广告
加载中

千诀科技完成Pre-A+轮融资 已与多家互联网科技企业达成合作|产业融资快报

亿邦动力 2025-12-22 10:19
亿邦动力 2025/12/22 10:19

邦小白快读

EN
全文速览

千诀科技融资与公司背景

1.公司完成Pre-A+轮融资,由钧山投资、祥峰投资和石溪资本联合注资,累计融资金额达数亿元,资金将用于核心技术演进、产品标准化及产业化交付能力提升。

2.公司成立于2023年,孵化自清华大学类脑中心,技术团队由清华大学自动化系硕博士组成,在人工智能和机器人领域有多年一线研发经验。

技术亮点和特点

1.“具身大脑”系统强调多模态实时感知、持续任务规划和不依赖预设的自主执行,与Physical Intelligence的π0.5模型理念一致,具备“功能迁移”能力,即在未训练场景中完成复杂任务。

2.系统特点包括跨环境适应性、无预设运行、超长时程工作(可持续数小时),适配二十余种硬件形态如自主家庭保姆机器人,现场演示上百场,最长超2小时。

商业化应用

1.系统已在家庭服务、物流配送、商业运营等场景稳定运行,与多家头部具身机器人厂商、消费电子公司及互联网科技企业合作,加速产业化落地。

产品研发与技术创新

1.千诀科技自主研发“具身大脑”系统,采用类脑计算技术路径,打通感知-推理-行为闭环,实现机器人从被动执行向主动规划的质变,推动产品智能化升级。

2.技术方向包括机器人系统设计、大规模模仿强化学习、类脑智能与生物计算等,提供研发参考点,如仿真-现实迁移学习和双臂控制多模态MPC。

消费趋势与用户行为观察

1.系统应用在家庭服务、物流配送、商业运营等场景,显示智能机器人满足新兴消费需求,如自主家庭保姆机器人可替代人力,观察用户指令下的长时程自主决策能力。

2.与消费电子公司和互联网科技企业合作,揭示用户行为在非结构化环境中的互动趋势,品牌可借此优化产品设计以契合市场。

品牌合作机会

1.与头部企业达成合作,提供品牌渠道建设和营销机会,如通过技术集成提升品牌溢价。

增长市场与机会提示

1.千诀科技在家庭服务、物流配送、商业运营等场景实现稳定运行,显示消费需求变化带来的新机会,如智能物流市场增长潜力大,系统功能迁移能力解决未训练场景任务,降低风险。

2.融资用于产业化交付提升,预示正面影响,卖家可把握具身智能风口,探索新兴业务领域。

合作方式与可学习点

1.与多家头部具身机器人厂商、消费电子公司及互联网科技企业合作,模式包括技术供应或产品集成,卖家可借鉴此合作框架。

2.技术路径如大规模模仿强化学习提供可学习点,卖家可应用类似方法优化事件应对策略。

最新商业模式与扶持启示

1.系统加速产业化落地,代表从技术研发到商业应用的模式转变,融资事件显示行业扶持趋势。

产品生产与设计需求

1.“具身大脑”系统可适配二十余种具身本体硬件形态,工厂需关注多形态兼容设计需求,如机器人系统设计和双臂控制技术。

2.技术如2D-3D语义视觉SLAM算法提供生产优化方向,强调仿真-现实迁移学习以提升制造精度。

商业机会与数字化启示

1.与头部具身机器人厂商合作,工厂可寻求零部件供应或代工机会,参与产业链。

2.类脑计算技术路径启示推进数字化,通过感知-推理-行为闭环提升生产效率,如应用在电商化生产流程中。

新技术与行业趋势

1.千诀科技采用大规模模仿强化学习、类脑智能与生物计算、仿真-现实迁移学习等新技术,显示行业向智能化、自主决策演进趋势。

2.系统打通感知-推理-行为闭环,实现机器人智能中枢,代表技术前沿发展方向。

客户痛点与解决方案

1.客户痛点如非结构化环境中的复杂任务执行能力不足,系统通过无预设、跨环境特点提供解决方案,支持超长时程决策。

2.全流程自主决策能力(覆盖识别、判断、路径生成)缓解操作风险,如物流配送场景中的高效执行痛点。

商业对平台需求与问题

1.具身大脑系统用于物流配送、商业运营等平台场景,显示平台需集成自主决策能力以解决非结构化环境操作问题。

2.系统跨环境适应性缓解平台现实挑战,如减少人为干预需求。

平台最新做法与招商机会

1.千诀科技进行开放场景演示,展现全流程能力,平台可借鉴以优化运营管理,如通过现场演示增强服务吸引力。

2.与多家互联网科技企业合作,提供平台招商机会,平台商可整合此技术提升风向规避策略。

产业新动向与商业模式

1.千诀科技融资事件加速产业化落地,显示具身智能领域蓬勃发展,商业模式包括技术授权和多场景合作应用。

2.在家庭服务、物流配送、商业运营等场景稳定运行,探索新兴商业路径,如主动规划机器人系统。

新问题与政策法规启示

1.新问题如长时程自主决策稳定性挑战,系统演示中已能无预设运行,提供研究点如神经推理机制。

2.首席技术顾问陈峰教授参与中国脑计划重大课题,启示政策支持方向,研究者可分析法规对产业发展的影响。

返回默认

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

Qianjue Technology secures Pre-A+ round funding from Junshan Capital, Vertex Ventures, and Shixi Capital, bringing total funding to hundreds of millions of RMB. The capital will be allocated to advancing core technologies, product standardization, and industrial delivery capabilities.

Founded in 2023 and incubated by Tsinghua University's Brain-Inspired Computing Center, the company's R&D team consists of Tsinghua automation department graduates with extensive experience in AI and robotics.

The company's 'Embodied Brain' system enables multimodal real-time perception, continuous task planning, and autonomous execution without pre-programming—similar to Physical Intelligence's π0.5 model. It demonstrates functional transfer, performing complex tasks in untrained scenarios.

Key features include cross-environment adaptability, preset-free operation, and multi-hour continuous operation. The system has been adapted to over 20 hardware types, including home service robots, with live demos lasting over 2 hours.

Commercial applications span home services, logistics, and business operations. Qianjue collaborates with leading robotics manufacturers, consumer electronics firms, and tech companies to accelerate industrial deployment.

Qianjue Technology's self-developed 'Embodied Brain' system adopts a brain-inspired computing approach, integrating perception, reasoning, and action into a closed loop. This enables robots to shift from passive execution to active planning, driving intelligent product upgrades.

Technical focuses include robotic system design, large-scale imitation reinforcement learning, and brain-inspired intelligence, offering R&D reference points such as simulation-to-reality transfer learning and dual-arm multimodal MPC control.

System applications in home services, logistics, and business operations reveal emerging consumer trends—for instance, autonomous home robots replacing manual labor while demonstrating long-term decision-making capabilities under user commands.

Collaborations with consumer electronics and internet firms highlight user interaction patterns in unstructured environments, providing brands insights to optimize product design.

Partnership opportunities exist with industry leaders to enhance brand value through technology integration and co-marketing initiatives.

Qianjue's stable system performance across home services, logistics, and business operations signals new market opportunities driven by shifting consumer demands. The system's functional transfer capability reduces risks in untrained scenarios, particularly in high-growth areas like smart logistics.

Recent funding for industrial delivery enhancement indicates positive industry momentum, suggesting sellers explore embodied intelligence as an emerging business frontier.

Collaboration models with robotics manufacturers and tech companies—ranging from technology supply to product integration—offer replicable frameworks for sellers. Technical approaches like large-scale imitation reinforcement learning can be adapted to optimize operational strategies.

The accelerated industrialization of embodied intelligence reflects a shift from R&D to commercial application, with funding trends indicating sustained industry support.

The 'Embodied Brain' system supports over 20 hardware configurations, necessitating factory focus on multi-form compatibility in areas like robotic system design and dual-arm control technologies.

Technologies such as 2D-3D semantic visual SLAM algorithms and simulation-to-reality transfer learning offer pathways to enhance manufacturing precision.

Partnerships with leading robotics manufacturers create opportunities for factories to supply components or provide OEM services within the supply chain.

The brain-inspired computing approach—emphasizing perception-reasoning-action loops—can digitize production processes, such as e-commerce-oriented manufacturing workflows.

Qianjue's adoption of large-scale imitation reinforcement learning, brain-inspired computing, and simulation-to-reality transfer learning signals an industry shift toward intelligent, autonomous systems.

The system's closed-loop integration of perception, reasoning, and action represents a cutting-edge technical architecture for robotic intelligence.

It addresses client pain points in unstructured environments by enabling preset-free, cross-context operation with extended decision-making endurance.

Full-cycle autonomy—covering recognition, judgment, and path generation—mitigates operational risks in scenarios like logistics delivery.

The Embodied Brain's applications in logistics and business operations highlight platform needs for autonomous decision-making in unstructured environments.

Its cross-environment adaptability reduces reliance on human intervention, addressing real-world platform challenges.

Qianjue's open-scene demonstrations showcase end-to-end capabilities, offering platforms models to enhance operational management and service appeal.

Collaborations with internet technology firms present partnership opportunities for platforms to integrate advanced risk-mitigation strategies.

Qianjue's funding round accelerates industrial adoption, reflecting robust growth in embodied intelligence with business models spanning technology licensing and multi-scenario applications.

Stable performance in home services, logistics, and commercial operations explores new commercial pathways for proactive planning systems.

Research challenges include long-term autonomous decision stability, with the system's preset-free operation offering study avenues like neural reasoning mechanisms.

CTO advisor Professor Chen Feng's involvement in China's Brain Project highlights policy support directions, inviting analysis of regulatory impacts on industry development.

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.

【亿邦原创】日前,具身大脑公司北京千诀科技有限公司(下称“千诀科技”)完成Pre-A+轮融资,本轮由钧山投资、祥峰投资和石溪资本投资,累计融资金额达数亿元。本轮融资将主要用于核心技术演进、产品标准化以及产业化交付能力的提升。

千诀科技成立于2023年,孵化自清华大学类脑中心。技术团队主体由清华大学自动化系硕博士组成,在人工智能和机器人领域有多年一线科学技术研发和高难度技术项目攻关经历。擅长技术方向包括:机器人系统设计、大规模模仿强化学习、类脑智能与生物计算、仿真-现实迁移学习、2D-3D语义视觉SLAM算法、双臂控制多模态MPC等。

创始团队层面,高海川为创始人兼CEO;2018年开始作为组长,带领类脑双臂机器人团队,从0到1设计多款双臂自主决策机器人。首席技术顾问陈峰教授为中国首批类脑研究专家,承担多项中国脑计划重大课题。

由千诀科技自主研发的“具身大脑”系统,强调通过多模态实时感知、持续任务规划和不依赖预设策略的自主执行,在理念与Physical Intelligence近期发布的π0.5模型高度一致。后者以“功能迁移”为核心,即在未训练场景中依然完成复杂任务的能力。

实际测试过程中,千诀科技具身大脑具有跨环境、无预设、长时程等特点,具备现场演示与落地应用能力,对比π0.5模型所搭建的家庭服务机器人场景更复杂,并可进行全开放现场演示。当参观人员下达指令后,该系统的超长时程的自主决策互动能力,可适配二十余种具身本体硬件形态;以自主家庭保姆机器人为例,可持续运行数小时直至电量耗尽。

基于千诀科技所采用的“类脑计算”技术路径,在构建决策大模型的同时,可以打通感知-推理-行为的系统闭环,形成机器人智能中枢,实现机器人从被动执行向主动规划的质变。

据官方消息,千诀科技将发布新一轮业务侧演示,现场展现其系统在开放式、非结构化环境下的全流程自主决策能力,覆盖识别、判断、路径生成、任务执行等多个维度,形成完整的“感知-决策-行动”闭环。在过去,千诀科技具身大脑累计现场演示达到上百场,最长演示时长在2小时以上。

创始人兼CEO高海川表示:“我们所展示的能产品化的技术,不是单纯地训练VLA模型,而是用类脑启发的神经推理机制重构了机器人决策的输入机制,让机器人仅依靠自身传感器进行长时程工作、真正拥有了任务自主性。在超长时程的开放场景演示中,我们的系统已能稳定运行、持续决策,不依赖人为指令触发而直接工作在真实非结构化环境中。”

商业化进度方面,千诀科技具身大脑已在家庭服务、物流配送、商业运营等多个场景实现稳定运行,与多家头部具身机器人厂商、消费电子公司及互联网科技企业达成合作,通用型具身智能的产业化落地加速。

文章来源:亿邦动力

广告
微信
朋友圈

这么好看,分享一下?

朋友圈 分享

APP内打开

+1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0