Robotics Public Signal Timeline v1
Date: 2026-06-10 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe S3-S5 signal timeline seed Visibility: PUBLIC Related files:
knowledge/robotics-research-harness.mdknowledge/robotics-signals.mdknowledge/robotics-site-outline-v1.mdknowledge/robotics-tesla-optimus-evidence-v2.mdknowledge/robotics-figure-unitree-evidence-v1.mdknowledge/robotics-leaderdrive-public-case-v2.mdknowledge/robotics-why-now-catalysts-v1.md
Public-safety: yes. No private portfolio data. No trade recommendation. No paid-report text.
0. One-line answer
截至 2026-06-10，humanoid / embodied AI robotics 的公开 S3-S5 信号已经足够支撑一条 public signal timeline：Tesla 提供产线和设计产能信号，Figure 提供真实部署 KPI 和制造爬坡指标，Unitree 提供低成本硬件和开发平台信号，Leaderdrive / 绿的谐波提供 filing-backed 供应链收入与风险提示。整体结论仍是：行业从 demo cycle 进入 pilot / early deployment / supply-chain validation 的交界处，但还没有全面进入 S5 commercialization proof。🟢/🟠
1. Timeline schema
| Field | Meaning |
|---|---|
| Date | Event date or filing / publication date |
| Company / layer | OEM, platform, supplier, AI stack, or market structure |
| Signal level | S3 monitoring / S4 material / S5 thesis-changing |
| Event | What changed |
| Material metric | Quantified claim with as-of date |
| Why it matters | Thesis relevance |
| What is still missing | Evidence required before upgrading |
| Source grade | 🟢 primary / 🟡 secondary / 🟠 estimate / 🔴 guess |
2. Public S3-S5 signal timeline
| Date | Company / layer | Signal level | Event | Material metric | Why it matters | Missing before S5 | Source grade |
|---|---|---|---|---|---|---|---|
| 2025-04-22 | Tesla Optimus / OEM + manufacturing | S3/S4 | Tesla Q1 2025 update says it remained on track for Optimus builds on Fremont pilot production line in 2025 and wider deployment of bots doing useful work across Tesla factories. | No unit count disclosed; material claim is existence of Fremont pilot production-line build target as of Q1 2025 filing. | Moves Optimus beyond pure demo language into pilot manufacturing and internal factory-use language. | Actual units produced, factory robot count, hours worked, uptime, intervention rate, task economics. | 🟢 Tesla Form 8-K Exhibit 99.1 |
| 2025-07-23 | Tesla Optimus / strategic identity | S3 | Tesla Q2 2025 update frames Tesla as transitioning from EV/energy leader to AI, robotics and related services. | Strategic positioning disclosure; no robot KPI. | Confirms management narrative, but not commercialization evidence. | Production, customers, robot revenue, economics. | 🟢 Tesla Form 8-K Exhibit 99.1 |
| 2025-10-22 | Tesla Optimus / production infrastructure | S4 | Tesla Q3 2025 update says first-generation production lines for Optimus are being installed in anticipation of volume production. | Installation disclosed as of Q3 2025 filing; no achieved output disclosed. | Production-line installation is stronger than demo evidence. | Output, yield, utilization, production cost, customer deployment. | 🟢 Tesla Form 8-K Exhibit 99.1 |
| 2025-11-19 | Figure AI / deployment | S4 | Figure says Figure 02 completed an 11-month deployment at BMW Group Plant Spartanburg. | 10-hour shifts Monday-Friday; 90,000+ parts loaded; 1,250+ runtime hours; contributed to 30,000+ BMW X3 vehicles; as of Figure post dated 2025-11-19. | One of the strongest public humanoid deployment KPI packets: real customer site, task volume, runtime, and production context. | Customer-confirmed economics, robot count, contract value, repeat order, full uptime/intervention distribution, payback. | 🟢 Figure official post |
| 2026-01-27 | Figure AI / autonomy | S3 | Figure introduces Helix 02 full-body autonomy and says shown demos are fully autonomous, not teleoperated. | Figure highlights a 4-minute continuous household task in a kitchen/dishwasher setting; as of 2026-01-27 official post. | Useful autonomy evidence, but still video/demo-centric relative to commercial deployment. | Long-duration customer-site reliability, many-trial failure distribution, intervention rate, economics. | 🟢 for company claim / 🟠 commercial implication |
| 2026-01-28 | Tesla Optimus / manufacturing plan | S4 | Tesla Q4 2025 update says Gen 3 Optimus is first mass-production-intended design; preparations underway for first production line, including supply-chain readiness, with start of production planned before end-2026 and eventual planned capacity of 1M robots/year. | 1M robots/year eventual designed capacity for first line; start of production planned before end-2026; as of Q4 2025 filing. | Quantifies the scale ambition and links design, line preparation, and supply chain. | Current production rate, achieved yield, cost curve, robot revenue, deployment KPIs. | 🟢 Tesla Form 8-K Exhibit 99.1 |
| 2026-04-22 | Tesla Optimus / designed capacity | S4 | Tesla Q1 2026 update says preparations for first large-scale Optimus factory will begin shortly in Q2; Fremont first-generation line designed for 1M robots/year; Texas second-generation line designed for long-term annual capacity of 10M robots. | Fremont 1M/year designed line; Texas 10M/year long-term designed line; Robotics entries listed as California Optimus — Construction and Texas Optimus — Construction. | Highest public Tesla Optimus scale signal so far, but it is capacity design / construction, not current production. | Actual output, uptime, yield, unit cost, external paid customer, robot revenue, safety / reliability KPIs. | 🟢 Tesla Form 8-K Exhibit 99.1 |
| 2026-04-23 | Leaderdrive / 绿的谐波 / supplier filings | S4 | 2025 annual report shows filing-backed growth in robotics-relevant components. | 2025 revenue RMB 570.714m, +47.31%; attributable net profit RMB 124.367m, +121.42%; harmonic reducer sales 425,158 units, +72.48%; industrial and embodied-intelligence robot components revenue RMB 422.528m, gross margin 34.88%. | Stronger than concept-stock narrative because it is filing-backed revenue / volume / gross-margin evidence. | Named humanoid customer, humanoid-specific revenue split, project volume, contract amount, single-robot value content, margin durability. | 🟢 2025 annual report |
| 2026-04-29 | Figure AI / manufacturing | S4 | Figure says BotQ has delivered 350+ Figure 03 robots and improved production rate from 1/day to 1/hour. | 350+ Figure 03 robots delivered; 24x throughput improvement in under 120 days; EOL first-pass yield >80%; battery-line first-pass yield 99.3%; 500+ battery packs; 9,000+ actuators across 10+ SKUs; as of 2026-04-29 official post. | Moves Figure from handcrafted prototype perception toward quantified manufacturing operations. | Link production to customer utilization, revenue, backlog, gross margin, field failure / service cost. | 🟢 Figure official post |
| 2026-05-26 | Figure AI / commercial deployment | S3/S4 | Figure announces commercial agreement with Catalyst Brands to deploy humanoids into Catalyst’s distribution and logistics network, starting at Reno, Nevada Distribution Logistics Center. | Named commercial agreement and starting location; no robot count, value, or timeline disclosed as of 2026-05-26 official post. | Broadens Figure from BMW manufacturing toward logistics/distribution use case. | Contract value, robot count, deployment timeline, task KPI, repeat order, customer-confirmed ROI. | 🟢 Figure official post |
| 2026-06-09 | Leaderdrive / 绿的谐波 / public-market risk | S3/S4 caution | Abnormal-trading announcement says 688017.SH share price rose sharply versus indices and company had no undisclosed major matter. | 2026-06-04 to 2026-06-08: share price +36.00% vs SSE Composite -3.05% and automation-equipment index +2.13%; company says no undisclosed major matter; as of 2026-06-09公告. | Important anti-hype signal: price action can outrun new fundamental disclosure. | New official customer/order/revenue disclosure; otherwise do not infer new contract from share-price move. | 🟢 company/SSE announcement |
| 2026-06-10 | Unitree / 宇树 / low-cost hardware | S3 | Unitree official product pages show public humanoid price anchors across R1, G1 and H2. | R1 from US$4,900; G1 from US$13.5K; H2 US$29,900; product pages accessed 2026-06-10. | Lowers experimentation cost and may expand developer / research / early pilot surface. | Humanoid unit shipments, gross margin, customer mix, deployment uptime, task ROI, repeat orders. | 🟢 Unitree product pages |
| 2026-06-10 | Unitree / 宇树 + NVIDIA ecosystem / platform tooling | S3 | Unitree H2 Plus and G1-D pages package compute, teleoperation, simulation, data acquisition and model-training workflows around humanoid hardware. | H2 Plus: NVIDIA Jetson T5000, FP4 2,070 TFLOPS, Isaac GR00T / TeleOp / Sim references; G1-D: data acquisition / processing / labeling / review / model training / inference toolchain; accessed 2026-06-10. | Suggests Unitree is becoming an embodied-AI development surface, not only a robot body vendor. | Active developer adoption, dataset scale, production deployments, software/service revenue, customer economics. | 🟢 Unitree product pages |
3. Signal clusters by evidence type
3.1 Production infrastructure signal
- Tesla: S4 because filings disclose production-line installation, construction status, and designed capacity. 🟢
- Figure: S4 because official posts disclose delivered robot count, production cadence, yields, battery packs, and actuator output. 🟢
- Current limitation: infrastructure and manufacturing metrics still need to connect to field utilization and economics. 🟢/🟠
3.2 Deployment KPI signal
- Figure BMW: S4 because it includes a named customer, real plant context, runtime hours, task count, working-day operation, and vehicle-production contribution. 🟢
- Catalyst: S3/S4 because “commercial agreement” and named starting site are better than generic partnership language, but still lack economics. 🟢/🟠
- Current limitation: customer-confirmed ROI, robot count, payback, repeat order, and intervention rate are still missing. 🟢/🟠
3.3 Cost-curve / platform-access signal
- Unitree R1/G1/H2: S3 because public pricing changes experimentation cost but not deployment proof. 🟢
- Unitree H2 Plus / G1-D: S3 because tooling around data, teleoperation, simulation and model training is relevant to the embodied-AI bottleneck. 🟢
- Current limitation: price and tooling do not equal reliability, margin, or customer ROI. 🟢/🟠
3.4 Supplier filing signal
- Leaderdrive / 绿的谐波: S4 because revenue, shipments, margin, R&D and risk disclosures come from filings / official announcements. 🟢
- Current limitation: still no named humanoid OEM, contract value, humanoid-specific revenue split, or durable margin proof. 🟢/🟠
4. What is not in the timeline
Excluded as noise unless upgraded:
- Viral demo videos without runtime, task definition, intervention rate, or trial count. 🟠/🔴
- TAM-only slides without value capture or margin evidence. 🟠
- LOI / MOU / strategic cooperation without economics. 🟢/🟠 depending on source.
- Unsourced supplier lists, teardown speculation, or market rumors. 🔴
- Share-price moves without new filing-backed fundamental disclosure. 🟢/🟠
5. What would upgrade the sector from S4-heavy to S5
The sector would move closer to S5 if public evidence shows at least two of the following:
- Multi-customer paid deployments with robot count, contract value, task scope and timeline. 🟢
- Customer-confirmed ROI / payback and repeat orders. 🟢
- Field reliability: uptime, intervention rate, task success rate, safety incidents, maintenance cost. 🟢
- Manufacturing: quarterly output, yield, cost curve, inventory / delivery reconciliation. 🟢
- Supplier filings: humanoid-specific revenue, named customer or program, margin durability and capacity utilization. 🟢
- Public robotics revenue / gross margin line item becoming trackable for listed companies. 🟢
6. Common misconceptions
Misconception 1: “有一个 S4 信号就说明产业已经进入规模商业化。”
Correction: S4 说明 thesis 被加强，但还不是 S5。Figure BMW、Tesla designed capacity、Leaderdrive filing growth 都很重要；它们仍缺 economics、repeatability 或 customer-confirmed proof。🟢/🟠
Misconception 2: “产能规划就是实际产量。”
Correction: Tesla 的 1M / 10M 是 designed capacity / long-term designed capacity，不是当前 production rate。需要 output、yield、uptime、cost、deployment data。🟢/🟠
Misconception 3: “低价硬件等于商业化成功。”
Correction: Unitree 低价改变 experimentation cost，但商业化还需要 uptime、support cost、gross margin、deployment ROI 和 repeat orders。🟢/🟠
Misconception 4: “供应链收入增长就等于 humanoid winner。”
Correction: filing-backed 增长比概念叙事强，但还要验证 humanoid-specific exposure、客户、volume、ASP、margin durability 和 architecture risk。🟢/🟠
7. Think Deeper
- 如果 Tesla 先实现内部工厂部署而不是外部销售，应该如何把 Optimus 归类：cost-reduction automation、platform option，还是 future service business？
- Figure 的 BMW KPI 如果不能复制到第二、第三个客户，S4 deployment signal 应该降级到什么程度？
- Unitree 如果把 humanoid body 价格继续压低，价值会迁移到 AI models、data、deployment integration，还是售后服务？
- Leaderdrive 的“工业及具身智能机器人零部件”收入如果持续增长但毛利率下行，是正常降本放量还是议价能力不足？
- 哪个 S5 信号更先出现：OEM paid deployment economics，还是 supplier humanoid-specific revenue disclosure？
8. Public site section draft
Signal timeline: 从 demo 新闻转向可验证证据
机器人行业过去最容易被 demo 和 TAM 叙事带偏。我们更关心 S3-S5 信号：哪些信息能被公开来源验证，哪些信息能量化，哪些信息真正改变 thesis。
截至 2026-06-10，最重要的公开信号不是单个“赢家”，而是证据类型的变化：Tesla 披露 Optimus 产线安装和 1M / 10M designed capacity；Figure 披露 BMW 产线 1,250+ runtime hours、90,000+ parts 和 BotQ 制造指标；Unitree 披露 R1 from US$4,900、G1 from US$13.5K、H2 US$29,900 以及 NVIDIA / data tooling；Leaderdrive / 绿的谐波通过 filings 披露 embodied-intelligence-related component revenue and margin。
这些信号让研究从“有没有机器人视频”转向“有没有生产、部署、客户、回本和收入”。但它们还没有完成 S5 commercial proof。下一步真正重要的是 repeat orders、uptime、intervention rate、customer payback、robot revenue 和 supplier humanoid-specific exposure。
9. Source list
Tesla:
- Tesla Q1 2025 Update, Form 8-K Exhibit 99.1, filed 2025-04-22: https://www.sec.gov/Archives/edgar/data/1318605/000162828025018851/exhbit991.htm 🟢
- Tesla Q2 2025 Update, Form 8-K Exhibit 99.1, filed 2025-07-23: https://www.sec.gov/Archives/edgar/data/1318605/000162828025035738/exhibit991.htm 🟢
- Tesla Q3 2025 Update, Form 8-K Exhibit 99.1, filed 2025-10-22: https://www.sec.gov/Archives/edgar/data/1318605/000162828025045861/exhibit991.htm 🟢
- Tesla Q4 2025 Update, Form 8-K Exhibit 99.1, filed 2026-01-28: https://www.sec.gov/Archives/edgar/data/1318605/000162828026003837/exhibit991.htm 🟢
- Tesla Q1 2026 Update, Form 8-K Exhibit 99.1, filed 2026-04-22: https://www.sec.gov/Archives/edgar/data/1318605/000162828026026551/exhibit991.htm 🟢
Figure AI:
- Figure, “F.02 Contributed to the Production of 30,000 Cars at BMW”, 2025-11-19, accessed 2026-06-10: https://www.figure.ai/news/production-at-bmw 🟢
- Figure, “Introducing Helix 02: Full-Body Autonomy”, 2026-01-27, accessed 2026-06-10: https://www.figure.ai/news/helix-02 🟢
- Figure, “Ramping Figure 03 Production”, 2026-04-29, accessed 2026-06-10: https://www.figure.ai/news/ramping-figure-03-production 🟢
- Figure, “Figure Signs Agreement with Catalyst Brands to Scale Humanoid Operations”, 2026-05-26, accessed 2026-06-10: https://www.figure.ai/news/figure-signs-agreement-with-catalyst-brands 🟢
Unitree / 宇树:
- Unitree R1 product page, accessed 2026-06-10: https://www.unitree.com/R1 🟢
- Unitree G1 product page, accessed 2026-06-10: https://www.unitree.com/g1/ 🟢
- Unitree H2 product page, accessed 2026-06-10: https://www.unitree.com/H2 🟢
- Unitree H2 Plus product page, accessed 2026-06-10: https://www.unitree.com/H2plus 🟢
- Unitree G1-D product page, accessed 2026-06-10: https://www.unitree.com/G1-D 🟢
Leaderdrive / 绿的谐波:
- 绿的谐波 2025 annual report PDF, disclosed 2026-04-23: https://static.sse.com.cn/disclosure/listedinfo/announcement/c/new/2026-04-23/688017_20260423_EU45.pdf 🟢
- 绿的谐波 2026 Q1 report, disclosed 2026-04-23: SSE / company filing 🟢
- 绿的谐波 abnormal trading announcement, dated 2026-06-09: SSE / company announcement 🟢
10. Public-safe flag
Public-safe: yes.
Do not include Hugo private portfolio weights. Do not frame TSLA, 688017.SH, or any supplier/OEM as buy/sell/hold. Do not treat designed capacity, official demos, low price, or share-price moves as commercial proof.