Robotics OEM Three Evidence Curves Public Brief v1

Date: 2026-06-11 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact for Codex packaging Visibility: PUBLIC Related files:

• knowledge/robotics-research-harness.md
• knowledge/robotics-tesla-figure-unitree-slide-evidence-v1.md
• knowledge/robotics-oem-public-evidence-matrix-v1.md
• knowledge/robotics-public-claim-bank-v1.md

Public-safety: yes. No Hugo private portfolio data. No trade recommendation.

0. One-line answer

截至 2026-06-11，Tesla / Figure / Unitree 最适合公开表达成“三条证据曲线”，而不是三家公司排名：

1. Tesla Optimus = capacity ambition / vertical integration curve：Fremont first-generation Optimus line designed for 1M robots/year；Texas second-generation line long-term designed annual capacity 10M robots。🟢 Tesla Q1 2026 Form 8-K Exhibit 99.1, filed 2026-04-22.
2. Figure AI = deployment KPI / manufacturing KPI curve：BMW Spartanburg 11-month deployment 披露 10-hour shifts Monday-Friday、90,000+ parts loaded、1,250+ runtime hours、contribution to 30,000+ BMW X3 vehicles；BotQ 披露 350+ Figure 03 delivered、cadence from 1/day to 1/hour、EOL first-pass yield >80%、battery-line first-pass yield 99.3%。🟢 Figure official posts, 2025-11-19 and 2026-04-29.
3. Unitree / 宇树 = cost-access / developer-platform curve：R1 from US$4,900、G1 from US$13.5K、H2 US$29,900；H2 Plus / G1-D pages reference NVIDIA Jetson T5000, Isaac GR00T / TeleOp / Sim, and data / training workflows。🟢 Unitree official product pages, accessed 2026-06-10.

Current conclusion: the trio collectively shows stronger S3-S4 evidence than a demo-only cycle, but none publicly proves complete S5 scaled commercial economics yet: repeat paid deployment, customer ROI/payback, low intervention rate, actual output/yield, robot revenue, and robot gross margin remain missing or incomplete. 🟢/🟠 as of 2026-06-11.

1. Why this artifact exists

Existing files already contain thick evidence. Codex packaging risk is compression error:

• compressing Tesla's designed capacity into current production;
• compressing Figure's BMW KPI into complete unit economics;
• compressing Unitree's low price into reliability / gross margin proof;
• turning three different evidence types into a misleading winner ranking.

This brief gives Codex exact public-safe copy, slide structure, and caveats for the Tesla / Figure / Unitree comparison.

2. The three evidence curves

3. Public-safe slide draft

Slide title

Tesla / Figure / Unitree: three evidence curves, not one horse race

Slide subtitle

Capacity ambition, deployment KPI, and cost-access curve answer different questions. None alone proves scaled commercial economics yet.

Main visual

What public evidence supports today

Tesla Optimus                     Figure AI                         Unitree / 宇树
Capacity ambition                 Deployment + manufacturing KPI    Cost-access platform
1M Fremont designed line          BMW: 1,250+ hours, 90k+ parts     R1 $4.9k / G1 $13.5k
10M Texas long-term design        BotQ: 350+ robots, >80% EOL FPY   H2 $29.9k + dev stack

What is still missing before S5

Tesla: actual output, yield, external paid customer, robot revenue, gross margin.
Figure: contract value, customer-confirmed ROI, repeat orders, intervention rate, gross margin.
Unitree: deployment KPI, shipment volume, uptime/reliability, service cost, margin.

Required footer

Evidence map, not winner ranking. Designed capacity ≠ production; customer KPI ≠ full economics; low price ≠ reliability or margin.

4. Public site section draft

Section heading

三家公司，三种证据，不是一个排行榜

Body copy

Tesla、Figure、Unitree 经常被放在同一张 humanoid 对比图里，但它们当前公开证据回答的是三道不同问题。

Tesla 的核心信号是制造尺度意图：截至 Tesla Q1 2026 8-K，Fremont first-generation Optimus line 的设计产能是 1M robots/year，Texas second-generation line 的 long-term designed annual capacity 是 10M robots。这个证据很强，但它证明的是 infrastructure / capacity ambition，不是当前实际产量或外部客户经济性。🟢/🟠

Figure 的核心信号是 deployment KPI 与 manufacturing KPI：BMW Spartanburg 部署披露 11 个月、10-hour shifts Monday-Friday、90,000+ parts loaded、1,250+ runtime hours、contribution to 30,000+ BMW X3 vehicles；BotQ 披露 350+ Figure 03 delivered、cadence from 1/day to 1/hour、EOL first-pass yield >80%、battery-line first-pass yield 99.3%。这比普通 demo 更接近商业化证据，但仍缺 contract value、robot count by customer、客户确认 ROI/payback、repeat-order economics、intervention rate 和 gross margin。🟢/🟠

Unitree 的核心信号是低成本硬件和开发者可获得性：R1 from US$4,900、G1 from US$13.5K、H2 US$29,900，并在 H2 Plus / G1-D 页面披露 Jetson / Isaac / TeleOp / Sim / data workflow 相关配置。低价格会显著降低实验门槛，但不能自动证明 industrial deployment reliability、maintenance cost、gross margin 或 customer ROI。🟢/🟠

更好的读法不是问“谁赢了”，而是问：谁在证明 capacity，谁在证明 deployment，谁在证明 cost curve？下一阶段要看的，是这些曲线能否汇合成 S5：repeat paid deployment + customer ROI + low intervention + production economics + financial disclosure。

5. Approved public claims

Claim A — Tesla

Approved:

Tesla Optimus 的公开强信号是超大产能规划和垂直整合制造路径；它还不是外部客户经济性的公开证明。

Source anchor:

• Tesla Q1 2026 Form 8-K Exhibit 99.1 disclosed Fremont first-generation Optimus line designed for 1M robots/year and Texas second-generation line long-term designed annual capacity of 10M robots. Filed 2026-04-22. 🟢

Do not say:

Tesla 已经有 1M 或 10M 台 Optimus 年产量。

Use instead:

Tesla 披露的是 designed capacity / long-term designed annual capacity，不是 current production rate。

Claim B — Figure

Approved:

Figure 当前最强的公开证据是 named customer deployment KPI 和 manufacturing KPI；但还不能把这些 KPI 直接等同于完整 unit economics。

Source anchor:

• BMW deployment metrics: 10-hour shifts Monday-Friday, 90,000+ parts loaded, 1,250+ runtime hours, contribution to 30,000+ BMW X3 vehicles. Figure official post, 2025-11-19. 🟢
• BotQ production metrics: 350+ Figure 03 delivered, cadence from 1/day to 1/hour, EOL first-pass yield >80%, battery-line first-pass yield 99.3%, 9,000+ actuators. Figure official post, 2026-04-29. 🟢

Do not say:

Figure 已经证明大规模商业化经济性。

Use instead:

Figure 的 public deployment evidence 很强，但 S5 仍需要 repeat orders、customer ROI/payback、intervention rate、uptime 和 gross margin。

Claim C — Unitree

Approved:

Unitree 的强信号是 cost-access curve：低价格会扩大实验和开发者使用面，但低价本身不是 industrial reliability、customer ROI 或 gross margin proof。

Source anchor:

• Unitree R1 from US$4,900, G1 from US$13.5K, H2 US$29,900. Unitree official product pages, accessed 2026-06-10. 🟢
• Unitree H2 Plus / G1-D pages reference NVIDIA Jetson T5000, Isaac GR00T / TeleOp / Sim, and data / training workflows. Accessed 2026-06-10. 🟢

Do not say:

低价证明 Unitree 会赢。

Use instead:

低价降低实验门槛，但商业化还需要可靠性、维护成本、毛利率、deployment workflow 和客户回本。

6. Signal vs noise rules for this comparison

Signal

1. Quantified designed capacity in filings or official financial updates. 🟢 Tesla Q1 2026 8-K.
2. Named customer + quantified operating KPI: runtime hours, shift schedule, task count, production contribution. 🟢 Figure BMW post, 2025-11-19.
3. Manufacturing KPI: delivered units, production cadence, first-pass yield, battery/actuator output. 🟢 Figure BotQ post, 2026-04-29.
4. Transparent product pricing that changes experimentation cost. 🟢 Unitree product pages, accessed 2026-06-10.
5. Future S5 signal: repeat paid deployment, customer-confirmed ROI/payback, low intervention rate, actual output/yield, revenue and gross margin. 🟢 if filing/customer/company-confirmed.

Noise unless upgraded

1. Demo video without task definition, reset count, runtime, intervention rate, or customer context. 🟠/🔴
2. Designed capacity treated as actual output. 🟢 for stated capacity; 🟠 if interpreted as production.
3. Customer logo without robot count, contract value, ROI/payback, or repeat-order evidence. 🟢/🟠
4. Low price treated as proof of gross margin or reliability. 🟠
5. AI-stack integration treated as generalized autonomy proof. 🟢 for integration claim; 🟠 for commercialization inference.

7. What would change our mind

Upgrade toward S5 if:

• Tesla discloses actual Optimus quarterly production output, line yield, internal factory task KPI, cost savings, or external paid deployment. 🟢/🟠
• Figure discloses repeat deployment across BMW / Catalyst / a third named customer with robot count, uptime/intervention, and customer-confirmed ROI/payback. 🟢/🟠
• Unitree discloses humanoid shipment volume, industrial customer KPI, uptime, support cost, gross margin, warranty / maintenance cost, or repeat orders. 🟢/🟠

Downgrade if:

• Tesla capacity plans slip without output, task KPI, or external customer evidence. 🟢/🟠
• Figure deployments remain one-off and economics stay undisclosed after repeated customer announcements. 🟢/🟠
• Unitree low-cost hardware expands demos but does not convert into reliable paid deployments or creates support / warranty burden. 🟠

8. Common misconceptions

1. “Tesla 产能最大，所以商业化最确定。”

   • Correction: designed capacity is S4 infrastructure evidence, not actual output, customer ROI, robot revenue, or margin. 🟢/🟠

2. “Figure 有 BMW KPI，所以已经 S5。”

   • Correction: BMW KPI is strong deployment evidence, but not contract economics, robot count by customer, customer-confirmed ROI, repeat order, or margin. 🟢/🟠

3. “Unitree 低价会自动 commoditize 全行业。”

   • Correction: low price may reduce experiment cost and pressure hardware ASP, but industrial reliability, software/data, service cost, and gross margin still decide value capture. 🟠

4. “三家公司应该直接按 market share / winner 排名。”

   • Correction: as of 2026-06-11, the public evidence is not mature enough for a clean market-share race; it is better treated as three evidence curves. 🟠 synthesis.

9. Think Deeper questions

1. If Tesla first deploys Optimus inside its own factories, should commercialization be measured by internal cost savings, external robot revenue, or both?
2. Which Figure metric would most quickly upgrade S4 to S5: repeat orders, uptime/intervention, customer ROI/payback, or gross margin?
3. Does Unitree's low-cost platform shift value away from humanoid bodies toward software/data/services, or simply accelerate hardware commoditization?
4. In humanoid robotics, which curve usually leads the investable cycle: cost-access, production yield, deployment KPI, or customer ROI?
5. What evidence is easiest for public markets to overprice before S5: Tesla capacity, Figure customer logos, Unitree low price, or viral demo quality?

10. Source list

Primary / official sources:

• Tesla Q1 2026 Form 8-K Exhibit 99.1, filed 2026-04-22: https://www.sec.gov/Archives/edgar/data/1318605/000162828026026551/exhibit991.htm 🟢 Used for Fremont 1M robots/year designed line and Texas 10M robots/year long-term designed capacity.
• Tesla Q4 2025 Form 8-K Exhibit 99.1, filed 2026-01-28. 🟢 Used in related claim map for Gen 3 / mass-production-oriented design context.
• Figure official post, “F.02 Contributed to the Production of 30,000 Cars at BMW”, 2025-11-19: https://www.figure.ai/news/production-at-bmw 🟢 Used for BMW deployment KPI.
• Figure official post, “Ramping Figure 03 Production”, 2026-04-29: https://www.figure.ai/news/ramping-figure-03-production 🟢 Used for BotQ manufacturing metrics.
• Figure official post, “Figure Signs Agreement with Catalyst Brands to Scale Humanoid Operations”, 2026-05-26. 🟢 for announcement; 🟠 for economics because robot count, contract value, ROI/payback, and margin are undisclosed.
• Unitree R1 product page: https://www.unitree.com/R1 🟢 Accessed 2026-06-10; used for R1 price.
• Unitree G1 product page: https://www.unitree.com/g1/ 🟢 Accessed 2026-06-10; used for G1 price.
• Unitree H2 product page: https://www.unitree.com/H2 🟢 Accessed 2026-06-10; used for H2 price.
• Unitree H2 Plus and G1-D official product pages. 🟢 Accessed 2026-06-10; used for Jetson / Isaac / TeleOp / Sim / data-training references.

Internal synthesis sources:

• knowledge/robotics-tesla-figure-unitree-slide-evidence-v1.md 🟠 synthesis artifact, dated 2026-06-11.
• knowledge/robotics-oem-public-evidence-matrix-v1.md 🟠 synthesis artifact, dated 2026-06-11.
• knowledge/robotics-public-claim-bank-v1.md 🟠 synthesis artifact, dated 2026-06-11.

11. Public-safety check

Safe to publish if used as written:

• No Hugo private portfolio weights or private rationale.
• No buy / sell / hold language.
• No paid-report excerpts.
• No private-channel claims.
• Source grades and as-of dates included for material claims.

Do not publish:

• “Tesla / Figure / Unitree winner ranking.”
• “Tesla has 1M or 10M current Optimus annual production.”
• “Figure has proven scaled humanoid unit economics.”
• “Unitree low price proves reliability or margin.”
• Any trade recommendation or implication.