Investment ResearchTopic research, evidence curves, and cycle signals
HomeSign inEN

Research library · updated 2026-06-15 · public

Warrior Brief｜Robotics Q5 China vs US Web Page Package

Date: 2026-06-15 Owner: Hugo / Genius Team Research agent: Finance / Charlie AGT-002 Target worker: Warrior / Team Fullstack Status: READY_FOR_WARRIOR Visibility: PUBLIC Output: site Priority: P0 Target route: /robotics/china-us-paths/

0. Page title

China vs US Robotics: Deployment Machine vs Frontier Stack

Subtitle:

China has the stronger deployment and supplier flywheel; the US has the stronger frontier humanoid and AI/model stack. Neither path has yet proven broad S5 humanoid economics.

1. Source files

Primary synthesis:

  • knowledge/robotics-q5-china-us-deployment-machine-vs-frontier-stack-v1.md

Supporting evidence:

  • knowledge/robotics-china-policy-deployment-machine-v1.md
  • knowledge/robotics-four-deep-questions-us-china-tesla-figure-leaderdrive-v1.md
  • knowledge/robotics-mainline-q0-q5-evidence-matrix-v1.md
  • knowledge/robotics-q0-cycle-stage-and-investment-timing-v2.md
  • knowledge/robotics-q1-value-stack-apple-android-or-something-else-v2.md
  • knowledge/robotics-q2-tesla-optimus-apple-style-candidate-still-missing-s5-proof-v2.md
  • knowledge/robotics-q3-figure-deployment-engineering-bottleneck-v2.md
  • knowledge/robotics-leaderdrive-s4-to-s5-evidence-debt-v1.md

2. Page thesis

China vs US should be framed as a path-conversion test, not a country-winner ranking.

The page should hold three ideas at once:

  1. China has stronger deployment machinery: industrial robot scale, domestic supplier share, policy-to-scene mechanisms, SOE/provincial deployment funnel, and supplier filing evidence.
  2. The US has stronger frontier stack concentration: Tesla capacity-intent, Figure deployment/manufacturing KPI, NVIDIA/model/data/simulation infrastructure, and mega-round private capital.
  3. Neither side has proven S5 broad humanoid economics because accepted units, repeat paid deployments, customer ROI/payback, intervention rate, robot revenue, margin, and service burden remain mostly undisclosed.

Memorable frame:

China has the deployment machine; the US has the frontier stack. The real test is S4-to-S5 conversion.

3. Recommended page architecture

Section A — Hero

Core message:

This is not a simple horse race. It is a conversion test.

Use a two-column hero:

  • China path:
    • industrial robot installation scale;
    • operational stock;
    • domestic supplier share;
    • MIIT/SASAC real-scene deployment funnel;
    • supplier filings like Leaderdrive.
  • US path:
    • Tesla Optimus capacity-intent filings;
    • Figure BMW + BotQ KPI;
    • NVIDIA / model / simulation stack;
    • frontier humanoid mega-rounds;
    • private-company concentration.

Then add a central S5 gate:

  • accepted units;
  • paid deployments;
  • repeat orders;
  • uptime / intervention;
  • customer ROI/payback;
  • robot revenue;
  • gross margin;
  • service burden.

Section B — China: deployment machine

Use cards with source grades.

Must include:

  • 2024 China industrial robot installations: 295,000 units. 🟢
  • China share of global industrial robot demand: 54%. 🟢
  • China operational stock: 2.027m robots. 🟢
  • China domestic supplier share: 57% in 2024 vs 47% in 2023. 🟢
  • 2026 MIIT/SASAC real-scene action: notice dated 2026-06-03, published 2026-06-08. 🟢
  • End-2026 routine deployment / representative scenario target. 🟢
  • 100+ high-value scenarios. 🟢
  • “万台级” landing capability target. 🟢 claim / 🟠 interpretation.
  • 2026-06-30 work-plan deadline and 2026-11-30 effectiveness-summary deadline. 🟢

Message:

China’s edge may be the evidence-production process: scene list → user data → consortium → validation → routine deployment → replication.

Guardrail:

Policy target ≠ delivered units. Scenario count ≠ economics.

Section C — US: frontier stack

Use cards with source grades.

Must include:

  • Tesla: filing-backed Optimus capacity-intent and manufacturing-infrastructure evidence. 🟢/🟠
  • Figure: BMW deployment KPI and BotQ manufacturing KPI. 🟢
  • Figure funding: US$675m Series B at US$2.6bn valuation, with frontier AI/compute investors/partners in existing artifacts. 🟢
  • NVIDIA / Isaac GR00T / model-data-simulation stack. 🟢/🟠
  • North America industrial robot orders: 31,311 units, US$1.963bn, +0.5% units, +0.1% value in 2024. 🟢

Message:

The US path is frontier-stack-heavy: AI/model/compute, private capital, and a few high-signal humanoid companies.

Guardrail:

Frontier capital and model stack do not equal repeatable customer economics.

Section D — Side-by-side path matrix

Use a compact matrix or paired cards:

China:

  • strongest in deployment scale;
  • stronger policy-to-scene machinery;
  • stronger domestic supplier flywheel;
  • needs proof of humanoid accepted units, ROI/payback, routine deployment, revenue/margin.

US:

  • strongest in frontier humanoid/model stack;
  • stronger named high-signal private/public frontier companies;
  • stronger AI/model/compute concentration;
  • needs proof of multi-customer paid deployments, low intervention, production economics, and customer ROI.

Section E — Why neither side is S5

Show two “not yet” cards.

China is not S5 because:

  • industrial robot installations are not humanoid accepted units;
  • policy targets are not delivered robots;
  • scenarios are not repeat paid deployments;
  • supplier revenue is not named humanoid OEM economics;
  • domestic supplier share is not customer ROI/payback.

US is not S5 because:

  • Tesla designed capacity is not actual production;
  • Figure BMW KPI is not ROI/payback;
  • BotQ output is not customer fleet utilization;
  • frontier funding is not product-market fit;
  • model/data evidence is not robot gross margin or service burden.

Section F — Upgrade dashboard

China upgrade signals:

  • MIIT/SASAC publishes named scenario outcomes with delivered/accepted units;
  • SOEs/customers disclose paid deployments and utilization;
  • RaaS pricing / payback / repeat expansion disclosed;
  • Chinese humanoid OEMs disclose revenue, backlog, margin, service cost;
  • supplier filings disclose named humanoid customers or humanoid-specific revenue/margin.

US upgrade signals:

  • Tesla discloses actual Optimus output, deployed robot count, factory productivity, intervention rate, ROI/payback, revenue or margin;
  • Figure/Apptronik/Agility disclose multi-customer paid deployments;
  • model/data layers demonstrably reduce deployment time or intervention in customer environments;
  • US frontier stack creates repeatable deployment playbooks rather than bespoke pilots.

Section G — Signal vs noise

Signal:

  • industrial installations / stock / supplier share;
  • dated policy funnel;
  • named-customer runtime/task KPI;
  • filing-backed supplier revenue/volume/margin;
  • SEC-filed capacity infrastructure;
  • model/data tooling tied to deployment outcome.

Noise unless upgraded:

  • country-winner rhetoric;
  • policy target as delivered units;
  • mega-round as product-market fit;
  • customer logo as ROI/payback;
  • industrial robot scale as humanoid economics;
  • frontier AI as deployment economics.

Section H — Common misconceptions

Include these:

  1. China’s industrial robot scale proves humanoid commercialization.
  2. US frontier AI/model strength automatically wins robotics.
  3. Policy support means all Chinese vendors benefit.
  4. Tesla/Figure evidence means US already has S5.
  5. China vs US should be the opening question.

Section I — Think Deeper questions

Use 5-7 questions:

  1. Which converts first into S5: China’s real-scene deployment funnel or the US frontier humanoid/model stack?
  2. Does the bottleneck move from robot body to deployment engineering?
  3. If China wins deployment scale but US wins model/data layer, where does value capture land?
  4. If US companies stay private longer, which public-market layers reflect their progress without forcing bad proxies?
  5. Could China produce data faster but economics slower?
  6. Could US companies achieve better autonomy but slower adoption because site integration is harder?
  7. Which 2026-2027 evidence matters most: accepted units, robot revenue, customer ROI, intervention rate, repeat order, or supplier humanoid revenue split?

4. Must-include source anchors

  • IFR, “China Tops World Record of 2 Million Factory Robots,” 2025-09-25. URL reached HTTP 200 on 2026-06-15; PDF text keywords were not extracted in the current environment, so numeric anchors rely on existing IFR-derived artifacts. 🟢
  • A3, “North American Robotics Market Holds Steady in 2024 Amid Sectoral Variability,” 2025-02-17. URL returned HTTP 200 on 2026-06-15; keyword anchors 31,311, 1.963, 0.5%, 0.1% matched. 🟢
  • USCC, “Humanoid Robots,” 2024-10-10. URL reached HTTP 200 on 2026-06-15; PDF text keywords were not extracted in the current environment, so claims rely on existing USCC-derived artifacts. 🟢
  • MIIT / SASAC, 2026 humanoid / embodied-AI real-scene action, dated 2026-06-03, published 2026-06-08. URL returned HTTP 200 on 2026-06-15; keyword anchors 100, 万台, 2026, 场景 matched. 🟢
  • Tesla Q1 2025 to Q1 2026 SEC Form 8-K Exhibit 99.1 sequence from Q2 artifacts. 🟢
  • Figure official BMW / BotQ / Catalyst / Helix sources from Q3 artifacts. 🟢
  • 绿的谐波 filings and Leaderdrive artifacts from Q4. 🟢
  • Charlie path-comparison synthesis and S4-to-S5 classification. 🟠

5. Do not say

  • Do not say China wins robotics.
  • Do not say the US wins robotics.
  • Do not turn geography into a stock recommendation.
  • Do not say industrial robot scale proves humanoid economics.
  • Do not say policy targets are delivered units.
  • Do not say mega-rounds prove product-market fit.
  • Do not say Tesla designed capacity is current production.
  • Do not say Figure BMW KPI proves customer ROI/payback.
  • Do not include Hugo private portfolio data or private rationale.

6. Preferred wording

Use:

  • “deployment machine vs frontier stack”
  • “path-conversion test”
  • “S4 evidence, not S5 economics”
  • “industrial robot scale ≠ humanoid economics”
  • “frontier AI ≠ deployment economics”
  • “policy funnel, not delivered-unit proof”
  • “country comparison after layer/evidence analysis”

Avoid:

  • “China wins”
  • “US wins”
  • “robotics race winner”
  • “policy guarantees commercialization”
  • “frontier stack guarantees adoption”
  • “geography equals investable conclusion”

7. Acceptance criteria

Warrior output is acceptable if:

  1. The hero frames the page as deployment machine vs frontier stack, not horse race.
  2. China and US evidence are separated by layer.
  3. China’s industrial scale is not overread as humanoid economics.
  4. US frontier stack is not overread as deployment economics.
  5. The MIIT/SASAC policy is framed as a funnel and target, not achieved units.
  6. Tesla/Figure/Leaderdrive references link back to Q2/Q3/Q4 evidence types.
  7. The S4-to-S5 conversion dashboard is clear.
  8. No country-winner rhetoric appears.
  9. No trade recommendation appears.
  10. No Hugo private portfolio data appears.

8. Public footer

Evidence map only. No country winner ranking. No trade recommendation. Industrial robot scale ≠ humanoid economics; frontier stack ≠ customer ROI.