s claim of more than 10 logistics centers, Q2 2026 thousand-unit deliveries, and long-term deployment validation. | 中 | SU013 |
| CU023 | EqualOcean says Robotera delivered more than 200 units in 2025, had more than 100 orders in progress, and received over half its orders from overseas clients. | 中 | SU014, SU020 |
| CU024 | China Daily says the L7 performs sorting, scanning, and screw-driving in logistics and factory settings. | 高 | SU020, SU004 |
| CU025 | China Daily says Robotera shipped 200 humanoid robots in 2025, had over 100 orders in progress, and counted nine of the world's top 10 most valuable tech companies as customers. | 高 | SU020, SU003 |
| CU026 | KrASIA says ByteDance’s robotics lab and Skild AI use Robotera robots for robotics R&D and technology transfer. | 中 | SU003 |
| CU027 | ByteDance’s official Seed Robotics page confirms ByteDance runs a general-purpose robotics team, so ByteDance should be treated as a developer-lab adopter rather than evidence of industrial revenue durability. | 中 | SU025 |
| CU028 | Mike Kalil says Robotera humanoids in logistics hubs perform the same conveyor-sorting and lane-placement jobs normally done by human workers. | 低 | SU012 |
| CU029 | Mike Kalil says Robotera appears to have found footing in logistics but still trails Unitree, UBTECH, AgiBot, and NOETIX on shipped-humanoid volume. | 低 | SU011 |
| CU030 | Across official and independent sources, SF or SF Group logistics is Robotera's clearest disclosed pilot-to-scale customer path because it combines named customer proof, deployed customs or sorting workflows, and large disclosed order or delivery figures. | 中 | SU001, SU002, SU006, SU013 |
| CU031 | Haier evidence is more specific than most other non-logistics logos because public sources say co-developed robots have already entered retail stores. | 中 | SU003, SU005 |
| CU032 | Geely evidence is credible but still thin on plant count, contract size, or throughput because public sources name the partner and manufacturing scenarios without disclosing site-level KPIs. | 中 | SU003, SU004, SU006 |
| CU033 | Renault, TCL, and Lenovo are publicly named industrial partners, but public sources in this pack do not disclose deployment counts, contract values, or site-specific outcomes for them. | 中 | SU001, SU004, SU006 |
| CU034 | China Post is named in Robotera and independent coverage, but this chapter does not have a China Post statement that names Robotera directly. | 中 | SU002, SU006, SU013, SU019 |
| CU035 | No public source in this pack discloses customer count, NRR, GRR, churn, contract length, or top-customer share. | 高 | SU015, SU016, SU017, SU018 |
| CU036 | IEEE Spectrum, McKinsey, Robotics & Automation News, and Forbes all argue that humanoid deployments can remain supervised, early-stage, or economically unproven even when partner logos and pilots are public. | 高 | SU015, SU016, SU017, SU018 |
| CU037 | Adverse sector commentary says structured logistics and factory tasks are the most realistic early commercialization path, but cost, battery life, reliability, and safety still limit broader scale. | 高 | SU015, SU016, SU021 |
| CU038 | Robotera's public pack does not separate audited production revenue from developer-lab, research, or pilot deployments, so headline partner lists may overstate durable revenue quality. | 中 | SU003, SU005, SU015, SU017, SU018 |
| CU039 | China Post’s own article shows the postal group is actively building unmanned logistics infrastructure, which makes it a plausible automation buyer environment but does not on its own prove Robotera deployment. | 中 | SU019 |
| CU040 | Aparobot’s secondary database also lists Lenovo, Geely, Haier, SF Express, and Renault among notable industrial partners, but that is lower-confidence corroboration rather than primary proof. | 低 | SU024 |
| CR001 | Robot Era’s L7 brochure describes an x86-plus-Orin compute stack on the robot platform. | 中 | SR003 |
| CR002 | The same L7 brochure lists binocular camera, 3D LiDAR, and microphone-array hardware on the platform. | 中 | SR003 |
| CR003 | BIS said its March 2025 Entity List action was intended to restrict the CCP’s ability to acquire and develop high-performance and exascale computing capabilities and quantum technologies for military applications. | 中 | SR014 |
| CR004 | BIS said 12 China/Taiwan entities were added for advanced AI, supercomputers, and high-performance AI chips, and 27 Chinese entities were added for support of China’s military modernization. | 中 | SR014 |
| CR005 | CRS says the U.S. government has strengthened export controls on advanced semiconductors to China since 2018. | 中 | SR015 |
| CR006 | Sidley says BIS’s January 2025 rules added controls on advanced computing items and, for the first time, on certain AI model weights. | 中 | SR019 |
| CR007 | Sidley says exports of covered advanced computing items or model weights outside close U.S. allies generally require licensing and face presumptive denial for frontier model weights. | 中 | SR019 |
| CR008 | FDD says a March 2026 Senate bill would prevent federal funds from purchasing certain Chinese-made unmanned ground vehicle systems, including humanoid robots. | 中 | SR016 |
| CR009 | FDD also argues Commerce should consider targeted controls on advanced sensors used by Chinese humanoid robot firms. | 中 | SR016 |
| CR010 | IEEE Spectrum says the proposed Chinese-robot ban could reshape supply chains because American robot makers still need Chinese-made components. | 中 | SR017 |
| CR011 | Rest of World says Chinese companies make up 63% of the global humanoid robot supply chain. | 中 | SR018 |
| CR012 | Rest of World says widened chip controls could deprive Chinese firms of cutting-edge chips while tariffs also make it harder for American companies to scale production quickly. | 中 | SR018 |
| CR013 | CSIS says export restrictions alone cannot substitute for broader industrial and research policy and may intensify China’s self-reliance push. | 中 | SR020 |
| CR014 | OSHA says there are currently no specific OSHA standards for the robotics industry. | 中 | SR021 |
| CR015 | OSHA says ISO 10218 and ANSI/RIA R15.06 are baseline references for industrial robots, but ISO 10218 does not apply to non-industrial robots. | 中 | SR021 |
| CR016 | OSHA’s technical manual says industrial robot systems include control systems, sensors, and communication interfaces and are used in industrial applications with growing AI-enabled capabilities. | 中 | SR022 |
| CR017 | The EU AI Act states that common rules for high-risk AI systems should be established to protect health, safety, and fundamental rights. | 中 | SR023 |
| CR018 | The EU Machinery Regulation says machinery accidents should be reduced by inherently safe design, proper installation, and maintenance, and it replaces the prior directive with a regulation. | 中 | SR024 |
| CR019 | McKinsey says ISO 10218 and ISO/TS 15066 focus on robot arms and collaborative robots rather than autonomous humanoids in unstructured settings. | 中 | SR029 |
| CR020 | McKinsey says ambiguity around humanoid-specific standards will constrain mainstream deployment until new standards are finalized and adopted. | 中 | SR029 |
| CR021 | Robotics & Automation News says China’s mass-production push is being accompanied by new humanoid-specific standards because the key question is what happens when something goes wrong. | 中 | SR025 |
| CR022 | Robotics & Automation News says serious robot injuries have occurred even in controlled industrial environments, showing that failures can still happen despite barriers and procedures. | 中 | SR025 |
| CR023 | IAPP says plaintiffs are bringing privacy complaints tied to tracking technologies and the use of consumer data to train AI. | 中 | SR032 |
| CR024 | Stinson says 20 U.S. states were actively enforcing comprehensive privacy laws as of January 2026. | 中 | SR033 |
| CR025 | Pearl Cohen says 2026 is bringing significant AI, digital-identity, and data-protection law changes across the United States and European Union. | 中 | SR034 |
| CR026 | DeepMind says useful physical-world robotics AI requires embodied reasoning and safe action, not just digital-domain competence. | 中 | SR026 |
| CR027 | DeepMind says VLA systems need generality, interactivity, and dexterity, framing Gemini Robotics as a step toward rather than completion of general-purpose robotics. | 中 | SR026 |
| CR028 | The Gemini Robotics paper says translating large multimodal models to physical agents remains a significant challenge. | 中 | SR027 |
| CR029 | The same paper says extra fine-tuning was required for long-horizon dexterous tasks, short-horizon learning from as few as 100 demonstrations, and adaptation to novel embodiments. | 中 | SR027 |
| CR030 | Bain says early humanoid deployments are mostly limited to highly structured environments. | 高 | SR028, SR029 |
| CR031 | Bain says handling and battery life remain gating factors and most current deployments remain pilot-stage with heavy human supervision. | 中 | SR028 |
| CR032 | McKinsey says large-scale deployment requires four bridges: safety for fenceless operations, sustained uptime, greater dexterity and mobility, and radical cost reduction. | 中 | SR029 |
| CR033 | McKinsey says near-term pilots should focus on structured environments because humanoids remain far from human dexterity in unstructured settings. | 中 | SR029 |
| CR034 | Interact Analysis says high prices and dexterity gaps are likely to remain barriers into the next decade. | 高 | SR028, SR030 |
| CR035 | IEEE Spectrum says no one has yet found an application that requires several thousand humanoids per facility. | 中 | SR031 |
| CR036 | Robot Era said in March 2026 that it had raised RMB 1 billion at a valuation above RMB 10 billion. | 中 | SR004 |
| CR037 | Robot Era said the March 2026 round brought its industrial-investor count to 16 across technology, automotive, logistics, semiconductors, telecommunications, and other sectors. | 中 | SR004 |
| CR038 | Robot Era said the March 2026 release listed industrial clients including Samsung, Geely, Renault, Lenovo, Haier, TCL, and Century Golden Resources. | 中 | SR004 |
| CR039 | Robot Era said in May 2026 that it had raised more than USD 200 million in a new financing round led by SF Group. | 中 | SR005 |
| CR040 | Robot Era said the same May 2026 release described deployments across more than ten logistics centers with China Post and SF Group. | 中 | SR005 |
| CR041 | Robot Era said it initiated thousand-unit deliveries in Q2 2026 with growth above 300%. | 中 | SR005 |
| CR042 | KrASIA says most investors and researchers expect the humanoid sector to reach maturity over the next five to ten years. | 中 | SR006 |
| CR043 | KrASIA says the sector still lacks a clear path to revenue. | 中 | SR006 |
| CR044 | KrASIA says some automotive-factory deployments are viewed by insiders as strategic tie-ups rather than real sales. | 中 | SR006 |
| CR045 | Humanoids Daily says the sector has entered a brutally expensive global capital race with a large valuation gap between leaders and followers. | 中 | SR007 |
| CR046 | CNBC says Unitree is pursuing an IPO that could value it at up to 50 billion yuan and that the company says it has been profitable since 2020 with revenue above RMB 1 billion. | 中 | SR008 |
| CR047 | UBTECH said Walker S orders exceeded RMB 800 million and several hundred robots were entering frontline industrial use. | 中 | SR009 |
| CR048 | UBTECH’s HKEX filing says humanoid-robot and intelligent-service-robot revenue reached RMB 820.6 million in 2025. | 中 | SR010 |
| CR049 | AGIBOT said it rolled out its 10,000th robot in March 2026. | 中 | SR011 |
| CR050 | TechCrunch says Figure’s Series C exceeded USD 1 billion at a USD 39 billion valuation. | 中 | SR012 |
| CR051 | Figure says Helix is a generalist humanoid VLA system that learns and improves as it acquires new skills. | 中 | SR013 |
| CR052 | Tsinghua’s profile says Jianyu Chen is an assistant professor at Tsinghua IIIS, earned a Berkeley robotics PhD, and runs a robotics and reinforcement-learning lab. | 中 | SR001 |
| CR053 | The Wire China says Jianyu Chen is RobotEra’s founder and that Tsinghua is the only academic backer it names in embodied AI. | 中 | SR002 |
| CR054 | No retained public source in this run provides Robot Era-specific CE, ISO, UL, incident-log, cybersecurity, OTA, MTBF, or uptime metrics. | 高 | SR021, SR022, SR023, SR024, SR025, SR029 |
| CV001 | March 2026 sources corroborate that Robot Era closed a strategic round of RMB 1 billion at a valuation above RMB 10 billion. | 高 | SV002, SV003 |
| CV002 | Pandaily described the same March 2026 financing as valuing Robot Era above $1.37 billion. | 中 | SV001 |
| CV003 | Robot Era’s retained March 2026 sources say cumulative orders exceeded RMB 500 million. | 高 | SV002, SV003 |
| CV004 | Using the disclosed >RMB 500 million cumulative orders as a rough proxy, a >RMB 10 billion valuation implies an estimated order-book multiple above 20x. | 中 | SV002, SV003 |
| CV005 | Robot Era’s disclosed commercial metrics are orders, shipments, and deployments rather than audited recognized revenue. | 中 | SV003, SV004, SV005 |
| CV006 | Robot Era’s March 2026 company-linked sources say about 50% of the disclosed order volume was overseas. | 中 | SV003, SV004 |
| CV007 | Robot Era’s March 2026 company-linked sources say a customs-inspection logistics solution had single orders exceeding RMB 50 million. | 中 | SV003, SV004 |
| CV008 | Automated Warehouse reported in June 2026 that Robot Era had raised more than $200 million. | 中 | SV005 |
| CV009 | Automated Warehouse reported that Robot Era had deployments across more than 10 logistics centers with China Post and SF Group. | 中 | SV005 |
| CV010 | Automated Warehouse reported that Robot Era initiated thousand-unit deliveries in the second quarter of 2026. | 中 | SV005 |
| CV011 | Figure announced a $675 million Series B at a $2.6 billion valuation in February 2024. | 中 | SV006 |
| CV012 | SiliconANGLE reported that Figure later raised more than $1 billion at a $39 billion valuation in September 2025. | 中 | SV007 |
| CV013 | CNBC reported that Physical Intelligence raised $400 million at a $2.4 billion post-money valuation in November 2024. | 中 | SV008 |
| CV014 | Apptronik announced a $350 million Series A in February 2025. | 高 | SV009, SV010 |
| CV015 | TechCrunch reported that Apptronik framed 2026 and beyond as the period for true commercialization and scaling. | 中 | SV010 |
| CV016 | Agility Robotics announced that Schaeffler made a minority investment in the company in November 2024. | 中 | SV011 |
| CV017 | Agility Robotics said Schaeffler saw potential to deploy a significant number of humanoids across its global network of 100 plants by 2030. | 中 | SV011 |
| CV018 | CNBC reported that Unitree filed for a Shanghai IPO in March 2026 seeking to raise 4.2 billion yuan. | 高 | SV013, SV014 |
| CV019 | CNBC reported that Unitree’s 2025 operating income grew 335% year over year to 1.708 billion yuan. | 中 | SV013 |
| CV020 | Humanoids Daily reported that Unitree’s updated disclosures showed Q1 2026 revenue up 68% year over year while adjusted net profit fell 52% to 40.3 million yuan. | 中 | SV014 |
| CV021 | Humanoids Daily reported that Unitree’s 2025 revenue reached about 1.70 billion yuan with gross margin of about 60.1%. | 中 | SV014 |
| CV022 | Unitree’s official G1 page lists a starting price of US$13.5K excluding tax and shipping. | 中 | SV015 |
| CV023 | CompaniesMarketCap listed UBTECH Robotics at roughly $7.22 billion of market capitalization in June 2026. | 中 | SV016 |
| CV024 | UBTECH’s 2025 annual results announcement said revenue increased 53.3% to RMB 2,001.0 million. | 中 | SV017 |
| CV025 | UBTECH’s 2025 annual results announcement said full-size embodied intelligent humanoid robot products and services generated RMB 820.6 million of revenue in 2025. | 中 | SV017 |
| CV026 | UBTECH’s 2025 annual results announcement said gross margin rose to 37.7% in 2025. | 中 | SV017 |
| CV027 | CompaniesMarketCap listed Symbotic at about $25.66 billion of market capitalization in June 2026. | 中 | SV018 |
| CV028 | CompaniesMarketCap listed Symbotic’s trailing twelve-month revenue at about $2.51 billion in June 2026. | 中 | SV019 |
| CV029 | Using June 2026 market cap and trailing revenue, Symbotic traded at an estimated market-cap-to-sales ratio of about 10x. | 中 | SV018, SV019 |
| CV030 | CompaniesMarketCap listed iRobot at about $14.84 million of market capitalization in June 2026. | 中 | SV020 |
| CV031 | Macrotrends listed iRobot’s 2024 revenue at $682 million. | 中 | SV021 |
| CV032 | Using June 2026 market cap and 2024 revenue, iRobot traded at an estimated market-cap-to-sales ratio close to zero. | 中 | SV020, SV021 |
| CV033 | Bain said early humanoid deployments were mostly limited to highly structured environments. | 中 | SV022 |
| CV034 | Bain said handling and battery life remained gating factors even as intelligence and perception improved. | 中 | SV022 |
| CV035 | McKinsey said the gap between technically demonstrated pilots and commercially viable scale remained wide. | 中 | SV023 |
| CV036 | IEEE Spectrum argued that well-funded humanoid companies could wave unresolved problems away while still raising extraordinary amounts of money. | 中 | SV024 |
| CV037 | China manufacturing depth, policy support, and strategic investors can justify some premium over a generic robotics story for Robot Era. | 中 | SV002, SV013, SV017, SV022 |
| CV038 | Robot Era still merits a discount to the top U.S. embodied-AI rounds because those companies command stronger software-brand and capital-access premiums. | 中 | SV006, SV007, SV008, SV003 |
| CV039 | Unitree and UBTECH show that Chinese humanoid leaders can publish meaningful revenue disclosures before or during capital-markets access. | 中 | SV013, SV014, SV017 |
| CV040 | Robot Era’s public file is stronger on funding momentum and order momentum than on revenue, margin, or liquidity disclosure. | 中 | SV001, SV002, SV003, SV005, SV022, SV023 |
| CV041 | The absence of public preference terms means the headline March 2026 valuation may not equal true common-equity entry value. | 低 | SV001, SV002, SV003 |
| CV042 | A simple 5x multiple on the disclosed >RMB 500 million cumulative orders proxy would imply only about RMB 2.5 billion of value. | 低 | SV003 |
| CV043 | A simple 10x multiple on the disclosed >RMB 500 million cumulative orders proxy would imply about RMB 5 billion of value. | 低 | SV003, SV018, SV019 |
| CV044 | The current >RMB 10 billion reference mark is roughly what a 20x multiple on the disclosed >RMB 500 million orders proxy would imply. | 中 | SV003 |
| CV045 | The public-evidence bear case is consistent with roughly RMB 6 billion to RMB 8.5 billion if order conversion disappoints or sector multiples compress. | 低 | SV003, SV020, SV021, SV022, SV023, SV024 |
| CV046 | The public-evidence base case is roughly RMB 8.5 billion to RMB 10.5 billion, leaving the current mark around the upper edge of defensible value. | 中 | SV001, SV002, SV003, SV013, SV017, SV018, SV019 |
| CV047 | The public-evidence bull case requires revenue conversion, margin disclosure, and repeat deployment proof that the public record does not yet provide. | 低 | |
| CV048 | The chapter’s investable conclusion is research-more with a stretched stance unless diligence closes the revenue, cash, and cap-table gaps. | 中 | SV001, SV002, SV003, SV022, SV023, SV024 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | RobotEra | ROBOTERA official homepage | ROBOTERA official homepage. |
| SO002 | RobotEra | ROBOTERA website application bundle | 7F, VIA Building, Tsinghua Science Park, Haidian District, Beijing. |
| SO003 | Tsinghua University IIIS | Tsinghua IIIS faculty page for Jianyu Chen | Jianyu Chen is an assistant professor at IIIS, Tsinghua University. |
| SO004 | KrASIA | More than a hardware play, Robot Era’s founder sets the record straight on the company’s true ambitions | Founded Aug 2023; Chen emphasizes brain plus body and Lego-like modular hardware. |
| SO005 | The Wire China | Chen Jianyu profile | Chen is RobotEra’s founder, age 33, and the company claims 9 of the world’s 10 largest tech firms as customers. |
| SO006 | PRNASIA | Embodied intelligence pioneer RobotEra secures RMB 1 billion in Series A funding | RobotEra secures RMB 1 billion in Series A funding. |
| SO007 | CnTechPost | Embodied AI startup Robotera raises $200 million new funding round | Embodied AI startup Robotera raises $200 million new funding round. |
| SO008 | Humanoids Daily | RobotEra secures $140M Series A+ backed by automakers Geely and BAIC, claims $70M in orders | RobotEra secures $140M Series A+ backed by automakers Geely and BAIC. |
| SO009 | RoboticsTomorrow | RobotEra raises over USD 200 million in new round led by SF Group, HSG and IDG Capital | 10+ logistics centers with China Post and SF Group; Q2 2026 thousand-unit deliveries and >300% growth. |
| SO010 | Yicai Global | Chinese robot startup RobotEra bags USD 140.5 million in latest fundraiser led by Geely | By Nov 2025 total disclosed capital reached CNY1.8 billion and orders exceeded CNY500 million, with half overseas. |
| SO011 | KrASIA | Backed by China’s automakers, Robot Era moves to fulfill major orders | Backed by China’s automakers, Robot Era moves to fulfill major orders. |
| SO012 | PRNewswire | RobotEra showcases its hexa-core robotics lineup at CES 2026 | RobotEra showcases its hexa-core robotics lineup at CES 2026. |
| SO013 | Humanoid.Guide | RobotEra L7 | L7 is listed at 171 cm, 55 DoF, about 70 kg, with 20 kg dual-arm payload. |
| SO014 | Humanoid.Guide | XHAND1 | XHAND1 is listed with 12 DoF, 80N grip force, and 25 kg lift. |
| SO015 | PRNASIA | RobotEra raises over USD 200 million in new round led by SF Group, HSG and IDG Capital | RobotEra raises over USD 200 million in new round led by SF Group, HSG and IDG Capital. |
| SO016 | PRNASIA | RobotEra raises RMB 1 billion in strategic round, valuation exceeds RMB 10 billion | Valuation exceeds RMB 10 billion with broad industry support. |
| SO017 | Humanoids Daily | RobotEra reaches $1.4 billion valuation in RMB 1 billion strategic round | RobotEra reaches $1.4 billion valuation in RMB 1 billion strategic round. |
| SO018 | Interesting Engineering | Chinese startup RobotEra raises big | Coverage supports the L7, Q5, and XHAND1 lineup narrative around CES 2026. |
| SO019 | The AI Insider | Chinese humanoid robotics company RobotEra secures USD 140M in Series A funding | Chinese humanoid robotics company RobotEra secures USD 140M in Series A funding. |
| SO020 | The AI Insider | China’s humanoid robot maker RobotEra raises over USD 200M in new funding round | China’s humanoid robot maker RobotEra raises over USD 200M in new funding round. |
| SO021 | Crunchbase News | Unicorn count hits four-year high with robotics and AI entrants including RobotEra | Crunchbase’s March 2026 unicorn board included RobotEra around $1.5 billion after about $145 million raised. |
| SO022 | TechCrunch | Almost 40 new unicorns have been minted so far this year. Here they are. | TechCrunch also listed RobotEra among new 2026 unicorns. |
| SO023 | CKGSB | Humanoid robots in China: progress, limits, and reality | China’s humanoid sector still faces progress limits and commercial reality gaps. |
| SO024 | The Next Web | China humanoid robot boom faces commercialization reality check | Buyer satisfaction is only 23%, with battery life and commercialization still problematic. |
| SO025 | The Robot Report | Chinese robotics outlook 2026 includes growth and competitive pressure | China’s robotics outlook for 2026 includes growth but also competitive pressure and likely shakeout. |
| SO026 | PRNewswire | RobotEra and UNIDO forge strategic partnership to advance global industrial transformation with embodied intelligence | RobotEra and UNIDO forge strategic partnership to advance global industrial transformation with embodied intelligence. |
| SM001 | BCC Research via Research and Markets | Humanoid Robot: Applications, Verticals and Global Markets | The global market for humanoid robots is expected to grow from $1.9 billion in 2025 and is projected to reach $11 billion by the end of 2030, at a compound annual growth rate (CAGR) of 42.8%. |
| SM002 | Humanoid.guide | Humanoid.guide Publishes Landmark 2026 Humanoid Robot Market Report | A central conclusion threads multiple chapters: manipulation matters most; dexterous, robust hands and the data to train them are gating factors for useful work at scale. |
| SM003 | Robozaps | Humanoid Robot Industry Report 2026 | This report tracks 26 humanoid robots across 7 countries, with over $5 billion in total industry investment and 14 robots commercially available today. |
| SM004 | International Federation of Robotics | Top 5 Global Robotics Trends 2026 | The global market value of industrial robot installations has reached an all-time high of US$ 16.7 billion, and humanoid applications in warehousing and manufacturing are coming into focus worldwide. |
| SM005 | The Robot Report | 2026 State of the Robotics Industry Report | Agility Robotics' Digit has been used by GXO Logistics, where the robot has moved more than 100,000 totes, and the report frames 2026 as a year of less dancing and more doing for humanoids. |
| SM006 | State Council Information Office / Xinhua | China releases national standard system for humanoid robotics and embodied AI | China took a significant step toward regulating its rapidly growing humanoid robotics industry with a standard system covering the entire industrial chain and lifecycle of humanoid robots and embodied AI. |
| SM007 | Shanghai Municipal Government | Shanghai maps out plan for embodied AI | The plan targets 100 leading enterprises, 100 application scenarios, 100 globally competitive products, and industry output of more than 50 billion yuan by 2027. |
| SM008 | International Federation of Robotics | World Robotics 2025 report – INDUSTRIAL ROBOTS – released by IFR | The new World Robotics 2025 statistics show 542,000 industrial robots installed in 2024, with China accounting for 295,000 installations and 54% of global deployments. |
| SM009 | International Federation of Robotics | World Robotics 2025 report – SERVICE ROBOTS – released by IFR | Transportation and logistics robots led professional service robotics with 102,900 units sold in 2024, while hospitality exceeded 42,000 and medical robots rose to around 16,700 units. |
| SM010 | Interact Analysis | Warehouse automation – What to expect in 2026 | Heading into 2026, Interact expects the disconnect between a weak macro backdrop and strong underlying demand for warehouse automation to narrow as broader conditions improve. |
| SM011 | Interact Analysis | Over One-Quarter of Warehouses Automated by 2027 | More than a quarter (26%) of warehouses are expected to be automated by 2027, from 18% in 2021 and far below current end-state potential. |
| SM012 | McKinsey & Company | Getting warehouse automation right | McKinsey says robot shipments could increase by up to 50 percent each year through 2030, with warehouse automation likely to grow by more than 10 percent per year. |
| SM013 | KrASIA | Backed by China’s automakers, Robot Era moves to fulfill major orders | Robot Era said it had secured more than RMB 500 million in commercial orders for 2025 and described use cases across logistics, manufacturing, and commercial services. |
| SM014 | PR Newswire | ROBOTERA and UNIDO Forge Strategic Partnership to Advance Global Industrial Transformation with Embodied Intelligence | ROBOTERA and UNIDO said they will focus on manufacturing, logistics, and commercial services to develop and promote embodied-intelligence solutions. |
| SM015 | PR Newswire / PRNASIA | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | ROBOTERA said it had deployments across more than ten logistics centers with China Post and SF Group and initiated thousand-unit deliveries in Q2 2026. |
| SM016 | Robotics & Automation News | The state of humanoid robotics: From research labs to real-world potential | In 2026, humanoids are less about abstract futurism and more about whether they can find real work in real environments such as logistics and manufacturing. |
| SM017 | RobotToday | CES 2026 Exposes a Structural Divide in Humanoid Robotics | Chinese manufacturers dominated physical presence, shipment volume, and price accessibility at CES 2026, while the commercialization gap between software ambition and real deployment remained visible. |
| SM018 | China Daily | Embodied AI turned into growth driver | China Daily says the Development Research Center forecasts the domestic embodied-AI market could reach 400 billion yuan by 2030 and exceed 1 trillion yuan by 2035. |
| SM019 | China Daily | Humanoid robotics making big strides | According to figures disclosed by MIIT in January, China has more than 140 humanoid robot manufacturers and over 330 product models, but scaling remains constrained by component and data bottlenecks. |
| SM020 | China Daily | Humanoid robots pivot to real-world use | Boao Forum speakers said the industry is pivoting from showpieces toward real-world utility, with early adoption in automotive and home appliance manufacturing plus logistics and sorting. |
| SM021 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | Automated Warehouse says ROBOTERA had deployments across more than 10 logistics centers with China Post and SF Group and reportedly initiated thousand-unit deliveries in Q2 2026. |
| SM022 | CNBC | Investors bet humanoid robots will transform industry and homes over the next decade | CNBC cites Barclays saying the humanoid market today is only $2 billion to $3 billion, but could reach $200 billion by 2035; China accounted for 85% of installations last year and produces robots at roughly half Western cost. |
| SM023 | China Daily | China introduces a standard framework for humanoid and embodied intelligence | The newly issued 2026 edition of the Humanoid and Embodied Intelligence Standard System is described as China's first comprehensive top-level standard design covering the full lifecycle of the industry. |
| SM024 | BusinessWire / Research and Markets | Global Humanoid Robots Market Report 2026-2040 | The release says Unitree's $5,900 R1 shocked the market, Goldman Sachs saw manufacturing costs down 40% year over year, and production targets are shifting from pilot scale to industrial scale. |
| SM025 | Humanoids Daily | Robotera Secures $280 Million Strategic Round as Logistics Deployment Scales | Humanoids Daily says Robotera reported 300% growth in 2026 and began delivering thousands of units across more than 10 Chinese logistics centers with China Post and SF Group. |
| SP002 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | ROBOTERA said it builds its own robotics hardware, with over 95% of core components developed internally. |
| SP004 | China Daily | Chinese startup Robotera unveils new humanoid robot | Standing at 171 cm tall, the bipedal robot can simultaneously achieve high-dynamic movement and fine manipulation. |
| SP005 | AGIBOT | AGIBOT Innovation (Shanghai) Technology Co., Ltd. | AGIBOT initiates the commercial mass production of general robots. |
| SP006 | AGIBOT / PR Newswire | AGIBOT Unveils New Generation of Embodied AI Robots and Models, Accelerating Real-World Deployment of Physical AI | In March 2026, AGIBOT announced the rollout of its 10,000th robot, marking a major milestone in large-scale production and deployment. |
| SP007 | CnTechPost | Agibot unveils 5 new robot platforms, targets over $1.46 billion in revenue by 2027 | Cumulative shipments of the company's full-size humanoid robot, the Expedition series, have now broken through the 2,000-unit milestone. |
| SP008 | Unitree Robotics | Humanoid robot G1_Humanoid Robot Functions_Humanoid Robot Price | Weight (With Battery) About 35kg ... Total Degrees of Freedom 23 - 43 ... Battery Life About 2h. |
| SP009 | Unitree Robotics | Unitree G1 | $13,500.00 USD. |
| SP010 | CnTechPost | Unitree plans $610 million China IPO amid humanoid robot boom | Hangzhou-based Unitree recorded revenue of 1.71 billion yuan in 2025, a 335.4% surge from the previous year. |
| SP011 | CNBC | Nvidia picks Unitree for humanoid robot platform as Chinese startup eyes IPO | Nvidia picks Unitree for humanoid robot platform as Chinese startup eyes IPO. |
| SP012 | Fourier Robotics | Fourier_General Purpose_Humanoid Robot_Assistant | The Fourier GR-1 breaks new ground in humanoid robotics for practical applications. |
| SP013 | Fourier Robotics | New_Milestone_of_Humanoid_Robotics | GR-2 introduces 12-DoF dexterous hands—double the dexterity of previous models. |
| SP014 | UBTECH Robotics | UBTECH Walker S Industrial Humanoid Robot | For Multi-task Industrial Scenarios | UBTECH Robotics | Walker S can now achieve precise and safe synchronized operations on the factory assembly line. |
| SP015 | UBTECH / PR Newswire | UBTECH Humanoid Robot Walker S2 Begins Mass Production and Delivery, with Orders Exceeding 800 Million Yuan | UBTECH has begun mass production and delivery of the first batch of several hundred full-size industrial humanoid robots, Walker S2. |
| SP016 | HKEX | UBTECH ROBOTICS CORP LTD annual results announcement for the year ended December 31, 2025 | Revenue from full-size embodied intelligent humanoid robot products and services grew rapidly ... to RMB820.6 million. |
| SP018 | The Robot Report | Galbot brings in $300M to scale mobile manipulator deployments | Galbot brings in $300M to scale mobile manipulator deployments. |
| SP019 | CnTechPost | Chinese robotics startup Galbot secures major state backing to scale AI models | Galbot has completed a new round of financing worth 2.5 billion yuan ($363 million), led by state-backed capital. |
| SP020 | Galbot / PR Newswire | Galbot Secures Over $300 Million in New Funding, Breaking Records with $3 Billion Valuation in China's Humanoid Robot Sector | Galbot has successfully completed a new funding round exceeding $300 million, bringing the company's total funding to $800 million. |
| SP021 | AgileX Robotics | AgileX Robotics official site | The site lists mobile chassis, robot arms, ALOHA series embodied-intelligence kits, and data-collection or education kits. |
| SP023 | Figure | Helix | Figure | Helix, our AI system, is a generalist humanoid Vision-Language-Action model. |
| SP024 | TechCrunch | BMW will deploy Figure’s humanoid robot at South Carolina plant | BMW will deploy Figure’s humanoid robot at South Carolina plant. |
| SP025 | TechCrunch | Figure reaches $39B valuation in latest funding round | The round exceeded $1 billion and values Figure at $39 billion. |
| SP026 | Physical Intelligence | Physical Intelligence (π) | Physical Intelligence is bringing general-purpose AI into the physical world ... to create a model that will control any robot to do any task. |
| SP027 | Physical Intelligence | A VLA with Open-World Generalization | Our latest generalist policy, π0.5, extends π0 and enables open-world generalization. |
| SP029 | 1X | NEO Home Robot | NEO works autonomously by default ... NEO $200 Deposit. |
| SP030 | TechCrunch | Norway's 1X is building a humanoid robot for the home | The company is quick to add that the Gamma is a long way from commercial scaling and deployment. |
| SP031 | Boston Dynamics | Atlas Humanoid Robot | Boston Dynamics | Atlas requires minimal supervision ... Battery Life 4 hours ... Weight Capacity (Instant) 50 kg. |
| SP032 | Boston Dynamics | An Electric New Era for Atlas | Boston Dynamics | This journey will start with Hyundai. |
| SP033 | The Robot Report | Boston Dynamics shows Atlas humanoid working at Georgia Hyundai plant | Boston Dynamics shows Atlas humanoid working at Georgia Hyundai plant. |
| SP035 | TechCrunch | Agility’s humanoid robots are going to handle your Spanx | Agility announced that it has entered into a formal deal following a successful pilot with logistics giant GXO. |
| SP036 | Forbes | Digit Gets A Job: Agility Robotics And Toyota Sign Robots-As-A-Service Deal | Toyota Motor Manufacturing Canada ... has signed a commercial Robots-as-a-Service agreement following a successful pilot with Digit. |
| SP037 | Figure | Figure | Figure 03 is a general purpose humanoid robot for every day. |
| SP038 | Figure | Figure 03 | Figure | Payload 20KG ... Runtime 5HR. |
| SP039 | Agility Robotics | Humanoid Solutions | Agility | Digit is the first humanoid to prove real value at scale. |
| SI001 | PRNASIA | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | |
| SI002 | PRNASIA | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | |
| SI003 | CnTechPost | Embodied AI startup Robotera raises over $200 million in new funding round | |
| SI004 | Yicai Global | Chinese Robot Startup Robotera Bags USD140.5 Million in Latest Fundraiser, Led by Geely | |
| SI005 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | |
| SI006 | KrASIA | Backed by China’s automakers, Robot Era moves to fulfill major orders | |
| SI007 | China Daily | Chinese startup Robotera unveils new humanoid robot | |
| SI008 | Pulse 2.0 | ROBOTERA Raises Over $200 Million To Scale Humanoid Robotics Commercialization | |
| SI009 | Unitree Robotics | Unitree G1 | |
| SI010 | BotInfo.ai | Unitree G1 Price 2026: Base $16K, EDU from $43.9K — All 16 Configs | |
| SI011 | CnTechPost | Unitree plans $610 million China IPO amid humanoid robot boom | |
| SI012 | UBTECH / PR Newswire | UBTECH Humanoid Robot Walker S2 Begins Mass Production and Delivery, with Orders Exceeding 800 Million Yuan | |
| SI013 | Hong Kong Exchanges and Clearing | UBTECH ROBOTICS CORP LTD annual results announcement for the year ended December 31, 2025 | |
| SI014 | SEC | iROBOT CORPORATION Form 10-K for fiscal year ended December 28, 2024 | |
| SI015 | Symbotic Investor Relations | Symbotic Reports Fourth Quarter and Fiscal Year 2025 Results | |
| SI016 | SEC | Serve Robotics Announces Third Quarter 2025 Results (Exhibit 99.1) | |
| SI017 | Business Wire / A3 | Robot Orders Hold Steady in Q1 2026 as Demand Broadens Across Non-Automotive Industries | |
| SI018 | Business Wire / ResearchAndMarkets | Global Humanoid Robots Market Report 2026-2040: Humanoid Robot Pricing Drops Below $10,000 as Market Expansion Accelerates | |
| SI019 | Bain & Company | Humanoid Robots: From Demos to Deployment | |
| SI020 | McKinsey & Company | Humanoid robots: Crossing the chasm from concept to commercial reality | |
| SI021 | Interact Analysis | Despite hype and large addressable market, humanoid robot adoption will remain low | |
| SI022 | SCIO | China's humanoid robots step from spectacle toward scalable industrial reality | |
| SI023 | OFweek | Ten Predictions for China's Humanoid Robot Industry in 2026 (Part 2) | |
| SI024 | KrASIA | Bubble or breakthrough? China’s humanoid robotics race faces reality check | |
| SI025 | RoboStore | Unitree Humanoid Robots | RoboStore | |
| SE001 | ROBOTERA | robotera.com client bundle (index.OeOBfVk3.js) | |
| SE002 | PR Newswire / ROBOTERA | ROBOTERA Showcases Its "Hexa-Core" Robotics Lineup at CES 2026 | Standing 171 cm tall and weighing 70 kg, L7 features human-like proportions, 55 degrees of freedom, and can carry up to 20 kg with both arms. |
| SE003 | PRNASIA / ROBOTERA | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | |
| SE004 | PRNASIA / ROBOTERA | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | |
| SE005 | Robotics & Automation News | Robot Era unveils ‘groundbreaking’ new large computing model for general-purpose robotics | ERA-42 integrates vision, language, touch, and body posture into a single model to generalize across diverse tasks and environments. |
| SE006 | Humanoids Daily | Robotera Unveils "End-to-End" Humanoid Solution to Close Logistics’ Final Automation Gap | Robotera claims this is the world's first application of an end-to-end VLA embodied model in a real-world logistics scenario. |
| SE007 | China Daily | Chinese startup Robotera unveils new humanoid robot | |
| SE008 | KrASIA / 36Kr | Backed by China’s automakers, Robot Era moves to fulfill major orders | Over the past two years, it has released modular components including dexterous hands, the Q5 commercial service robot, the bipedal L7 robot, and the M7 upper-body modular platform. |
| SE009 | KrASIA / 36Kr | Bubble or breakthrough? China’s humanoid robotics race faces reality check | The industry has yet to find a clear path to revenue. |
| SE010 | Aparobot | L7 - Robot Details, Use Case and Specifications | |
| SE011 | Aparobot | Q5 - Robot Details, Use Case and Specifications | |
| SE012 | Aparobot | ROBOTERA - Robotics Company on Aparobot.com | |
| SE013 | ChipSilicon | Robotera L7 | Humanoid Robot | |
| SE014 | RoboHorizon | RobotEra's Humanoid Aims to End Warehouse Chaos | |
| SE015 | Origin of Bots | L7 by Robot Era | Origin of Bots | |
| SE016 | Origin of Bots | Q5 by Robot Era | Origin of Bots | |
| SE017 | Humanoid Press | RobotEra Q5 — 44‑DoF Ultra‑Humanoid Service Robot | |
| SE018 | GitHub / RobotEra TECHNOLOGY CO., LTD. | RobotEra TECHNOLOGY CO., LTD. organization page | |
| SE019 | GitHub / ROBOTERA | roboterax/ros2_sdk | This repository contains example ROS 2 packages for STAR1 robot control, developed by RobotEra. |
| SE020 | GitHub / ROBOTERA | roboterax/xbot_sdk_api | RobotEraSDKAPI is a ROS2-based Python library that provides complete robot-control functions including trajectory control, MPC, and XHand/XHand Lite control. |
| SE021 | Bain & Company | Humanoid Robots: From Demos to Deployment | Intelligence and perception are nearing parity with humans, while handling and battery life remain gating factors. |
| SE022 | McKinsey & Company | Humanoid robots: Crossing the chasm from concept to commercial reality | |
| SE023 | Interact Analysis | Humanoid-Robots-May-2025.pdf | Major barriers include regulatory and safety concerns, dexterity limitations, cost, and questions over whether humanoid is the optimal form-factor for most tasks. |
| SE024 | State Council Information Office of China | China's humanoid robots step from spectacle toward scalable industrial reality | Most humanoid robots can only operate for two to three hours per charge—insufficient for serious industrial use. |
| SE025 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | |
| SE026 | CnTechPost | Embodied AI startup Robotera raises over $200 million in new funding round | |
| SU001 | PR Newswire | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | Cumulative orders have exceeded RMB 500 million, 50% from international clients, including nine of the world's top 10 publicly listed tech companies, some placing repeat orders up to six times. |
| SU002 | PR Newswire | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | ROBOTERA has achieved the first product-market fit (PMF) in the embodied intelligence sector, with deployments across more than ten logistics centers in collaboration with China Post and SF Group. In Q2 2026, ROBOTERA initiated thousand-unit deliveries. |
| SU003 | KrASIA / 36Kr English | Backed by China’s automakers, Robot Era moves to fulfill major orders | Robot Era’s products have entered North America, Europe, Japan, and South Korea, with overseas revenue accounting for about half of total revenue. |
| SU004 | Yicai Global | Chinese Robot Startup Robotera Bags USD140.5 Million in Latest Fundraiser, Led by Geely | Since the beginning of the year, Robotera has been scaling up its commercial rollout, with total orders exceeding CNY500 million. Overseas business now accounts for half of this figure. |
| SU005 | Humanoids Daily | Robotera Secures $140M Series A+ Backed by Automakers Geely and BAIC, Claims $70M in Orders | Robotera claims to have secured over RMB 500 million (~$70 million) in commercial orders for 2025. |
| SU006 | Humanoids Daily | Robotera Secures $280 Million Strategic Round as Logistics Deployment Scales | These robots are currently deployed across more than 10 logistics centers in North, East, and South China through partnerships with China Post and SF Group. |
| SU007 | Humanoids Daily | Robotera Unveils "End-to-End" Humanoid Solution to Close Logistics’ Final Automation Gap | Robotera claims this is the world's first application of an end-to-end VLA embodied model in a real-world logistics scenario. |
| SU008 | RoboticsTomorrow | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | ROBOTERA has achieved the first product-market fit in embodied intelligence, with deployments across over ten logistics centers in collaboration with China Post and SF Group. |
| SU009 | CnEVPost / CnTechPost | Embodied AI startup Robotera raises over $200 million in new funding round | |
| SU010 | Ventureburn | ROBOTERA Raises USD 200 Million To Scale Global Robotics Expansion | |
| SU011 | Mike Kalil | The Rise of Beijing’s RobotEra | RobotEra hasn't shipped nearly as many humanoids as Chinese competitors like Unitree, UBTECH, AgiBot, and NOETIX Robotics, but it appears to have found footing in the logistics sector. |
| SU012 | Mike Kalil | RobotEra’s Humanoids Enter China’s Logistics Sector | According to RobotEra, its flagship L7 humanoids are already operating in more than 10 logistics centers across the mainland through partnerships with major delivery operators like SF Express and China Post. |
| SU013 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | ROBOTERA asserted that it has achieved a product-market fit in the embodied intelligence sector, with deployments across more than 10 logistics centers in collaboration with China Post and SF Group. |
| SU014 | EqualOcean | Embodied Intelligence Company ROBOTERA Raises Nearly CNY 500 Mn in Series A Round | As of June, more than 200 units have been delivered this year, with hundreds more orders in mass production. Over 50% of all orders come from overseas clients. |
| SU015 | IEEE Spectrum | Humanoid Robots: The Scaling Challenge | But as of now, the market for humanoid robots is almost entirely hypothetical. Even the most successful companies in this space have deployed only a small handful of robots in carefully controlled pilot projects. |
| SU016 | McKinsey & Company | Humanoid robots: Crossing the chasm from concept to commercial reality | For humanoid robots to leave pilot purgatory and deliver real value at scale in the workplace, tech providers must focus on building four essential bridges. |
| SU017 | Robotics & Automation News | Investors warn ‘AI hype is fueling a bubble in humanoid robotics’ | Industrial and logistics robots already generate revenue and can deliver measurable results, while humanoids can't yet prove their commercial value. |
| SU018 | Forbes | The Billion-Dollar Robot Race Is Moving Faster Than The Robots | The evidence points toward a sector that has drawn more capital than its current technological maturity can absorb. |
| SU019 | China Post党建网 | 企业文化实践案例∙守正创新│“无”中生“有” 打造新质生产力——浙江金华邮政构建无人化配送服务网络 | |
| SU020 | China Daily | Chinese startup Robotera unveils new humanoid robot | Robotera revealed that, so far it has shipped 200 humanoid robots in 2025, with over one hundred orders currently in progress. |
| SU021 | State Council Information Office / Xinhua | China's humanoid robots step from spectacle toward scalable industrial reality | The sensible strategy... is to begin commercialization in structured, high-demand sectors such as manufacturing and logistics, building self-sustaining business models there. |
| SU022 | PR Newswire Asia | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | |
| SU023 | PR Newswire Asia | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | |
| SU024 | Aparobot | ROBOTERA - Robotics Company on Aparobot.com | Its ability to raise significant capital, ship hundreds of robots, and attract notable industrial partners (e.g., Lenovo, Geely, Haier, SF Express, Renault) reflects early commercial traction beyond mere lab demonstrations. |
| SU025 | ByteDance Seed | Bytedance Seed-Robotics | The Seed Robotics team tackles challenges in general-purpose intelligent robotics, developing industry-leading technologies in foundational models, perception, manipulation, and interaction. |
| SR001 | Tsinghua University IIIS | Chen Jianyu - Institute for Interdisciplinary Information Sciences (IIIS) | I am an assistant professor in Institute for Interdisciplinary Information Sciences (IIIS) at Tsinghua University. Prior to that, I was working with Prof. Masayoshi Tomizuka at the University of California, Berkeley and received my Ph.D. degree in 2020. |
| SR002 | The Wire China | Chen Jianyu (陈建宇) | Chen Jianyu, 33, is founder of RobotEra, a Beijing-based maker of humanoid robots. |
| SR003 | ROBOTERA | L7 brochure (full-size bipedal humanoid robot) | |
| SR004 | PRNASIA / ROBOTERA | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | ROBOTERA has closed a strategic funding round of RMB 1 billion, pushing its valuation past RMB 10 billion. |
| SR005 | PRNASIA / ROBOTERA | ROBOTERA Raises Over USD 200 Million in New Round Led by SF Group, HSG and IDG Capital | In Q2 2026, ROBOTERA initiated thousand-unit deliveries, with growth exceeding 300%. |
| SR006 | KrASIA / 36Kr | Bubble or breakthrough? China’s humanoid robotics race faces reality check | Still, while those events jolted sentiment, they didn’t solve the core problem: the industry has yet to find a clear path to revenue. |
| SR007 | Humanoids Daily | The Great Valuation Chasm: A 2025 Guide to the Humanoid Robotics Capital Race | |
| SR008 | CNBC | China’s Unitree heats up humanoid robot race as IPO valuation reportedly hits $7 billion | Unitree Robotics is planning an initial public offering that could value the company at up to 50 billion yuan ($7 billion). |
| SR009 | PR Newswire / UBTECH | UBTECH Humanoid Robot Walker S2 Begins Mass Production and Delivery, with Orders Exceeding 800 Million Yuan | Since early 2025, UBTECH's Walker series humanoid robots have accumulated orders exceeding 800 million yuan. |
| SR010 | Hong Kong Stock Exchange | UBTECH Robotics annual results filing | services increased by approximately 2,203.7% from RMB35.6 million in 2024 to RMB820.6 million in 2025. |
| SR011 | PR Newswire / AGIBOT | AGIBOT Unveils New Generation of Embodied AI Robots and Models, Accelerating Real-World Deployment of Physical AI | In March 2026, AGIBOT announced the rollout of its 10,000th robot, marking a major milestone in large-scale production and deployment. |
| SR012 | TechCrunch | Figure reaches $39B valuation in latest funding round | The round, which “exceeded $1 billion,” said Figure, was led by Parkway Venture Capital. |
| SR013 | Figure | Helix | Helix, our AI system, is a generalist humanoid Vision-Language-Action model that learns and improves over time as it acquires new skills. |
| SR014 | U.S. Bureau of Industry and Security | Commerce Further Restricts China’s Artificial Intelligence, Advanced Computing Capabilities | Restrict the Chinese Communist Party’s ability to acquire and develop high-performance and exascale computing capabilities. |
| SR015 | Congressional Research Service | U.S. Export Controls and China: Advanced Semiconductors | Since 2018, the U.S. government has sought to strengthen U.S. export controls to restrict the export of advanced semiconductors to China. |
| SR016 | Foundation for Defense of Democracies | As Chinese Robotics Industry Surges, Senate Considers Limited Federal Procurement Ban | The proposed bill would prevent federal funds from being used to purchase unmanned ground vehicle systems produced by firms under the jurisdiction of a foreign adversary. |
| SR017 | IEEE Spectrum | US Ban on Chinese Robots Could Reshape Supply Chains | The catch: American robot makers still need Chinese-made components. |
| SR018 | Rest of World | The trade war is delaying the future of humanoid robot workers | Chinese companies make up 63% of the global humanoid robot supply chain. |
| SR019 | Sidley Austin | New U.S. Export Controls on Advanced Computing Items and Artificial Intelligence Model Weights: Seven Key Takeaways | For the first time, BIS will control AI “model weights.” |
| SR020 | CSIS | The Limits of Chip Export Controls in Meeting the China Challenge | Export restrictions alone cannot substitute for comprehensive industrial and research policy measures necessary to ensure U.S. leadership. |
| SR021 | Occupational Safety and Health Administration | Robotics - Standards | There are currently no specific OSHA standards for the robotics industry. |
| SR022 | Occupational Safety and Health Administration | OSHA Technical Manual (OTM) - Section IV: Chapter 4 | This chapter is intended as a guide to robot systems found in industrial applications. |
| SR023 | EUR-Lex | Regulation (EU) 2024/1689 (Artificial Intelligence Act) | Common rules for high-risk AI systems should be established. |
| SR024 | EUR-Lex | Regulation (EU) 2023/1230 on machinery | The social cost of the large number of accidents caused directly by the use of machinery can be reduced by inherently safe design and construction of machinery. |
| SR025 | Robotics & Automation News | Why China’s new humanoid robot standards could change the industry | The emergence of humanoid robots introduces a different set of challenges. |
| SR026 | Google DeepMind | Introducing Gemini Robotics and Gemini Robotics-ER, AI models designed for robots to understand, act and react to the physical world | To be useful and helpful to people, AI models for robotics need three principal qualities: they have to be general, interactive, and dexterous. |
| SR027 | arXiv | Gemini Robotics: Bringing AI into the Physical World | Their translation to physical agents such as robots remains a significant challenge. |
| SR028 | Bain & Company | Humanoid Robots: From Demos to Deployment | Early deployments are mostly limited to highly structured environments. |
| SR029 | McKinsey & Company | Humanoid robots: Crossing the chasm from concept to commercial reality | Humanoids cannot move from concept to coworker until all four are built. |
| SR030 | Interact Analysis | Humanoid Robots May 2025 | Our outlook remains cautious due to several key barriers that hinder widespread adoption, including high prices and the gap in the dexterity needed to match human productivity levels. |
| SR031 | IEEE Spectrum | Humanoid Robots: The Scaling Challenge | I don’t think anyone has found an application for humanoids that would require several thousand robots per facility. |
| SR032 | IAPP | Understanding emerging digital litigation trends in the US | Plaintiffs are also bringing privacy complaints for a host of emerging technologies, such as tracking pixels and using consumer data to train AI. |
| SR033 | Stinson LLP | A New Era of Comprehensive Privacy Laws and the Surge in Data Privacy Litigation: Important Updates for 2026 | As of January 2026, 20 states are actively enforcing comprehensive privacy laws. |
| SR034 | Pearl Cohen | New Privacy, Data Protection and AI Laws in 2026 | Several significant legislative developments are reshaping the legal landscape governing artificial intelligence, digital identity, and data protection across the United States and European Union. |
| SV001 | Pandaily | Robot Era Raises USD 137 Million Strategic Round, Valuation Exceeds USD 1.37 Billion | |
| SV002 | TNGlobal | Singtel Innov8 backs ROBOTERA $145M strategic round | |
| SV003 | PR Newswire / PRN Asia | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | RMB 500 Million in Orders Marks ROBOTERA’s Commercial Leadership. |
| SV004 | Newswire.ca | ROBOTERA Raises RMB 1 Billion in Strategic Round, Valuation Exceeds RMB 10 Billion with Broad Industry Support | |
| SV005 | Automated Warehouse | Chinese humanoid developer ROBOTERA raises over $200M | |
| SV006 | PR Newswire | Figure Raises $675M at $2.6B Valuation and Signs Collaboration Agreement with OpenAI | |
| SV007 | SiliconANGLE | Humanoid robot startup Figure raises $1B+ at $39B valuation | |
| SV008 | CNBC | Jeff Bezos and OpenAI invest in robot startup Physical Intelligence at $2.4 billion valuation | Physical Intelligence ... has raised $400 million at a $2.4 billion post-money valuation. |
| SV009 | Apptronik | Apptronik Raises $350 Million in Series A Funding | |
| SV010 | TechCrunch | Apptronik, which makes humanoid robots, raises $350M as category heats up | |
| SV011 | Agility Robotics | Agility Robotics Announces Strategic Investment and Agreement with Motion Technology Company Schaeffler Group | |
| SV012 | TechCrunch | Agility Robotics lays off some staff amid commercialization focus | |
| SV013 | CNBC | Unitree plans Shanghai IPO, testing interest in humanoid robots | |
| SV014 | Humanoids Daily | Inside Unitree’s Prospectus: Revenue Climbs and Profits Dip as Star Market IPO Hearing Approaches | |
| SV015 | Unitree Robotics | Humanoid robot G1_Humanoid Robot Functions_Humanoid Robot Price | |
| SV016 | CompaniesMarketCap | UBTECH Robotics (9880.HK) - Market capitalization | |
| SV017 | HKEX | UBTECH ROBOTICS CORP LTD annual results announcement for the year ended December 31, 2025 | |
| SV018 | CompaniesMarketCap | Symbotic (SYM) - Market capitalization | |
| SV019 | CompaniesMarketCap | Symbotic (SYM) - Revenue | |
| SV020 | CompaniesMarketCap | iRobot (IRBT) - Market capitalization | |
| SV021 | Macrotrends | IRobot Revenue 2011-2025 | IRBT | |
| SV022 | Bain & Company | Humanoid Robots: From Demos to Deployment | Humanoid robotics are drawing capital and headlines, but early deployments are mostly limited to highly structured environments. |
| SV023 | McKinsey | Humanoid robots: Crossing the chasm from concept to commercial reality | The gap between what is technically demonstrated in pilots and what is commercially viable at scale remains wide. |
| SV024 | IEEE Spectrum | Humanoid Hype Dominates Top Robotics Stories of 2025 | Well-funded humanoid companies seem quite happy to wave those problems away while continuing to raise extraordinary amounts of money. |
| SV025 | Grandall Law Firm | Grandall Shanghai Assisted State-owned Companies Completing Equity Investment in Frontier | |
| SV026 | CompaniesMarketCap | UBTECH Robotics (9880.HK) - Revenue | |
| SV027 | Macrotrends | Symbotic Revenue 2021-2025 | SYM | |
| SV028 | Macrotrends | Symbotic Price to Sales Ratio 2021-2025 | SYM | |
| SV029 | Macrotrends | IRobot Price to Sales Ratio 2010-2023 | IRBT | |
| SV030 | CompaniesMarketCap | UBTECH Robotics (9880.HK) - Revenue |