10月10日，无锡巨蟹智能、墨现科技两笔具身智能产业链融资同步官宣。自变量机器人同时出现在两家公司的投资方名单中——前者做一体化关节模组，后者做柔性触觉传感器。一家本体企业同日押注驱动与感知两个上游环节，动作密集。《科创板日报》记者注意到，近一年来，从银河通用、星海图到智元机器人、鹿明机器人，越来越多仍在融资阶段的具身智能本体企业，开始向上游零部件、数据服务、场景应用公司出手。据记者不完全统计，截至2026年9月，国内已有超20家具身智能企业累计对外投资超100起。“边融边投”不再是头部企业的个例，正在逐步成为人形机器人赛道里多数企业的选择。
On October 10, two embodied intelligence supply chain funding rounds were announced simultaneously—Wuxi Juxie Intelligent and Moxian Technology. AgiBot appeared on both companies' investor lists—the former makes integrated joint modules, the latter makes flexible tactile sensors. One body company betting on both actuation and perception upstream links on the same day is密集 action. Science and Technology Innovation Board Daily reporters noted that over the past year, from Galbot and Galaxea to AgiBot and Luming Robotics, more and more embodied intelligence body companies still in funding stages are beginning to invest in upstream components, data services, and scenario application companies. According to incomplete reporter statistics, as of September 2026, over 20 embodied intelligence companies in China have cumulatively made over 100 external investments. "Funding while investing" is no longer an isolated case of leading companies—it is gradually becoming the choice of most companies in the humanoid robot track.
巨蟹智能本轮A+轮融资规模超亿元，由华泰紫金领投，德昌电机、九号资本、自变量机器人跟投，老股东科沃斯、南创投继续追投。作为机器人一体化关节核心零部件厂商，公司预计2026年关节模组产能将达20万套，2027年提升至100万套。同日，墨现科技完成数千万元产业轮融资，投资方包括自变量机器人、松山湖产投等。墨现科技是国内柔性触觉传感器领域厂商，产品覆盖触觉传感器与数据采集手套，本轮资金将用于新一代数采手套研发、供应链优化及量产能力建设。两笔融资的投资方结构高度相似：产业资本占据主导，既有科沃斯、德昌电机、自变量机器人等产业链上下游企业，也有地方产业基金参与。
Juxie Intelligent's Series A+ round exceeded 100 million yuan, led by Huatai Zijin, with Johnson Electric, Ninebot Capital, and AgiBot participating, and existing shareholders Ecovacs and Nancheng Venture Capital continuing to invest. As a robot integrated joint core component manufacturer, the company expects joint module capacity to reach 200,000 units in 2026, rising to 1 million in 2027. On the same day, Moxian Technology completed a tens-of-millions-yuan industrial funding round, with investors including AgiBot and Songshan Lake Industrial Investment. Moxian Technology is a domestic flexible tactile sensor manufacturer, with products covering tactile sensors and data collection gloves. This round's funds will be used for next-generation data collection glove R&D, supply chain optimization, and mass production capacity building. The investor structures of the two rounds are highly similar: industrial capital dominates, including upstream and downstream companies like Ecovacs, Johnson Electric, and AgiBot, as well as local industrial funds.
自变量同时押注驱动与感知两大上游环节，是当前具身智能产业投资热潮的一个缩影。从整体布局来看，头部企业的对外投资主要围绕供应链绑定和数据协同两大主线展开。供应链层面，在行业集中攻坚量产落地的背景下，整机企业正通过投资锁定核心零部件产能。除自变量机器人布局巨蟹智能、墨现科技外，科沃斯连续两轮跟投巨蟹智能；智元机器人已投资孵化15家产业链上下游早期项目；灵心巧手通过直接投资与LP出资双重方式布局产业链；星动纪元投资潜界科技，补齐上游感知技术短板。智元机器人邓泰华早前透露，智元从投早投小的布局中取得8倍的投资年化收益率。但智元并不以财务投资为目标，而是想借此构建起具身智能产业链的生态。
AgiBot betting on both actuation and perception upstream links simultaneously is a microcosm of the current embodied intelligence industrial investment boom. From an overall布局 perspective, leading companies' external investments mainly revolve around two main lines: supply chain binding and data collaboration. On the supply chain side, against the backdrop of the industry concentrating on mass production deployment, complete machine companies are locking in core component capacity through investment. Besides AgiBot布局 Juxie Intelligent and Moxian Technology, Ecovacs has participated in Juxie Intelligent's two consecutive rounds; AgiBot has invested in and incubated 15 upstream and downstream early-stage projects; Linker Hand布局 the supply chain through both direct investment and LP contributions; Xingdong Era invested in Qianjie Technology to fill upstream perception technology gaps. AgiBot's Deng Taihua earlier revealed that AgiBot achieved an 8x annualized investment return from its "invest early, invest small"布局. But AgiBot does not aim for financial investment—it wants to build an embodied intelligence supply chain ecosystem.
数据协同层面，真实物理交互数据是当前具身大模型训练较缺的一环。今年8月，鹿明机器人、觅蜂科技、自变量机器人等5家头部企业联合投资AI数据公司恺望数据，金额达数亿元。同期，银河通用联合百度入股数据采集方案商简智机器人。星海图也投资了德塔智能、简智机器人等多家数据采集公司。此外，头部企业不再满足于单点投资，开始通过产业基金成批投资上下游公司。2026年6月，星海图联合凯辉基金推出“星途计划”，计划5年投入10亿元，投资100家具身智能领域上下游创业公司，覆盖数据、零部件、场景应用等全链条。灵心巧手联合开创电气设立1亿元产业基金，优先投向具身智能、核心关节模组、智能装备及产业链上下游。
On the data collaboration side, real physical interaction data is a relatively scarce link in current embodied foundation model training. In August this year, five leading companies including Luming Robotics, Mifeng Technology, and AgiBot jointly invested in AI data company Kaiwang Data, with an amount reaching hundreds of millions of yuan. In the same period, Galbot partnered with Baidu to invest in data collection solution provider Jianzhi Robotics. Galaxea also invested in multiple data collection companies including Deta Intelligent and Jianzhi Robotics. Moreover, leading companies are no longer satisfied with single-point investments, beginning to invest in upstream and downstream companies in batches through industrial funds. In June 2026, Galaxea partnered with Cathay Capital to launch the "Star Journey Plan," planning to invest 1 billion yuan over 5 years in 100 upstream and downstream startups in embodied intelligence, covering the full chain of data, components, and scenario applications. Linker Hand partnered with Kaichuang Electric to establish a 100-million-yuan industrial fund, prioritizing embodied intelligence, core joint modules, intelligent equipment, and supply chain upstream and downstream.
鹿明机器人相关负责人告诉《科创板日报》记者，上游战略投资的意义在于深化与核心供应链伙伴的协同，在联合开发、技术验证和产能规划等环节建立更紧密的合作关系。投资能够强化长期合作的基础，但供应链竞争力最终仍要依靠技术能力、工程经验、质量体系和交付表现来建立。这股本体企业做投资的热潮，真实战略需要更多，还是市场造势更多？水木资本董事长唐劲草向《科创板日报》记者表示，两种成分都有，但到2026年这个节点，真实战略需要的成分已经多于市场造势。
A Luming Robotics executive told Science and Technology Innovation Board Daily that the significance of upstream strategic investment lies in deepening collaboration with core supply chain partners, establishing closer cooperative relationships in joint development, technical validation, and capacity planning. Investment can strengthen the foundation for long-term cooperation, but supply chain competitiveness ultimately must be built on technical capability, engineering experience, quality systems, and delivery performance. Is this wave of body companies doing investment more about real strategic need or market hype? Shuimu Capital Chairman Tang Jincao told Science and Technology Innovation Board Daily that both elements exist, but by 2026, the real strategic need component already exceeds market hype.
“整机厂天然面临一个矛盾：融资窗口在缩短，但供应链又没成熟到可以完全市场化采购，核心零部件供应还没稳定，各家都在抢。市场上我们看到有对外投资的具身企业里，17家是做本体整机的，占比将近六成。这种情况下，花几千万锁住一个灵巧手或者关节电机的团队，比什么都重要。”唐劲草说，“如果只是想造势，没必要投这么多笔、这么早，有几笔PR性质的联合投资就够了，部分项目二十三个月投三十七笔的密度，显然是在做供应链纵深布局，不是做给市场看的。”在他看来，不排除部分年轻企业有跟风性质，尤其是成立不到一年就出手投资的，决策质量存疑，但整体趋势上，对外投资已经从可选项变成了具身智能企业的必选项。
"Complete machine companies inherently face a contradiction: funding windows are shortening, but the supply chain is not mature enough for fully market-based procurement, core component supply is not yet stable, and everyone is grabbing. In the market, we see that among embodied companies with external investments, 17 are body complete machine makers, nearly 60%. In this situation, spending tens of millions to lock in a dexterous hand or joint motor team is more important than anything." Tang Jincao said, "If it were just for hype, there would be no need to invest so many times, so early. A few PR-nature joint investments would suffice. Some projects with 37 investments in 23 months are clearly doing deep supply chain布局, not performing for the market." In his view, it is possible some young companies are following the trend, especially those investing within a year of founding, where decision quality is questionable, but overall, external investment has moved from optional to mandatory for embodied intelligence companies.
产业资本密集下场，直接改变了早期项目的定价方式。唐劲草坦言，以前看零部件公司，主要看技术参数、工艺成熟度、客户验证进度，估值有相对稳定的锚。现在如果一家电机或灵巧手公司拿了本体厂商的钱，估值里就多了一层战略背书的溢价，而且这个溢价不好量化。产业资本对估值的容忍度比财务投资机构高，“财务投资人在谈判里的位置变被动了，信息上也比产业方慢一拍。”对于边融边投模式的风险，唐劲草认为，最大的风险在于押错技术路线，以及被自己的投资绑住手脚。“这些本体公司投上游，本质上是在用自己的资金给某一条技术路线投票。比如说某家投了A方案的关节模组，结果半年后发现B方案在批量和寿命上反超了，那这个投资就变成了包袱，反而让它自己在供应链选择上动作变形。”
Industrial capital entering densely directly changes how early-stage projects are priced. Tang Jincao admitted that previously, looking at component companies, one mainly looked at technical parameters, process maturity, and customer validation progress, with valuation having a relatively stable anchor. Now if a motor or dexterous hand company takes money from a body manufacturer, valuation gains an additional layer of strategic endorsement premium, and this premium is hard to quantify. Industrial capital has higher tolerance for valuation than financial investment institutions. "Financial investors' position in negotiations has become passive, and they are one step slower than industrial parties in information." Regarding the risks of the funding-while-investing model, Tang Jincao believes the biggest risk lies in betting on the wrong technology route and being tied down by one's own investment. "These body companies investing upstream are essentially using their own funds to vote for a certain technology route. For example, if one invests in a Plan A joint module, then six months later finds Plan B has overtaken in batch and lifespan, that investment becomes a burden, actually distorting its own supply chain choices."
对于联合投资模式下的竞争与协同问题，鹿明机器人相关负责人对《科创板日报》记者表示，推动具身智能产业落地从来不是单一企业的单打独斗，需要场景、数据、模型、本体等各环节深度协同。“我们坚持核心技术自主研发，对于产业链上能形成互补、具备长期战略价值的环节，会通过战略合作建立更紧密的协同关系。”鹿明方面称，其自研的Lumos NexCore产业具身技能生产系统，打通了数据获取、模型训练、技能评测、部署运行等关键流程，这套体系要发挥价值，既需要自身全栈技术能力，也需要产业链伙伴的配合。“产业协同的价值是共同提升产业链整体效率，各家企业可以在产品定义、模型研发、客户服务上发挥差异化竞争能力，二者并不矛盾。”2026年被普遍视为人形机器人量产元年。只靠自主研发，速度已经跟不上竞争节奏。通过对外投资、合资孵化、产业基金等方式，快速锁定供应链、技术和场景资源，头部企业正用这种方式拉开差距。
Regarding competition and collaboration under the joint investment model, a Luming Robotics executive told Science and Technology Innovation Board Daily that advancing embodied intelligence industrial deployment has never been a single company's solo effort—it requires deep collaboration across scenarios, data, models, and bodies. "We insist on independent R&D of core technologies. For links in the supply chain that can form complementarity and have long-term strategic value, we will establish closer collaborative relationships through strategic cooperation." Luming said its self-developed Lumos NexCore industrial embodied skill production system connects key processes including data acquisition, model training, skill evaluation, and deployment operation. For this system to deliver value, it needs both its own full-stack technical capability and supply chain partners' cooperation. "The value of industrial collaboration is jointly improving overall supply chain efficiency. Each company can exert differentiated competitive capabilities in product definition, model R&D, and customer service—the two are not contradictory." 2026 is widely regarded as the first year of humanoid robot mass production. Relying only on independent R&D, speed can no longer keep up with competition pace. Through external investment, joint venture incubation, and industrial funds to quickly lock in supply chain, technology, and scenario resources, leading companies are using this approach to widen the gap.
