About

关于我们 About Us

商务联系 Business contact@synkrotron.ai
电话 Phone 029-89851182

清科灵境致力于为物理世界构建通用的具身智能引擎。公司由图灵奖得主姚期智院士担任院长的清华交叉信息核心技术研究院孵化,由知名学者杨子江教授创办并出任 CEO,原Uber高级科学家潘余曦博士出任 CTO。公司拥有一支高素质的研发团队,大多来自世界知名高校和世界500强企业,硕博学历占比超过 50%。 Synkrotron is dedicated to building universal embodied intelligence engines for the physical world. Incubated by the Institute for Interdisciplinary Information Core Technology at Tsinghua University, chaired by Turing Award laureate Prof. Andrew Chi-Chih Yao, the company was founded by renowned scholar Prof. Zijiang Yang as CEO, with Dr. Yuxi Pan, former Senior Scientist at Uber, serving as CTO. Our R&D team draws talent from top-tier universities and Fortune 500 companies, with over 50% holding advanced degrees.

我们秉承“让机器理解物理世界”的理念,围绕 World Action Model(世界动作模型)、Dora 具身智能中间件、数据闭环与合成数据开展内部研发。Lingo Brain、Lingo Trace、Lingo Data,以及具身仿真与数据合成技术,是支持模型、数据采集、治理与仿真的内部技术和工具;面向客户的产品与方案请见产品和解决方案栏目。 Guided by our belief in teaching machines the physical world, we develop internal technologies around the World Action Model, Dora middleware, data loops and synthetic data. Lingo Brain, Lingo Trace, Lingo Data, and our embodied simulation and synthetic data technologies support models, data capture, governance and simulation. Customer-facing products and solutions are described in their respective sections.

清科灵境

联系我们 Contact Us

提交咨询需求,或通过电话与邮箱直接联系我们。 Send an enquiry, or contact us directly by phone or email.

告诉我们您的需求 Tell us what you need

请填写合作需求、应用场景或技术问题,我们会与您确认后续沟通方式。 Tell us about your collaboration needs, application scenario, or technical questions, and we will follow up.

咨询类别 Type of consultation

What we believe

我们相信 We believe

三个驱动我们向前的核心信念。 Three convictions that drive us forward.

物理世界是智能的终极考场 The physical world is the ultimate test

真正的智能必须能够理解、预测并作用于物理世界。感知只是起点,操纵与交互才是终点。 True intelligence must understand, predict, and act in the physical world. Perception is only the beginning — manipulation and interaction are the destination.

世界模型是具身智能的基石 World models are the foundation of embodied AI

没有对物理规律的深层理解,机器人只能运行在狭窄的预设场景中。世界模型打破了这一限制。 Without deep understanding of physical laws, robots remain confined to narrow, pre-scripted scenarios. World models break this limitation.

真实场景进化胜过实验室完美 Real-world evolution beats lab perfection

我们从真实部署中获取反馈,以数据飞轮驱动持续学习。每一次失败与纠偏,都是系统进化的燃料。 We harvest feedback from real deployments and drive continuous learning through data flywheels. Every failure and correction fuels system evolution.

Partners

合作伙伴 Partners

已与多家顶尖企业与高校开展深度合作。 Deep collaborations with leading enterprises and universities.