Robotics Company Map Morning Review v1

Date: 2026-06-13 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact for morning review Visibility: PUBLIC Source discipline: 🟢 primary / 🟡 secondary / 🟠 estimate / 🔴 guess

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

0. One-line answer

截至 2026-06-13，humanoid / embodied robotics 的公开公司地图不应该被压缩成“谁赢了”，而应该被压缩成“谁在证明哪一种证据曲线”：Tesla 证明 capacity ambition，Figure 证明 deployment + manufacturing KPI，Unitree 证明 cost-access / developer-platform，UBTECH 证明 filing-backed humanoid revenue / units，Agility 证明 bounded logistics RaaS deployment，Apptronik 证明 partner-rich scaling path，1X 证明 home-humanoid preorder / capacity-intent + human-guidance data loop。🟢/🟠 synthesis from primary-source packets reviewed through 2026-06-13.

当前结论：这些公司 collectively 把行业从 demo-only 推到 S3/S4-heavy evidence cycle，但还没有形成 S5 scaled commercial economics。S5 仍需要 repeat paid deployment、customer-confirmed ROI/payback、low intervention rate、actual output/yield、robot revenue、robot gross margin、service cost 和 repeat-order quality。🟠 Charlie synthesis, 2026-06-13.

1. Core question

早晨 review / Codex 打包时，最容易犯的错误是什么？

不是缺素材，而是压缩错误：

1. 把 Tesla designed capacity 当成 current production。
2. 把 Figure BMW KPI 当成完整 unit economics。
3. 把 Unitree low price 当成 reliability / margin proof。
4. 把 UBTECH filing revenue 当成 S5 return-cycle proof。
5. 把 Agility / Apptronik / 1X 的 agreement / capacity / preorder 当成 repeat-order economics。
6. 把七条不同证据曲线压成一个“winner ranking”。

2. Company-map compression table

3. The morning-review framing

Recommended headline

Humanoid robotics is seven evidence curves, not one horse race.

Recommended subheadline

Capacity, deployment KPI, low-cost hardware, filing revenue, bounded RaaS, partner stack, and home preorders answer different questions. None alone proves S5 scaled commercial economics.

60-second spoken version

机器人公司地图已经比 2023-2024 年更值得跟踪，因为证据单位变了：Tesla 披露的是产线和长期设计产能，Figure 披露的是客户现场 KPI 和 BotQ 制造 KPI，Unitree 披露的是低价硬件和开发平台，UBTECH 披露的是 humanoid revenue / units，Agility 披露的是 bounded logistics RaaS 进展，Apptronik 披露的是客户/制造/AI/资本 partner stack，1X 披露的是家庭场景 preorder / capacity-intent 和 Expert Mode。🟢/🟠

但这些证据不是同一种东西。不能把 capacity、runtime、price、revenue、RaaS agreement、partnership、preorder 混成一个总分。当前更准确的判断是：行业处在 S3/S4-heavy evidence cycle，尚未进入 S5 return cycle。🟠

4. Signal vs noise rules for Codex packaging

Signal

1. Filing-backed revenue / unit disclosure: e.g. UBTECH 2025 full-size embodied humanoid revenue RMB 820.557m and 1,079 units sold. 🟢 2025 annual report disclosed 2026-04-15.
2. Named customer + quantified task KPI: e.g. Figure BMW 1,250+ runtime hours and 90,000+ parts loaded. 🟢 Figure official post 2025-11-19.
3. Manufacturing KPI: e.g. Figure BotQ >80% EOL FPY and 99.3% battery-line FPY. 🟢 Figure official post 2026-04-29.
4. Designed capacity / production-line disclosure: e.g. Tesla Optimus Fremont 1M/year designed line and Texas 10M/year long-term design. 🟢 Q1 2026 8-K.
5. Product price points that change experimentation cost: e.g. Unitree R1 US$4,900 / G1 US$13.5K / H2 US$29,900. 🟢 product pages accessed 2026-06-10.
6. Bounded workflow agreements after pilots: e.g. Agility GXO multi-year RaaS agreement after pilot. 🟢 Agility official release 2024-06-27.

Noise unless upgraded

1. Viral demo without runtime / task / reset / intervention / customer context. 🔴 if unsupported.
2. Customer logo without robot count, contract value, ROI/payback, uptime, or repeat-order terms. 🟢 for logo/announcement; 🟠 for economics inference.
3. Designed capacity interpreted as current production. 🟢 for capacity claim; 🟠 for production inference.
4. Low price interpreted as reliable deployment or margin quality. 🟢 for price; 🟠 for commercial inference.
5. Funding / valuation / partnership interpreted as revenue, margin, or customer ROI. 🟢 for announcement; 🟠 for economics inference.
6. Preorder / deposit interpreted as recognized revenue or durable retention. 🟢 for deposit mechanics; 🟠 for revenue/retention inference.

5. Slide-ready layout

Slide title:
Humanoid robotics is seven evidence curves, not one horse race

Columns:
1. Evidence curve
2. Best public anchor
3. Missing before S5

Rows:
Tesla      Capacity ambition        1M Fremont designed line / 10M Texas long-term design     Actual output, yield, revenue, margin
Figure     Deployment + mfg KPI     BMW 1,250+ h / 90k+ parts; BotQ 350+ robots, >80% FPY    ROI, repeat order, intervention, margin
Unitree    Cost-access curve        R1 $4.9k / G1 $13.5k / H2 $29.9k                         Deployment KPI, support cost, margin
UBTECH     Filing revenue/units     RMB 820.557m humanoid revenue; 1,079 units sold           Humanoid margin, ROI, repeat quality
Agility    Bounded RaaS             GXO multi-year RaaS; 100k+ totes                          Contract value, robot count, margin
Apptronik  Partner stack            Mercedes/GXO/Jabil/NVIDIA/DeepMind + >$935m Series A      Deployment count, ROI, revenue
1X         Home preorder/capacity   $200 deposit; 10k/year factory capacity; Expert Mode      Delivery, autonomy, retention, margin

Footer:
Evidence map only. No winner ranking. No trade recommendation. S4-heavy evidence does not equal S5 scaled-commercial economics.

6. Public site section draft

三个不要：不要排行榜，不要总分，不要把不同证据混在一起

机器人公司图谱的关键不是“谁赢了”，而是“谁在证明哪一种证据”。截至 2026-06-13，Tesla、Figure、Unitree、UBTECH、Agility、Apptronik、1X 的公开证据回答的是七个不同问题：能不能建产线？能不能在客户现场跑 KPI？硬件成本能不能足够低？有没有 filing-backed humanoid revenue？能不能在 bounded logistics workflow 里收费部署？有没有制造/AI/客户 partner stack？家庭场景有没有 preorder 和 data loop？🟢/🟠

这也是为什么不能把所有公司放进一个简单排名。Tesla 的 capacity ambition 不是 Figure 的 deployment KPI；Unitree 的 low price 不是 UBTECH 的 filing revenue；Agility 的 RaaS workflow 也不是 1X 的 home preorder。它们共同说明行业证据质量在提高，但并没有共同证明 S5 commercial economics。

更好的 public-safe 读法是：现在机器人行业从“看 demo”进入“看 KPI”的阶段。下一阶段真正会改变 thesis 的，不是更多视频，而是 repeat paid deployment、customer-confirmed ROI/payback、低 intervention、实际产量和良率、robot revenue、robot gross margin 与服务成本。

7. Common misconceptions

1. “Tesla 产能最大，所以商业化最确定。”

   • Correction: Tesla disclosed designed capacity / long-term designed annual capacity, not current Optimus output, external customer revenue, or robot margin. 🟢/🟠

2. “Figure 有 BMW KPI，所以已经 S5。”

   • Correction: BMW KPI is strong S4 deployment evidence, but contract value, robot count by customer, customer-confirmed ROI/payback, intervention rate, and gross margin remain undisclosed. 🟢/🟠

3. “Unitree 低价会自动证明商业化。”

   • Correction: low price reduces experiment cost, but industrial reliability, maintenance burden, gross margin, deployment workflow, and customer payback remain unproven. 🟢/🟠

4. “UBTECH 有 humanoid revenue，所以 return cycle 已经开始。”

   • Correction: UBTECH filing evidence is unusually strong, but humanoid-specific margin, customer ROI/payback, repeat-order quality, service cost, and durable profitability are still missing. 🟢/🟠

5. “Partnership stack 等于订单经济性。”

   • Correction: Apptronik / 1X / Agility partnerships and capacity/preorder signals matter, but partner logos without deployment count, revenue, margin, and ROI are not S5. 🟢/🟠

8. What would change our mind

Upgrade toward S5 if:

1. Any OEM discloses multi-customer paid deployments with robot count, runtime, task KPI, uptime, intervention rate, and repeat orders. 🟢 if official/customer-confirmed.
2. Customers confirm ROI/payback and operational savings over multi-month deployments. 🟢/🟡 depending on source.
3. Filings disclose humanoid-specific robot revenue, gross margin, service/warranty cost, and cash conversion. 🟢.
4. Tesla discloses actual Optimus production output/yield or internal factory cost-savings KPI, not only capacity. 🟢/🟠.
5. Unitree discloses shipment volume, industrial deployment KPI, support cost, gross margin, and repeat customers. 🟢/🟠.

Downgrade if:

1. Capacity targets slip without output/yield data. 🟢/🟡.
2. Customer deployments remain one-off and do not convert into repeat orders. 🟢/🟡.
3. Low-cost hardware produces broad demos but poor reliability / high support burden. 🟡/🟠.
4. Filing-backed revenue grows but margins, receivables, inventory, or cash flow deteriorate. 🟢/🟠.
5. Public-market narratives infer specific customers or orders not confirmed by filings/company/customer sources. 🔴/🟠.

9. Think Deeper questions

1. 哪一条证据曲线最可能先变成 S5：deployment KPI、filing revenue、manufacturing yield、customer ROI，还是 gross margin？
2. 如果 Tesla 先在内部工厂部署 Optimus，应该用 internal cost savings 还是 external robot revenue 定义 commercialization？
3. Unitree 的低价硬件会把价值迁移到 software/data/deployment services，还是只是扩大 hardware experiment base？
4. UBTECH 的 humanoid revenue 是一次性 early adoption wave，还是可重复的 deployment economics？需要哪些指标区分？
5. Public markets 会先奖励 S4 evidence，还是等 S5 margins / ROI 才重估？

10. Public-safety flag

Safe to publish if used as an evidence map.

Do not publish:

• Hugo private portfolio weights or private rationale.
• Buy / sell / hold language for TSLA, 9880.HK, 688017.SH, suppliers, or any public/private company.
• Claims that designed capacity equals current production.
• Claims that deployment KPI equals customer ROI or full economics.
• Claims that low price equals reliability or margin.
• Claims that partnership / funding / preorder equals scaled commercial proof.

11. Source list

Primary / official sources already captured in existing knowledge artifacts:

• Tesla Q1 2026 Form 8-K Exhibit 99.1, filed 2026-04-22: Fremont first-generation Optimus line designed for 1M robots/year and Texas second-generation line long-term designed annual capacity 10M robots. 🟢
• Tesla Q1 2025-Q1 2026 Form 8-K Exhibit 99.1 sequence reviewed in robotics-tesla-optimus-filing-evidence-audit-v1.md. 🟢
• Figure official post “F.02 Contributed to the Production of 30,000 Cars at BMW”, 2025-11-19: BMW deployment metrics. 🟢
• Figure official post “Ramping Figure 03 Production”, 2026-04-29: BotQ manufacturing metrics. 🟢
• Figure official post “Figure Signs Agreement with Catalyst Brands to Scale Humanoid Operations”, 2026-05-26. 🟢 for announcement; 🟠 for economics because key terms undisclosed.
• Unitree R1 / G1 / H2 / H2 Plus / G1-D official product pages, accessed 2026-06-10. 🟢
• UBTECH Robotics Corp Ltd Annual Report 2025, disclosed 2026-04-15. 🟢
• Agility Robotics official GXO / tote milestone / Mercado Libre / Toyota / Schaeffler releases, reviewed in robotics-agility-digit-evidence-v1.md. 🟢
• Apptronik official Mercedes-Benz / GXO / NVIDIA / Google DeepMind / Jabil / Series A releases, reviewed in robotics-apptronik-apollo-evidence-v1.md. 🟢
• 1X official NEO / manufacturing / factory / NVIDIA / World Model Lab pages and releases, reviewed in robotics-1x-neo-evidence-v1.md. 🟢

Internal synthesis sources:

• knowledge/robotics-oem-public-evidence-matrix-v1.md 🟠 synthesis artifact dated 2026-06-11.
• knowledge/robotics-oem-three-evidence-curves-public-brief-v1.md 🟠 synthesis artifact dated 2026-06-11.
• knowledge/robotics-public-claim-bank-v1.md 🟠 synthesis artifact dated 2026-06-11.
• knowledge/robotics-morning-review-why-now-evidence-ladder-v1.md 🟠 synthesis artifact dated 2026-06-13.