• 亿邦公众号
    • APP下载

累计近10亿 一家清华系“具身智能”又融资了

作者:陈美 文章来源:投中网

2026-05-12 09:01

作者:陈美 文章来源:投中网

2026-05-12 09:01

邦小白快读

EN
全文速览

本文核心报道鹿明机器人融资近10亿元,聚焦具身智能在工业场景的真实价值落地。

1.融资详情:鹿明机器人完成A1和A2轮融资,由三菱电机连续领投,其他机构跟投,资金用于具身智能大模型研发和多场景商业化,助力产业规模化落地。

2.技术亮点:自研Lumos Nexcore物理AI引擎,融合世界模型等技术,提升动作精度和安全性;通过无本体数采技术构建数据飞轮,加速算法迭代;硬件全栈自研如LUS和MOS机器人,优化成本控制。

3.应用案例:与三菱电机合作设立联合实验室,部署MOS机器人进行PLC产线AI质检,包括自主识别和瑕疵比对,提升效率;行业趋势显示2026年将迎来具身智能量产和规模化卡位战。

4.实用建议:绑定产业方拿真实订单已成融资标配,纯技术驱动团队易被淘汰,企业需以客户为中心,解决产业痛点。

鹿明机器人品牌战略以工业具身智能为核心,强调价值驱动和产业合作。

1.品牌营销:通过融资新闻提升影响力,三菱电机等产业巨头连续投资背书,增强品牌可信度;合作案例如与中远海运等伙伴拓展应用场景,传递可靠品牌形象。

2.产品研发:自研物理AI引擎和硬件产品(如LUS全尺寸人形机器人),内置工业视觉注意力优化,降低BOM成本,提升产品性价比和竞争力。

3.消费趋势:消费需求转向工业场景真价值,避免To C闭环难问题;用户行为表现为对高精度、安全性的偏好,例如在工厂部署机器人进行质检操作。

4.合作方式:与三菱电机深度绑定,共享资源如测试平台,实现双赢;品牌渠道建设体现在与广运集团等产业伙伴合作,加速千行百业渗透。

行业机会在于绑定产业方和规模落地,需应对分化竞争风险。

1.增长市场:2026年具身智能将进入量产和规模化阶段,工业场景如质检需求旺盛;案例中鹿明与三菱合作部署机器人,提升效率10-20%,显示市场潜力。

2.机会提示:绑定头部产业方拿真实订单成融资标配,可学习鹿明模式,与中远海运等合作拓展业务;扶持政策如三菱提供自动化技术资源,降低部署成本。

3.风险提示:竞争加剧导致梯队分化,第二队聚焦细分场景跑通闭环;正面影响是技术迭代快,负面影响是纯技术团队难获融资,需完整闭环能力。

4.事件应对:以客户为中心响应需求,如鹿明从产业痛点出发,扩展场景迁移能力;最新商业模式是数据驱动底座,需强调泛工业应用。

具身智能提供新生产工具,启示数字化升级和效率优化。

1.产品生产需求:工业场景需高精度运动控制机器人,如鹿明MOS重载轮臂式机器人用于产线外观检测,实现六面无盲区瑕疵比对,提升质检准确率。

2.商业机会:部署具身智能可降低边际成本,案例中三菱电机产线试点,机器人自主抓取产品,优化效率;与鹿明合作可获得联合实验室资源,解决协同作业难题。

3.推进数字化启示:无本体数采技术构建数据飞轮,加速模型泛化,减少依赖本体数据;通过硬件自研控制BOM成本,启示工厂采用低成本数字方案落地电商需求。

行业转向数据驱动解决方案,新技术解决工业痛点。

1.行业发展趋势:具身智能2026年量产卡位战开启,竞争分化,第一梯队企业融资超百亿;服务机会在数据飞轮之争绑定产业方,如鹿明与三菱合作模式。

2.新技术:鹿明自研Lumos Nexcore引擎,融合VLA联合训练和延迟晚融合架构,提升复杂场景适配;MOE专家网络和工业视觉优化新技术,增强安全性与精度。

3.客户痛点:工业客户需高效、安全、通用解决方案,痛点如高精度控制难;鹿明通过完整闭环(数据基建、引擎、硬件、方案)提供应对,实例在质检应用提升准确性。

4.解决方案:针对泛工业场景提供迁移能力,如从抓取技能扩展到商业服务;合作方式包括设立实验室共享资源,满足客户数据需求。

商业需求推动平台化合作,需聚焦落地和数据管理。

1.商业对平台需求:产业方需场景定义平台,如三菱电机与鹿明共建联合实验室,开放测试平台,解决协同作业和技术难题;平台招商需强调真实订单,满足产业转型升级需求。

2.平台的最新做法:部署机器人试点(如PLC产线应用),整合数据飞轮,例如通过无本体数采加速迭代;运营管理体现在成本控制优化边际成本,提升平台效率。

3.风向规避:竞争加剧分化市场,平台应锚定落地场景,避免纯技术驱动,学习鹿明绑定头部产业方策略;风险是未绑定方易被淘汰,机会在泛工业领域拓展。

产业进入卡位阶段,商业模式需完整闭环。

1.产业新动向:2026年具身智能规模化落地年,竞争分化,第一梯队融资超百亿;案例显示三菱电机FA业务营收占比13.5%,推动工业自动化需求。

2.新问题:如何从工业场景迁移通用能力至商业服务或家庭,如鹿明沉淀的操作技能扩展;政策启示是产业法规需支持数据飞轮创新,避免技术壁垒。

3.商业模式:强调以客户为中心的绑定产业方模式,拿订单成标配;完整闭环能力(数据、引擎、硬件、方案)是关键,减少对单一技术依赖。

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

This article covers Luming Robotics' nearly 1 billion RMB in new funding, focused on delivering real-world value from embodied intelligence in industrial scenarios.

1. Funding details: Luming Robotics has closed its Series A1 and A2 rounds, with Mitsubishi Electric leading both rounds consecutively, followed by other participating investors. The capital will be used for R&D on large embodied intelligence models and multi-scenario commercialization to support large-scale industrial adoption.

2. Technical strengths: The company developed its proprietary Lumos Nexcore physical AI engine, which integrates world modeling technology to improve motion precision and safety. Its ontology-agnostic data collection technology builds a data flywheel to accelerate algorithm iteration, and full-stack in-house hardware development (such as its LUS and MOS robots) optimizes cost control.

3. Use cases: It established a joint lab with Mitsubishi Electric, where MOS robots have been deployed for AI-powered quality inspection on PLC production lines, handling autonomous defect identification and comparison to boost efficiency. Industry forecasts project that 2026 will see mass production of embodied AI solutions and a race for large-scale market positioning.

4. Key takeaway: Securing real orders through partnerships with established industry players has become a standard requirement for fundraising; purely technology-driven teams are at high risk of being eliminated. Companies must prioritize client-centricity and focus on solving core industrial pain points.

Luming Robotics' brand strategy centers on industrial embodied intelligence, with a focus on value delivery and industrial collaboration.

1. Brand building: Its fundraising news has boosted brand visibility, and consecutive lead investments from industry giant Mitsubishi Electric serve as strong endorsement to enhance brand credibility. Partnerships with players like COSCO Shipping to expand application scenarios reinforce its image as a reliable industry player.

2. Product R&D: In-house development of its physical AI engine and hardware products (such as the LUS full-size humanoid robot) includes optimized industrial visual attention to reduce BOM costs, improving product cost-performance and competitiveness.

3. Market trends: Demand is shifting toward delivering real value in industrial scenarios, which avoids the difficulty of building closed-loop operations that plagues many consumer-facing (To C) projects. End users prioritize high precision and safety, as seen in the adoption of inspection robots on factory floors.

4. Collaboration approach: Deep strategic ties with Mitsubishi Electric enable resource sharing such as access to testing platforms for mutual benefit. Partnerships with industry players including Guangyun Group support channel expansion to accelerate penetration across a wide range of industries.

Industry opportunities lie in partnering with established industry players and scaling operations, but companies must navigate growing competitive divergence.

1. Growth market: Embodied intelligence will enter the mass production and large-scale adoption phase by 2026, with strong demand for applications like industrial quality inspection. The case of Luming and Mitsubishi's deployed robots, which delivered 10-20% efficiency gains, demonstrates the market's significant potential.

2. Opportunity insights: Securing real orders through partnerships with leading industry players is now a standard requirement for fundraising. Sellers can learn from Luming's model and expand business through partnerships such as its collaboration with COSCO Shipping. Support from strategic partners like Mitsubishi, which provides access to automation technology resources, also lowers deployment costs.

3. Risk alerts: Intensifying competition is driving market stratification, with second-tier players focusing on running closed-loop operations in niche segments. While this accelerates overall technological iteration, purely technology-focused teams struggle to secure funding, and full closed-loop operational capabilities have become a requirement to compete.

4. Strategic response: Companies must adopt a client-centric approach to respond to market demand, just as Luming built its business around industrial pain points and expanded its cross-scenario transfer capability. The emerging business model centers on a data-driven infrastructure layer, with a focus on general-purpose applications across broad industrial sectors.

Embodied intelligence introduces new production tools that offer insights for digital transformation and efficiency optimization.

1. Production requirements: Industrial scenarios require robots with high-precision motion control. For example, Luming's MOS heavy-duty wheeled arm robot conducts six-sided full-coverage defect comparison for production line visual inspection, improving inspection accuracy.

2. Business opportunities: Deploying embodied intelligence reduces marginal production costs, as demonstrated in Mitsubishi Electric's pilot production line, where robots autonomously grasp products to streamline operations. Partnering with Luming also provides access to joint lab resources to solve collaborative operation challenges.

3. Insights for digital transformation: Ontology-agnostic data collection builds a data flywheel that accelerates model generalization and reduces dependence on ontology-specific data. In-house hardware development to control BOM costs also offers a lesson for factories: low-cost digital solutions can successfully meet e-commerce-driven production demands.

The industry is shifting toward data-driven solutions, with new technologies addressing longstanding industrial pain points.

1. Industry trends: The race for mass production and market positioning in embodied intelligence will kick off in 2026, bringing increased market stratification; total fundraising for first-tier players already exceeds 10 billion RMB. Service opportunities lie in the competition to build data flywheels through industrial partnerships, as seen in the Luming-Mitsubishi collaboration model.

2. New technologies: Luming's proprietary Lumos Nexcore engine integrates VLA joint training and delayed late-fusion architecture to improve adaptability to complex scenarios. New technical optimizations including MOE expert networks and improved industrial vision enhance both safety and operational precision.

3. Customer pain points: Industrial customers require efficient, safe, and general-purpose solutions, and high-precision motion control remains a core unmet need. Luming addresses this through a full closed-loop offering covering data infrastructure, AI engine, hardware, and end solutions, which has delivered improved accuracy in quality inspection applications.

4. Solution offerings: The company provides cross-scenario transfer capability for general industrial use cases, for example expanding from grasping skills to commercial service applications. Collaborative models such as joint labs enable resource sharing to meet customers' data-related needs.

Growing commercial demand is driving platform-based collaboration, with a focus on real-world adoption and data management.

1. Platform demand from industry: Industrial players require scenario-definition platforms. For example, Mitsubishi Electric and Luming co-founded a joint lab with an open testing platform to solve collaborative operation and technical challenges. Platform recruitment should prioritize opportunities tied to real orders to meet the demand for industrial transformation and upgrading.

2. Latest platform practices: Platforms can run robot pilots (such as the PLC production line use case) and integrate data flywheels, using ontology-agnostic data collection to accelerate model iteration. Operational optimization to control costs and reduce marginal costs also improves overall platform efficiency.

3. Risk mitigation: Intensifying competition is fragmenting the market, so platforms should anchor themselves in落地场景, avoid backing purely technology-driven projects, and follow Luming's example by partnering with leading industry players. Unaligned players face a high risk of elimination, while the biggest opportunity lies in expansion across broad industrial sectors.

The industry is entering a market positioning phase, where full closed-loop business models are required to compete.

1. New industry developments: 2026 will be the year of large-scale落地 of embodied intelligence, with growing competitive divergence; total fundraising for first-tier players already exceeds 10 billion RMB. The case study notes that FA (Factory Automation) accounts for 13.5% of Mitsubishi Electric's total revenue, driving growing demand for industrial automation.

2. Emerging research questions: It remains unclear how to transfer general capabilities built in industrial scenarios to commercial services or household applications, an extension Luming is pursuing by refining its accumulated operational skills. For policymakers, this highlights the need for industry regulations to support data flywheel innovation and prevent unnecessary technological barriers.

3. Business model insights: The client-centric, industry partnership-focused model, where securing confirmed orders is standard, has become the dominant approach. Full closed-loop capabilities spanning data, engine, hardware and end-to-end solutions are the core requirement for success, reducing reliance on any single technical advantage.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

今日,投中网独家获悉,主打工业与物流落地的鹿明机器人,连续完成A1及A2融资。具体来看,A1轮由三菱电机智能制造科技(中国)集团有限公司(以下简称“三菱电机”)领投,老股东普华资本、吴中金控等机构超额跟投;A2轮仍由三菱电机持续领投,恒生电子、海高集团、昆石投资等机构跟投。

自此,两轮融资完成后,鹿明机器人已累计融资近10亿元。据悉,新一轮连续融资将用于具身智能大模型的研发、多场景商业化的拓展,以及更规模化的产业落地。

清华学霸带队,剑指具身智能规模化落地

众所周知,鹿明机器人的创业初心,源于创始人喻超对于“物理世界”的理解,以及对“物理世界AGI”的技术信仰。在通往信仰的路上,喻超表现出了极强的务实主义——持续为具身智能寻找真正的“价值出口”。

“具身智能不能只停留在实验室或视频里,它必须去‘真干活’,产生‘真价值’。”喻超谈到。相比于目前尚难找到商业闭环的To C场景,鹿明机器人将工业场景作为具身智能价值落地的关键出口之一。

面向工业场景,鹿明机器人自研Lumos Nexcore物理AI引擎,融合世界模型与VLA联合训练,内置工业视觉注意力优化,搭配延迟晚融合架构与MOE专家网络,兼顾动作精度、物理安全与复杂场景适配,驱动工业具身智能高效落地。

为了快速跑通工业场景下的ROI,鹿明机器人从底层构建了差异化竞争力——数据侧,率先布局数据基建。 通过“无本体数采”技术,形成“数据-模型”的高效飞轮,提升模型泛化能力、加速算法快速迭代,让机器人快速习得工业级的通用技能。

硬件侧,鹿明机器人选择全栈自研,推出全尺寸人形机器人LUS、重载轮臂式具身智能机器人MOS等产品,不仅在机器人的运动能力、操作精度与环境适应性上进行了全面创新,更通过对BOM成本的极致控制,降低了硬件在产业落地过程中的边际成本。

“我们希望解决一个又一个具体的产业痛点”,喻超表示,当机器人“真干活”的价值被不断验证,具身智能的规模化落地也就水到渠成。

产业资本加注,三菱电机连续两轮领投

在新A1及A2融资中,产业资本成为关键词。尤其是三菱电机,连续两轮领投、持续加码鹿明机器人,此外,恒生电子、海高集团也出现在A2轮投资方名单中。

作为一家老牌电机巨头,三菱电机为何会持续出手?一位投资人告诉我,产业资本的出手,从来就不是“讲故事”,而是在于“技术互补”与“会算账”。

“三菱电机作为全球工业自动化领域巨头,有深度布局智能制造、切入前沿智能技术赛道的必要性。在统筹工业自动化事业的中国子公司成立后,迅速与鹿明机器人达成战略合作,本身反映出该巨头在工业自动化业务上迫切的转型升级需求。”上述投资人分析道。

另一方面,对于鹿明机器人而言,借助这一工业自动化巨头的产业资源与工业场景,鹿明能快速找到具身智能技术产业化的落地路径,实现双赢。

截至目前,双方已在苏州、上海设立联合实验室,并开放鹿明机器人测试平台与三菱电机自动化技术资源,着重攻克高精度运动控制、多机协同作业等技术难题。

与此同时,三菱电机工厂也以PLC产线为试点,部署鹿明MOS重载轮臂式具身智能机器人,实现AI外观检测的柔性产线应用,包括机器人自主识别、抓取产品,以及六面无盲区外观瑕疵比对等任务,以提升质检效率与准确性。

值得一提的是,《财富》2025年世界500强榜单显示,三菱电机以约‌370亿美元的营收,‌排名第‌439位。其中,与鹿明合作的工业自动化(FA)业务,是三菱电机最重要的业务之一。2025财年(截至2026年 3月)显示,三菱电机的FA业务营收占总营收的13.5%,营业利润占比约17.7%。

除了三菱电机,以及新入局的产业资本,投中网获悉,鹿明机器人还与中远海运、德马科技、广运集团等产业伙伴合作,以全面提速通用具身智能在千行百业的应用。

绑定产业方、拿订单,已成融资标配?

回顾具身智能的发展,如果说2025年是“破局”之年,那么2026年便是“量产与规模化落地”之年。

就今年而言,赛道竞争已明显加剧,呈现分化态势:一方面,第一梯队具身智能企业融资较快,估值基本突破百亿级别,并且持续获得资本青睐与产业方的扶持;另一方面,第二梯队企业进一步聚焦下沉细分场景,深耕零部件、数据服务等环节,降低成本,跑通商业闭环。

这其中,不管是哪一梯队的“选手”,都无一例外地将目光锁定在“产业化”上。这从一定程度上讲,谁能真正走进工厂、落地场景,谁就能拥有继续留在牌桌上的话语权。

不知不觉间,一场由“场景定义的数据飞轮之争”悄然爆发。过程中,有投资人透露,绑定头部产业方、拿下真实订单,已成为融资标配,而纯靠技术驱动的研发团队很难获得市场认可,未来将很快被市场淘汰。

作为鹿明机器人的创始人,喻超也认为,“以客户为中心”是鹿明机器人重要的价值观之一,因为只有产业客户才最了解产业真实需求和痛点。

不过,在喻超看来,从产业方需求出发只是第一步,如何扩展和迁移才是最重要的。“这一代具身智能技术与传统机器人及自动化技术相比,最大的区别在于,采取数据驱动的技术底座和通用的硬件方案,以及解决通用场景问题的能力。”

喻超认为,在工业场景中,从最简单的抓取到复杂的工具使用,沉淀下来的自主操作能力,可以向商业服务、家庭等更广阔场景迁移。

在产业落地与数据采集之外,全栈技术的闭环之争也是核心看点。就目前情况来看,单一技术优势已明显不再奏效,数据基建、核心引擎、硬件平台、场景方案的完整闭环能力,成为每个具身智能“玩家”都在强调的事。

喻超表示,鹿明机器人正朝着行业关注的这几个核心方向进行布局。未来,在泛工业领域,鹿明机器人还将继续夯实基本盘,锚定“产业具身”,为机器人在工业场景下规模化落地夯实技术根基根基。

由此看来,2026年,一场具身智能的“产业卡位战”真的来了。

注:文/陈美,文章来源:投中网(公众号ID:China-Venture),本文为作者独立观点,不代表亿邦动力立场。

【版权提示】亿邦动力倡导尊重与保护知识产权。未经许可,任何人不得复制、转载、或以其他方式使用本网站的内容。如发现本站文章存在版权问题,烦请提供版权疑问、身份证明、版权证明、联系方式等发邮件至run@ebrun.com,我们将及时沟通与处理。

已点赞

+1

{{ praise_text }}

好文章,值得被更多人看到

FAQ回顾

鹿明机器人最新一轮融资的用途是什么?

新一轮融资将用于具身智能大模型的研发、多场景商业化的拓展,以及更规模化的产业落地。

三菱电机为何连续领投鹿明机器人?

三菱电机作为工业自动化巨头,旨在通过投资实现技术互补与业务转型升级,双方已在苏州、上海设立联合实验室,并在PLC产线试点部署机器人。

2026年具身智能行业的发展趋势是什么?

2026年被视为具身智能的量产与规模化落地之年,竞争加剧,企业普遍聚焦产业化,绑定产业方、获取真实订单成为融资和生存的关键。

广告

下一篇

征集报名《品牌数字化服务商资源手册》

《品牌数字化服务商资源手册》展示数字化服务商风采,亿邦全年零售大会通发,触达更多资源,马上加入!

马上入驻

电商早报:阿里千问与淘宝全面打通;拼多多618长达41天

阿里千问与淘宝全面打通;拼多多618长达41天;东方甄选首家线下体验店北京试营业

电商早报

2026-05-12 06:43:47