PHYSICAL AI · INDUSTRIAL APPLICATION LAYER

Physical AI 时代的
工业应用层
The Industrial
Application Layer
of Physical AI.

The Industrial Application Layer of the Physical AI Era.

大模型时代,Cursor、Harvey 把 token 变成生产力。物理 AI 时代,我们做同样的事——把基模的 Action,变成真实场景里稳定运行的生产力。整机 + 智能,把具身智能真正用进制造、物流、能源等真实场景。In the LLM era, Cursor and Harvey turned tokens into productivity. In the Physical AI era we do the same — we turn a foundation model's Action into stable, real-world productivity. Embodiment + intelligence, bringing embodied AI into real manufacturing, logistics and energy scenarios.

具身智能应用的先行者 · A pioneer in applied embodied AI
01 · WHAT WE TAKE ON · XJ-01

还有大量作业,
今天只能靠人。
A great deal of work
still depends on people.

The processes fixed machines can't do and general robots can't generalize.

大量作业仍停在"人 + 半自动化设备"协作:多品种混线、频繁换型、依赖判断和手感。固定专机做不了,通用机器人也通用不了。A huge amount of work is still stuck at "humans + semi-automation": mixed models, frequent changeovers, reliance on judgment and feel. Fixed machines can't handle it; general robots can't generalize to it.

大批量标准件交给专机;我们专注传统自动化做不进、必须靠人柔性的那类工序。High-volume standard parts go to dedicated machines; we focus on the processes that need human flexibility, where traditional automation can't reach.

而且不只工业产线。凡是这类"靠人柔性"的场景,都是我们的战场——And not only production lines. Any scenario that still relies on human flexibility is our field —

制造生产Manufacturing

打磨、检测、装配、上下料Grinding, inspection, assembly, loading

物流场站Logistics

分拣、装卸、多机调度Sorting, handling, multi-robot dispatch

能源 · 特种Energy · Special

能源装备、矿洞、数据中心Energy equipment, mining, data centers

02 · FOUNDATIONS TO PHYSICAL AI · XJ-02

把基模能力,
变成产线生产力。
We turn model capability
into production.

We plug into the ecosystem, and own the layer that turns capability into production.

· PLUG IN · 我们接入的生态 ·· PLUG IN · ecosystems we connect to ·
通用基模Foundation modelsVLA / WAM · PI · 智元 · 自变量
机器人本体 · 硬件Robot bodies · hardware成熟本体与标准零部件Mature bodies & standard parts
客户 · 集成商Customers · integrators真实产线与渠道网络Real lines & channel network
· 星炬 OWNS · 我们自研、拥有的那一层 ·· XINGJU OWNS · the layer we build & own ·
垂类场景数据Scenario data真机采集 + 仿真 + 标注Real capture + sim + labeling
真机后训练Real-machine post-trainingRL + 人在环,训到可交付RL + human-in-loop, to deliverable
整机 · 协同控制Integration · coordination多机多设备实时协同Real-time multi-robot control
场景落地 · 真实生产力Deployment · real productivity稳定运行,数据反哺,越用越强Runs, feeds data back, keeps improving

基模越强,应用层越值钱 —— 那层"胶水",把分散的能力,变成能干活的部署。The stronger the base models, the more valuable the application layer — the glue that turns scattered capability into working deployments.

03 · WHAT WE DO · XJ-03

我们能做什么。What we do.

Four capabilities, one integrated stack.

/ 01 · DATASETS
垂类场景数据集Vertical scenario datasets

围绕真实场景做采集、仿真、标注,沉淀"面向具身 AI 训练的真实世界操作数据"——视觉、力反馈、动作轨迹,仿真造不出、别人买不到,可开源共享。We capture, simulate and label around real scenarios to build real-world operation data for embodied-AI training — vision, force feedback, motion trajectories that simulation can't create and others can't buy. Partly open-sourced.

/ 02 · POST-TRAINING
真机强化学习后训练Real-machine RL post-training

不训基模。用 RL + 人在环把通用基模调成场景专用能力,把成功率从"能跑"训到"能交付"——这是落地成败真正发生的地方。We don't train foundation models. With RL + human-in-the-loop we tune a general base model into scenario-specific capability, lifting success rates from "it runs" to "it delivers" — where deployment is truly won or lost.

/ 03 · INTEGRATION
整机集成方案Whole-machine integration

工作站本体、柔性工装、末端执行器,加上多机 · 多设备的实时协同控制(节拍、防干涉、坐标系),组合成一台可交付的整机。Workstation bodies, flexible fixtures and end-effectors, plus real-time multi-robot / multi-device coordination (cycle time, anti-collision, coordinate frames) — assembled into one deliverable machine.

/ 04 · APPLICATION
具身智能应用方案Embodied AI applications

从场景分析、能力拆解到落地交付的完整方案——把具身智能真正用进你的生产,不是给你一个 demo。A complete solution from scenario analysis and capability breakdown to delivery — bringing embodied AI into your production, not just a demo.

真机后训练 · 任务成功率Real-machine post-training · task success rate
未经后训练Before~60%
真机后训练后After95%
从"能跑"到"能交付"From "it runs" to "it delivers"
04 · SOLUTIONS · XJ-04

应用场景。Solutions.

Where our Workers go to work.

05 · DATA ENGINE · XJ-05

面向具身 AI 训练的
真实世界操作数据。
Real-world operation data,
built for embodied-AI training.

Scenario × Data — the flywheel a bought machine never gives you.

有场景Scenario采真机数据Real data后训练Post-training越用越懂你的产线Fits your line better over time
SIG · 01
真机采集Real capture

联合真实场景,采集视觉、力反馈、动作轨迹等一手操作数据。First-hand operation data — vision, force, trajectories — from real scenarios.

SIG · 02
仿真生成Simulation

模块化仿真,低成本扩展长尾工艺与节拍,用于预训练与迁移。Modular sim to cheaply expand long-tail processes, for pretraining and transfer.

SIG · 03
标注与校验Labeling & QA

质量闭环,把原始数据清洗成可直接训练的高质量数据。A quality loop that turns raw data into training-ready data.

SIG · 04
数据集开源Open datasets

沉淀垂类场景数据集,一部分开源共享,连接生态。Vertical datasets, partly open-sourced to connect the ecosystem.

场景和数据两个轮子互相带动:越用越懂你的产线 —— 这是买一台设备得不到的东西。Scenarios and data drive each other: the longer it runs, the better it fits your line — something you never get from buying a machine.

06 · WHY WE CAN DELIVER · XJ-06

这是一项复杂的
系统工程。
This is complex
systems engineering.

It takes four capabilities at once. We have all four.

把具身智能落到真实场景,不是训好一个模型就行 —— 它是一整套系统工程,要同时具备四种能力。纯基模厂只有 AI,纯集成商只有产线;四样同时具备,才做得成。Bringing embodied AI into the real world isn't just training a model — it's a full systems-engineering effort that needs four capabilities at once. A pure model shop has only AI; a pure integrator has only the line. You need all four to pull it off.

场景拆解 · 懂工艺Scenario analysis · process

把每个场景拆到底层能力需求,配出对的方案。Break each scenario down to its capability needs, then design the right solution.

真机后训练 · 懂具身 AIPost-training · embodied AI

RL + 人在环,把通用基模训成场景可交付。RL + human-in-loop to train a base model into deliverable capability.

多机协同控制 · 懂工业自动化Coordination · automation

具身机器人 + 工业设备 + PLC 协同成一条线。Embodied robots + industrial equipment + PLC working as one line.

集成交付 · 懂产线Delivery · the shop floor

现场部署、调试、稳定运行,真正交付到生产。On-site deploy, tune and run stably — delivered into production.

这四种能力的交叉,是最难复制的地方 —— 我们把 AI 与工业产线两边,做进了同一支团队。The intersection of these four is the hardest thing to replicate — we've built both AI and the shop floor into one team.

合心集团既是我们的天使投资方,也是我们的产线客户 —— 一边投,一边下场用。这是工业 Physical AI 公司最理想的早期验证。HEXIN Group is both our angel investor and a production-line customer — investing and using at the same time. The ideal early validation for an industrial Physical AI company.

07 · ABOUT XINGJU · XJ-07

关于星炬。About Xingju.

A Tsinghua-rooted team, focused on real-world scenarios.

星炬智械,是一家专注真实场景的具身智能应用公司Xingju is an embodied-AI application company focused on real-world scenarios.
我们是谁Who we are

把具身智能真正落进一个场景,从来不是单点技术能解决的 —— 它是一整套系统工程:从场景数据采集、真机后训练与技能沉淀,到多机多设备协同控制、模块化整机,再到现场集成落地,每一环都不能缺Bringing embodied AI into a real scenario is never solved by a single technology — it is a full systems-engineering effort: from scenario data capture, real-machine post-training and skill libraries, to multi-robot / multi-device coordination, modular hardware, and on-site deployment. Every link has to be there.

星炬是一支清华系团队,同时具备顶尖 AI 与强化学习算法能力、具身智能研究能力,以及多年产业工程经验。在真实场景的 AI 方案落地上,我们积累了超过十年的经验,也最擅长这件事 —— 一整套系统工程里要补的每一环,都在我们的能力范围之内。Xingju is a Tsinghua-rooted team combining top-tier AI and reinforcement-learning talent, embodied-AI research, and years of industrial engineering. In deploying AI solutions into real-world scenarios we carry over a decade of experience — and it is what we do best. Every link a full systems-engineering effort demands sits within our reach.

凭什么能把一个场景做成Why we can make a scenario work
  • 方法 · 真机后训练Method · real-machine post-training
    不是拿通用模型直接上,而是在真实场景里反复打磨,把任务成功率从 ~60% 做到 95% —— 跨过"从能跑,到能落地"的那道坎。Not deploying a general model as-is, but refining it in the real scene until task success climbs from ~60% to 95% — crossing the gap from "it runs" to "it works on the floor."
  • 验证 · 头部厂商场景Validation · leading-manufacturer scenarios
    已在汽车零部件、3C 精密制造、能源装备等领域头部厂商的真实场景中验证与落地。Validated and deployed in the real scenarios of leading manufacturers across auto parts, precision electronics and energy equipment.
给我们一个场景,我们负责让它在现场落地、干下去Give us a scenario — we make it land on the floor and keep running.
( CONTACT ) — XJ-001
聊聊你要应用
具身智能的场景。
Tell us about a scenario
you'd bring embodied AI into.

凡是"必须靠人的柔性、传统自动化做不进"的场景,把它发给我们,一起看能不能做。Any scenario that still needs human flexibility and can't be automated the old way — send it over and let's see if we can do it.

或直接邮件 Or email us at contact@xingju-ai.com