Research library · updated 2026-06-18 · public

Robotics why-now slide 3 source appendix — measurable before proven

Date: 2026-06-18 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: SYNTHESIS_CANDIDATE Visibility: PUBLIC Target use: slide 3 evidence appendix / /robotics/why-now/ source cards Public-safety: no Hugo private portfolio data; no trade recommendation; no paid-report excerpts; no private channel checks; no supplier rumors.

Related files:

  • knowledge/robotics-why-now-source-backed-catalyst-claim-control-2026-06-18.md
  • knowledge/robotics-why-now-catalysts-v1.md
  • knowledge/robotics-public-claim-bank-v1.md
  • knowledge/robotics-sharing-backlog.md

0. One-line answer

截至 2026-06-18,slide 3 的最稳 public claim 是:机器人行业“现在值得研究”不是因为 S5 商业化已经证明,而是因为证据单位已经从 TAM / demo 变成可量化、可证伪的 S3/S4 指标:AI/data stack、低价硬件 access、客户现场 KPI、制造/产能 KPI、上市公司 filing-backed revenue。🟢 primary / official anchors below; 🟠 Charlie synthesis dated 2026-06-18.

1. Packaging-ready thesis

Suggested title: Why robotics now: measurable before proven

Suggested subtitle: The evidence unit changed from demos and TAM to measurable catalysts — but S5 economics is still unpaid.

Allowed conclusion:

  • The robotics cycle is becoming researchable because multiple public, dated, quantified evidence streams now exist. 🟢/🟠
  • Current public evidence supports an S3/S4-heavy early-deployment / supply-chain-validation frame, not an S5 return-cycle claim. 🟠

Do not say:

  • Do not say robots are already broadly commercialized.
  • Do not say Tesla is producing 1M or 10M Optimus robots/year.
  • Do not say low-priced humanoids prove reliability, autonomy, customer ROI, or gross margin.
  • Do not rank companies as winners or recommend any trade.

2. Five source cards for slide 3

CatalystQuantified public anchorAs-of dateGradeSignalMissing denominator
AI / data stackNVIDIA Isaac GR00T page positions GR00T as a humanoid robot foundation-model / data / simulation / runtime stack; prior local artifacts also capture NVIDIA synthetic-motion figures including 780,000 synthetic trajectories, 6,500 human-demo-hour equivalent, 11-hour generation time, and reported +40% GR00T N1 lift when synthetic + real data are combined.Reviewed through 2026-06-18; direct URL HTTP 200 from this machine on 2026-06-18🟢 NVIDIA official page / technical materials; 🟠 commercialization implicationS3/S4 platform-enablementCustomer-site intervention-rate decline, deployment-time decline, paid software economics, multi-OEM adoption metrics.
Hardware accessUnitree official pages list R1 AIR from US$4,900 / R1 from US$5,900, G1 from US$13.5K, and H2 at US$29,900, tax/shipping excluded.Product pages accessed in prior artifacts 2026-06-10; direct URLs HTTP 200 from this machine on 2026-06-18🟢 Unitree official product pages; 🟠 adoption implicationS3/S4 cost-access / productizationShipment volume, active deployed fleet, uptime, intervention rate, warranty/support burden, gross margin, customer ROI/payback.
Customer-site KPIFigure says its BMW Plant Spartanburg deployment ran across an 11-month period, 10-hour shifts Monday-Friday, loaded 90,000+ parts, achieved 1,250+ runtime hours, and contributed to production of 30,000+ BMW X3 vehicles.Figure official post dated 2025-11-19; URL HTTP 200 from this machine on 2026-06-18🟢 Figure official post; 🟠 economics implicationS4 deployment KPIRobot count, contract value, uptime/intervention distribution, customer-confirmed ROI/payback, repeat order, vendor gross margin, service burden.
Manufacturing KPI / capacity intentFigure says BotQ delivered 350+ Figure 03 robots, improved cadence from 1/day to 1/hour, reached >80% end-of-line first-pass yield, 99.3% battery-line FPY, 500+ battery packs, and 9,000+ actuators. Tesla Q1 2026 Exhibit 99.1 says Fremont first-generation Optimus line is designed for 1M robots/year and Texas second-generation line for long-term annual capacity of 10M robots.Figure post dated 2026-04-29; URL HTTP 200 from this machine on 2026-06-18. Tesla filing dated/filed 2026-04-22; SEC URL returned 403 from this machine on 2026-06-18, while local handoff records prior 2026-06-18 HTTP 200 verification.🟢 Figure official post; 🟢 Tesla SEC filing as captured in prior verified artifacts; 🟠 comparison synthesisS4 manufacturing / capacity-intentActual quarterly output, line utilization, accepted units, sustained yield, inventory/service quality, customer demand, revenue and margin.
Filing-backed revenue / supplier disclosureUBTECH 2025 annual report: RMB 2.001bn revenue (+53.3% YoY), full-size embodied humanoid revenue RMB 820.557m (+2,203.7% YoY; 41.1% of revenue), 1,079 units sold. 绿的谐波 2025 annual report: RMB 570.714m revenue (+47.31% YoY), net profit RMB 124.367m (+121.42% YoY), harmonic reducer sales 425,158 units (+72.48% YoY), industrial/embodied-intelligent robot component revenue RMB 422.528m with 34.88% GM.UBTECH disclosed 2026-04-15 in prior artifacts; 绿的谐波 disclosed 2026-04-23; SSE PDF URL HTTP 200 from this machine on 2026-06-18 but PDF body not parsed in this pass.🟢 listed-company annual reports as captured in prior artifacts; 🟠 sector implicationS4 filing-backed commercial / supplier signalHumanoid-specific margin, named customer mix, order volume, ASP, repeat orders, customer ROI/payback, margin durability.

3. Signal vs noise

Signal:

  • Dated, quantified, source-backed numbers: runtime hours, task count, production contribution, delivered robot count, cadence, first-pass yield, revenue, unit sales, gross margin. 🟢
  • Evidence chains appearing together: model/data stack + hardware access + deployment KPI + manufacturing KPI + filings. 🟠
  • Future falsification gates: accepted units, uptime/intervention, repeat order, ROI/payback, service burden, margin. 🟠

Noise unless upgraded:

  • Viral video without duration, reset count, intervention rate, failure cases, or customer-site denominator. 🔴/🟠
  • Customer logo without robot count, contract value, uptime/intervention, ROI/payback, repeat order. 🟠
  • Designed capacity treated as current production. 🟢/🟠
  • Low price treated as reliability, margin, or adoption proof. 🟢/🟠
  • Supplier product fit treated as named customer/order proof. 🔴/🟠

4. S5 upgrade gates

Upgrade from “why-now S4 evidence stack” toward “S5 commercialization proof” only if at least four of seven gates are answered with primary, customer-side, or filing-backed evidence:

  1. Accepted robots by customer/site/quarter. 🟢 required future evidence
  2. Uptime and intervention-rate distribution under production conditions. 🟢 required future evidence
  3. Task output and baseline replacement metric. 🟢/🟠 if baseline estimated
  4. Repeat order, multi-site expansion, or multi-period deployment from the same customer. 🟢 required future evidence
  5. Customer ROI/payback or transparent inputs for payback calculation. 🟢/🟠
  6. Robot revenue, gross margin, service/support cost, warranty reserve, and cash conversion. 🟢
  7. Safety, downtime, product-liability, cyber, maintenance, and insurance risk allocation. 🟢/🟡

5. Source list and this-pass verification notes

6. Public-safe footer

Evidence map only. No company ranking. No trade recommendation. S3/S4 evidence is not S5 commercialization proof. Designed capacity is not current production; low price is not reliability or margin proof; filing-backed robotics revenue still needs customer economics and margin durability.