Joyce Chen
    • Create new note
    • Create a note from template
      • Sharing URL Link copied
      • /edit
      • View mode
        • Edit mode
        • View mode
        • Book mode
        • Slide mode
        Edit mode View mode Book mode Slide mode
      • Customize slides
      • Note Permission
      • Read
        • Only me
        • Signed-in users
        • Everyone
        Only me Signed-in users Everyone
      • Write
        • Only me
        • Signed-in users
        • Everyone
        Only me Signed-in users Everyone
      • Engagement control Commenting, Suggest edit, Emoji Reply
    • Invite by email
      Invitee

      This note has no invitees

    • Publish Note

      Share your work with the world Congratulations! 🎉 Your note is out in the world Publish Note No publishing access yet

      Your note will be visible on your profile and discoverable by anyone.
      Your note is now live.
      This note is visible on your profile and discoverable online.
      Everyone on the web can find and read all notes of this public team.

      Your account was recently created. Publishing will be available soon, allowing you to share notes on your public page and in search results.

      Your team account was recently created. Publishing will be available soon, allowing you to share notes on your public page and in search results.

      Explore these features while you wait
      Complete general settings
      Bookmark and like published notes
      Write a few more notes
      Complete general settings
      Write a few more notes
      See published notes
      Unpublish note
      Please check the box to agree to the Community Guidelines.
      View profile
    • Commenting
      Permission
      Disabled Forbidden Owners Signed-in users Everyone
    • Enable
    • Permission
      • Forbidden
      • Owners
      • Signed-in users
      • Everyone
    • Suggest edit
      Permission
      Disabled Forbidden Owners Signed-in users Everyone
    • Enable
    • Permission
      • Forbidden
      • Owners
      • Signed-in users
    • Emoji Reply
    • Enable
    • Versions and GitHub Sync
    • Note settings
    • Note Insights New
    • Engagement control
    • Make a copy
    • Transfer ownership
    • Delete this note
    • Save as template
    • Insert from template
    • Import from
      • Dropbox
      • Google Drive
      • Gist
      • Clipboard
    • Export to
      • Dropbox
      • Google Drive
      • Gist
    • Download
      • Markdown
      • HTML
      • Raw HTML
Menu Note settings Note Insights Versions and GitHub Sync Sharing URL Create Help
Create Create new note Create a note from template
Menu
Options
Engagement control Make a copy Transfer ownership Delete this note
Import from
Dropbox Google Drive Gist Clipboard
Export to
Dropbox Google Drive Gist
Download
Markdown HTML Raw HTML
Back
Sharing URL Link copied
/edit
View mode
  • Edit mode
  • View mode
  • Book mode
  • Slide mode
Edit mode View mode Book mode Slide mode
Customize slides
Note Permission
Read
Only me
  • Only me
  • Signed-in users
  • Everyone
Only me Signed-in users Everyone
Write
Only me
  • Only me
  • Signed-in users
  • Everyone
Only me Signed-in users Everyone
Engagement control Commenting, Suggest edit, Emoji Reply
  • Invite by email
    Invitee

    This note has no invitees

  • Publish Note

    Share your work with the world Congratulations! 🎉 Your note is out in the world Publish Note No publishing access yet

    Your note will be visible on your profile and discoverable by anyone.
    Your note is now live.
    This note is visible on your profile and discoverable online.
    Everyone on the web can find and read all notes of this public team.

    Your account was recently created. Publishing will be available soon, allowing you to share notes on your public page and in search results.

    Your team account was recently created. Publishing will be available soon, allowing you to share notes on your public page and in search results.

    Explore these features while you wait
    Complete general settings
    Bookmark and like published notes
    Write a few more notes
    Complete general settings
    Write a few more notes
    See published notes
    Unpublish note
    Please check the box to agree to the Community Guidelines.
    View profile
    Engagement control
    Commenting
    Permission
    Disabled Forbidden Owners Signed-in users Everyone
    Enable
    Permission
    • Forbidden
    • Owners
    • Signed-in users
    • Everyone
    Suggest edit
    Permission
    Disabled Forbidden Owners Signed-in users Everyone
    Enable
    Permission
    • Forbidden
    • Owners
    • Signed-in users
    Emoji Reply
    Enable
    Import from Dropbox Google Drive Gist Clipboard
       Owned this note    Owned this note      
    Published Linked with GitHub
    • Any changes
      Be notified of any changes
    • Mention me
      Be notified of mention me
    • Unsubscribe
    # From Concept to Production: Humanoid Robotics at Scale ## NVIDIA GTC 2026 產業技術研究報告 --- **Session ID**:S81645 **時間**:2026/03/16(週一)3:00–3:40 PM PDT **主題領域**:Robotics – Humanoid Robots **產業類型**:Manufacturing **技術等級**:General Interest **目標受眾**:Business Executive **NVIDIA Technology**:Jetson・Isaac・Omniverse・Cosmos **Panel 參與者**: NVIDIA(Head of Robotics and Edge Computing Ecosystem)、Tesla(VP, AI Software)、Hexagon Robotics(CEO)、Agility Robotics(CTO)、Skild AI(CEO & Co-Founder)、Stanford / Physical Intelligence(Assistant Professor / Co-Founder) --- ## 一、場次定位與核心命題 人形機器人正從原型(prototype)演進到量產就緒(production-ready)的系統,背後驅動力來自推理型 AI、規模化模擬、與即時邊緣運算的突破。但要真正實現,產業必須縮小「數位智能」與「物理現實」之間的落差。 本場 Panel 匯集全球領先機器人公司的關鍵人物,討論 Physical AI 的未來,以及解鎖真正有能力的人形機器人所需的條件。 ### 官方 Key Takeaways 1. 領先團隊如何從資料蒐集、模擬到實地 rollout 進行訓練與部署 2. Physical AI 的技術堆疊(foundation models、world models、control systems)如何跨平台泛化 3. 未來 3–5 年最具潛力的應用、商機與開放研究問題 --- ## 二、Panel 參與者與各自代表的路線 本場 Panel 的價值在於:把「整機量產」、「工業部署」、「foundation model 軟體層」、「學界/新創的 robot learning」放在同一張桌上。 | 機構 | 代表角色 | 路線定位 | |------|----------|----------| | NVIDIA | Robotics & Edge Computing Ecosystem | 以 Jetson/Isaac/Omniverse/Cosmos 支撐 physical AI 通用基礎設施 | | Tesla | AI Software VP | 整機 + 軟硬整合 + 量產思維 | | Hexagon Robotics | CEO | 工業 humanoid「AEON」,強調量測/感測、定位與工業自治 | | Agility Robotics | CTO | Digit 主打倉儲/物流,以 production deployment / safety / ROI 為主軸 | | Skild AI | CEO & Co-Founder | 機器人大腦/基座模型,強調跨機型(omni-bodied)與模擬+真實資料閉環 | | Stanford / Physical Intelligence | Assistant Professor / Co-Founder | 學界+創業雙重視角,robot interaction at scale 的可泛化學習 | --- ## 三、從 Prototype 到量產:人形機器人規模化的四大難關 ### 3.1 資料:從「少量 demo」到「長尾情境」的資料工廠化 人形機器人最大的瓶頸是長尾情境(跌倒、滑動、遮擋、物件變異、碰撞、抓取失敗、地形差異)。要量產,資料不能只靠少量人工示範,而需建立可擴張的資料閉環: | 步驟 | 內容 | |------|------| | 模擬產生情境 | 大規模平行模擬生成多樣化訓練資料 | | 真機蒐集失敗案例 | Failure mining,從部署中挖掘長尾事件 | | 回灌再訓練 | 新資料回流至訓練管線,持續改進策略 | | 回歸測試 | 每次更新後驗證既有能力未退化 | Skild AI 的公開敘述即強調用模擬與人類動作資料訓練,再用所有部署機器人的資料持續改進。 ### 3.2 模擬:從「看起來像」到「可轉移(Sim-to-Real)」 量產需要可轉移,而可轉移依賴: | 關鍵因素 | 說明 | |----------|------| | 模擬 fidelity | 物理行為(接觸、摩擦、平衡)的逼真度 | | Domain randomization | 視覺、物理參數的隨機化範圍 | | 物理一致性 | 接觸/摩擦/平衡等與真實世界的一致程度 | | 可重現測試環境 | 回歸測試需要確定性的模擬重播 | NVIDIA 的 Omniverse/Isaac 路線以及 Cosmos world foundation models 主打「為 physical AI 提供可生成/可推理的世界模型與資料管線」。 ### 3.3 控制:把「推理 AI」落到「可驗證的控制系統」 Humanoid 不是只有 planning;它需要連到多層控制: | 控制層級 | 功能 | |----------|------| | Foundation / World Model | 任務理解與情境推理 | | 任務規劃 | 技能選擇、步驟排序 | | Whole-body Control | 動態平衡、接觸切換 | | Manipulation | 手部抓取、插拔、鎖付 | | Safety Envelope | 力/速度/區域限制、碰撞停止 | 實務上常見分層架構:foundation/world model → 任務規劃 → 控制器(MPC/阻抗/足端控制)→ 即時安全監控。這也是「從數位智能到物理現實」要補的工程段落。 ### 3.4 部署:即時邊緣運算與可維運性 量產意味著在現場長時間可靠運作: | 部署要件 | 說明 | |----------|------| | 即時邊緣運算 | Jetson 等平台支援現場推論 | | 遠端監控 | 即時狀態、告警、遠端介入 | | 版本治理 | 模型/韌體/策略/安全規則的 OTA 更新 | | 可維運性 | MTBF/MTTR、零件供應、維修訓練 | | 能耗與散熱 | 長時間運作的功耗管理 | Agility 對外以「production deployment / safety / ROI」作為敘事主軸,即反映這些落地要件。 --- ## 四、Physical AI 的新技術堆疊 ### 4.1 三層堆疊架構 Session 第二個 Key Takeaway 明確指出 foundation models、world models、control systems 形成通用堆疊: ``` ┌─────────────────────────────────────┐ │ Foundation Model(泛化能力底座) │ │ 感知、語意理解、技能組合 │ ├─────────────────────────────────────┤ │ World Model(世界生成與理解) │ │ 情境生成、物理推理、資料擴增 │ ├─────────────────────────────────────┤ │ Control System(可驗證行為) │ │ 全身協調、安全監控、即時控制 │ └─────────────────────────────────────┘ ``` ### 4.2 各層對應的 NVIDIA 技術 | 堆疊層級 | 功能 | NVIDIA 技術 | |----------|------|-------------| | Foundation Model | 跨機型可遷移的能力表示 | Isaac GR00T N1(humanoid foundation model) | | World Model | 情境生成/理解/資料擴增 | Cosmos(open world foundation models + guardrails + data pipeline) | | Simulation | 高保真模擬與規模化訓練 | Isaac Sim / Omniverse / Newton | | Control & Deployment | 即時推論與邊緣運算 | Jetson | | 資產與場景標準 | 機器人與環境描述 | OpenUSD / Omniverse | ### 4.3 跨平台泛化的意義 這套堆疊的設計目標是「跨不同機器人平台泛化」——同一套 foundation model + world model 可以部署到不同廠商的 humanoid 上,差異化跑在 control 層與機械設計層。這也是 Skild AI 所強調的「omni-bodied」路線。 --- ## 五、最具潛力的落地場域(3–5 年) ### 5.1 以「可量化 ROI」優先排序 | 場域 | 典型任務 | ROI 論述 | 成熟度 | |------|----------|----------|--------| | 倉儲/物流 | 搬運、上料、轉運、拆疊箱 | 環境標準化、ROI 容易算 | 最高(已有部署案例) | | 工廠彈性工序 | 補料、取放、簡易組裝、巡檢 | 產線變更頻繁時,人形泛用性成本正當 | 中高 | | 高風險/高負載 | 危險搬運、夜間作業、惡劣環境 | 安全與人力缺口 | 中 | | 零售/服務 | 理貨、導引、簡單服務 | 勞力短缺驅動 | 中低 | | 家用通用 | 家務、照護輔助 | 場景極度多樣,可靠度門檻最高 | 低(3–5 年不宜作主戰場) | ### 5.2 導入策略建議 不要把「家用通用機器人」當 3–5 年主戰場。商業導入通常會先從工業可控場域建立可靠度與維運模式,再逐步外擴。建議的優先順序: 1. **先選結構化場域**(倉儲/工廠)建立基線 2. **累積可靠度數據**(MTBF/MTTR、安全事件率) 3. **建立維運模式**(遠端監控、OTA 更新、維修流程) 4. **再擴展至半結構化/非結構化場域** --- ## 六、導入評估指標 若要把 Panel 討論轉成可落地的企業導入/合作評估,建議至少用以下 KPI: | 指標 | 說明 | 量測方式 | |------|------|----------| | 任務成功率 | 按情境分層(常態/長尾) | 模擬 + 真機測試 | | 安全事件率 | 力超限、侵入禁區、跌倒/碰撞 | 即時監控 + 事件回報 | | 吞吐與節拍 | 與現場產線節拍對齊 | 實地量測 | | 可維運性 | MTBF / MTTR | 長期追蹤 | | 回歸測試覆蓋 | 模型/策略更新後的驗證覆蓋率 | 模擬 + 少量真機 | | TCO | 部署、保固、維修、人員訓練、停機成本 | 財務分析 | --- ## 七、開放研究問題(3–5 年) | 研究問題 | 說明 | |----------|------| | 長尾情境的系統性處理 | 如何自動發現、生成、與驗證 edge cases | | Sim-to-Real gap 的量化與監控 | 不只是「看起來能用」,而是有指標可追蹤 | | Foundation model 的可驗證性 | 大模型輸出如何與安全約束對齊 | | 多機協作 | 多台 humanoid 在同一場域的協調與避障 | | 人機協作安全 | 共享工作空間的動態安全策略 | | 硬體可靠度與成本下降 | 關節、致動器、感測器的量產成本與壽命 | --- ## 八、決策者應帶走的關鍵結論 | 結論 | 說明 | |------|------| | 量產的關鍵不是 demo 能做什麼 | 而是資料閉環、回歸測試、維運能力能否規模化 | | 技術堆疊正在收斂 | Foundation model + World model + Control + Simulation-first pipeline | | 先從可控場域建立基線 | 倉儲/工廠等結構化場域,以安全與維運作為規模化門檻 | | 跨平台泛化是趨勢 | 同一套 AI 堆疊可部署到不同廠商的 humanoid | | 資料閉環是護城河 | 模擬生成 + 真機 failure mining + 回灌再訓練的完整迴圈 | | ROI 必須可量化 | 不是技術最先進的贏,而是能算清楚 TCO 的先落地 | | 安全是量產的前提 | 功能安全、協作安全、安全監控不是附加功能 | --- ## 九、延伸學習資源 | 主題 | 建議資源 | |------|----------| | GTC26 S81645 場次資訊 | GTC Session Catalog | | NVIDIA Isaac GR00T N1 | Humanoid foundation model(公開報導與技術摘要) | | NVIDIA Cosmos | Open world foundation models(官網、Cookbook、docs) | | NVIDIA Isaac Sim / Lab | 機器人模擬與學習框架 | | Newton Physics Engine | 開源 GPU 物理引擎(Warp + OpenUSD) | | Hexagon AEON | 工業 humanoid 與 Robotics division 公告 | | Agility Robotics Digit | Production deployment 與安全能力公開資訊 | | Skild AI | Skild Brain、訓練資料路線公開報導 | | Physical Intelligence (Pi) | Chelsea Finn / Stanford 研究與創業脈絡 | | 會後回看 | NVIDIA On-Demand(會後以 S81645 搜尋錄影/投影片) | --- *— 報告完 —*

    Import from clipboard

    Paste your markdown or webpage here...

    Advanced permission required

    Your current role can only read. Ask the system administrator to acquire write and comment permission.

    This team is disabled

    Sorry, this team is disabled. You can't edit this note.

    This note is locked

    Sorry, only owner can edit this note.

    Reach the limit

    Sorry, you've reached the max length this note can be.
    Please reduce the content or divide it to more notes, thank you!

    Import from Gist

    Import from Snippet

    or

    Export to Snippet

    Are you sure?

    Do you really want to delete this note?
    All users will lose their connection.

    Create a note from template

    Create a note from template

    Oops...
    This template has been removed or transferred.
    Upgrade
    All
    • All
    • Team
    No template.

    Create a template

    Upgrade

    Delete template

    Do you really want to delete this template?
    Turn this template into a regular note and keep its content, versions, and comments.

    This page need refresh

    You have an incompatible client version.
    Refresh to update.
    New version available!
    See releases notes here
    Refresh to enjoy new features.
    Your user state has changed.
    Refresh to load new user state.

    Sign in

    Forgot password
    or
    Sign in via Google Sign in via Facebook Sign in via X(Twitter) Sign in via GitHub Sign in via Dropbox Sign in with Wallet
    Wallet ( )
    Connect another wallet

    New to HackMD? Sign up

    By signing in, you agree to our terms of service.

    Help

    • English
    • 中文
    • Français
    • Deutsch
    • 日本語
    • Español
    • Català
    • Ελληνικά
    • Português
    • italiano
    • Türkçe
    • Русский
    • Nederlands
    • hrvatski jezik
    • język polski
    • Українська
    • हिन्दी
    • svenska
    • Esperanto
    • dansk

    Documents

    Help & Tutorial

    How to use Book mode

    Slide Example

    API Docs

    Edit in VSCode

    Install browser extension

    Contacts

    Feedback

    Discord

    Send us email

    Resources

    Releases

    Pricing

    Blog

    Policy

    Terms

    Privacy

    Cheatsheet

    Syntax Example Reference
    # Header Header 基本排版
    - Unordered List
    • Unordered List
    1. Ordered List
    1. Ordered List
    - [ ] Todo List
    • Todo List
    > Blockquote
    Blockquote
    **Bold font** Bold font
    *Italics font* Italics font
    ~~Strikethrough~~ Strikethrough
    19^th^ 19th
    H~2~O H2O
    ++Inserted text++ Inserted text
    ==Marked text== Marked text
    [link text](https:// "title") Link
    ![image alt](https:// "title") Image
    `Code` Code 在筆記中貼入程式碼
    ```javascript
    var i = 0;
    ```
    var i = 0;
    :smile: :smile: Emoji list
    {%youtube youtube_id %} Externals
    $L^aT_eX$ LaTeX
    :::info
    This is a alert area.
    :::

    This is a alert area.

    Versions and GitHub Sync
    Get Full History Access

    • Edit version name
    • Delete

    revision author avatar     named on  

    More Less

    Note content is identical to the latest version.
    Compare
      Choose a version
      No search result
      Version not found
    Sign in to link this note to GitHub
    Learn more
    This note is not linked with GitHub
     

    Feedback

    Submission failed, please try again

    Thanks for your support.

    On a scale of 0-10, how likely is it that you would recommend HackMD to your friends, family or business associates?

    Please give us some advice and help us improve HackMD.

     

    Thanks for your feedback

    Remove version name

    Do you want to remove this version name and description?

    Transfer ownership

    Transfer to
      Warning: is a public team. If you transfer note to this team, everyone on the web can find and read this note.

        Link with GitHub

        Please authorize HackMD on GitHub
        • Please sign in to GitHub and install the HackMD app on your GitHub repo.
        • HackMD links with GitHub through a GitHub App. You can choose which repo to install our App.
        Learn more  Sign in to GitHub

        Push the note to GitHub Push to GitHub Pull a file from GitHub

          Authorize again
         

        Choose which file to push to

        Select repo
        Refresh Authorize more repos
        Select branch
        Select file
        Select branch
        Choose version(s) to push
        • Save a new version and push
        • Choose from existing versions
        Include title and tags
        Available push count

        Pull from GitHub

         
        File from GitHub
        File from HackMD

        GitHub Link Settings

        File linked

        Linked by
        File path
        Last synced branch
        Available push count

        Danger Zone

        Unlink
        You will no longer receive notification when GitHub file changes after unlink.

        Syncing

        Push failed

        Push successfully