Humanoid Supply Chain · Mexico Initiative · Launching 2026
We help Series A–C humanoid robotics companies qualify Mexico-based suppliers for the components they buy today: wire harnesses, structural metals, battery enclosures, PCB assemblies, final integration. We are upfront about the components Mexico cannot deliver. We orchestrate. We do not pretend to manufacture.
The thesis in 60 seconds
The opportunity is not "build the whole robot in Mexico." It is the nearshore industrialization of selected subcomponents, subassemblies, testing, and integration services where Mexico's existing manufacturing depth creates immediate advantage.
Cumulative tariffs on Chinese components keep accelerating. The Section 232 robotics investigation closes May 30, 2026 — outcome could add 25–50% on top of existing duties for Chinese-origin imports.
Compliance-first nearshoring qualifies under USMCA today. Anti-transshipment enforcement is tightening — pretending will not work, but real substantial transformation in Mexico will.
150+ harness plants in Juárez. Nemak global #1 in aluminum casting. Foxconn, Flex, Jabil, Sanmina all present. BMW $800M battery plant in SLP. CATL plant planned. The base is there.
Latin America's first humanoid factory announced Feb 2026. ~2,000 jobs. ~$10K target price. Strong market signal that Mexico is the right geography for humanoid manufacturing in 2026.
The wedge
Categorized by Mexico's realistic near-term capability based on existing manufacturing adjacency — not aspirational planning. These are the components we orchestrate.
| Component family | % of BOM | Mexico fit | Anchor industrial base |
|---|---|---|---|
| Wire harnesses · cable assemblies | 3–5% | World-class | Aptiv 31 plants · Lear · Yazaki — Cd. Juárez cluster |
| Connectors · junction boxes | 2–3% | Strong | Molex (GDL) · TE Connectivity · auto supply chain |
| Aluminum housings · brackets · shafts | 4–6% | Very strong | Nemak global #1 · FRISA · Metalsa — Monterrey cluster |
| Sheet-metal subassemblies | 3–5% | Strong | 100+ CNC shops across 4 clusters |
| Battery enclosures · thermal plates | 5–8% | Strong with EV partner | BMW $800M (SLP) · CATL planned · Katcon thermal |
| PCB assembly · controller enclosures | 3–5% | Strong | Foxconn · Flex · Sanmina · Jabil — all in MX |
| Final integration · test fixtures | 2–4% | Strong | 3.4M+ vehicles exported 2024 — deep QA culture |
The Mexico industrial map
Mexico's manufacturing capabilities for humanoid-adjacent components are concentrated in four main clusters, each with distinct strengths. We have mapped them; we work with maquila partners across all four.
Nemak HQ (global #1 aluminum casting), FRISA (forged rings), Metalsa, Katcon, Ternium. 100+ CNC shops. Reflex Robotics chose this cluster for their humanoid factory. AS9100 + IATF 16949 commonly certified.
150+ wire harness plants. Aptiv 31 plants. Lear, Yazaki. Foxconn 1M+ sqft. Jabil flagship MX site (already mfg Apptronik Apollo). CATL plant planned. Largest electronics-mfg concentration in Mexico.
$17.9B EMS market. Flex, Sanmina, Molex, Continental, Intel design center. Sensor integration, vision modules, ECUs, software/firmware. Deep talent pool in electronics engineering.
Bombardier, Safran, ITP Aero, GE Aviation. AS9100 + NADCAP cluster. BMW $800M battery plant (SLP). Nidec e-axle plant. Aerospace tolerance standards directly transferable to humanoid precision requirements.
What we do NOT sell
Saying "we cannot deliver harmonic drives" is the move that makes the rest of the pitch credible. Mexico cannot do these components today. We will not pretend otherwise.
Green-zone components represent 20–35% of total BOM cost. These are high-volume, lower-margin layers — not the premium actuator stack (47–56% of BOM). The opportunity is real but must be sized correctly: we win on volume, compliance, and proximity, not margin-per-unit.
Our approach
Capital-light at the start. Validation gates between phases. We do not build a factory before demand says we should.
20–30 structured interviews with humanoid builders. Map 5–10 Mexican maquilas with relevant capability. Broker 2–3 small orders end-to-end to prove the supply chain works. Engage trade compliance counsel on USMCA qualification per component family.
Formalize JV with best-performing maquila partner. Hire senior manufacturing engineer (EMS or Tier-1 auto background). Target 2–3 signed LOIs for prototype-to-small-batch production. Begin proprietary software for AI-powered quoting + USMCA compliance automation.
Scale to serve 5–10 humanoid companies with recurring component supply. Deploy software platform as a SaaS layer for MX-US robotics supply chain management. Evaluate dedicated manufacturing investment based on actual volume + margin data.
Why us — honestly
Validation discipline
The strategy has a binary kill criterion at month 6 — published in our internal thesis and shared with anyone who asks. Most "consultants" tell you what you want to hear. We will not.
If by month 6 we cannot identify at least 3 humanoid companies willing to issue a Letter of Intent for Mexico-sourced subcomponents, the thesis is invalidated for this market cycle and the venture is shelved until volumes grow. This is a binary test — no rationalization.
If you engage with us, you are working with a partner who is structurally incentivized to be honest about whether the Mexico path serves your specific BOM. We do not need to keep the venture alive at your expense.
Who we serve
Sub-1,000 unit/year volumes. Jabil's MOQs and onboarding cycles do not match your prototype-to-pilot speed. We orchestrate around that constraint.
Your customers, investors, or board expect non-Chinese-origin components in your BOM. Section 232/301 + anti-transshipment enforcement makes that requirement structural, not preference.
You are not Tesla. Building your own factory is 12–18 months and tens of millions of dollars away. You need a Mexico-based supply chain that works for the next 24 months, not the next 7 years.
A 30-minute strategic call. No deck. We share our thesis under your standard NDA. We listen to your BOM and tell you honestly whether we can help — and where we can't.