Humanoid Robot Supply Chain Bottlenecks: The Components Holding Back Mass Production

The $30,000–$150,000 Problem

Every humanoid robot shipped today costs between $30,000 and $150,000 to build. The widely cited long-term target — the price point at which humanoid robots become viable for mass deployment in logistics, manufacturing, and domestic settings — is under $20,000. That gap is not primarily a software problem, a motor problem, or an AI problem. It is a supply chain problem.

McKinsey & Company's April 2026 analysis, "Turning Humanoid Supply Chain Constraints into Billion-Dollar Wins," identifies the component supply chain as the binding constraint on humanoid robot scale. Venture capital funding for robotics surged more than threefold between 2023 and 2025 to $40.7 billion annually. China committed a $138 billion state fund to AI and robotics. OEMs from Tesla and Figure AI to Unitree and AgiBot are scaling pilots. But the components that go into each unit — the reducers, screws, sensors, and magnets — are produced by a narrow set of manufacturers operating at volumes far below what humanoid scale requires.

The result is a scaling dilemma: low volumes prevent suppliers from investing in dedicated production lines, but without cost reductions, end-user demand stays constrained. Breaking this cycle is the central challenge of the next three to five years.

The Anatomy of a Humanoid Robot's Bill of Materials

McKinsey maps the humanoid hardware stack across five domains:

Together these five domains account for 85 to 90 percent of total unit cost. The remaining 10 to 15 percent covers cooling systems, wiring harnesses, and other subsystems.

The distribution matters because it determines where supply chain constraints have the most leverage. Actuators represent both the largest cost bucket and the highest bottleneck risk — a combination that makes the actuator supply chain the primary lever for reaching the sub-$20,000 target.

Three High-Bottleneck Component Clusters

McKinsey identifies three component clusters where structural supply constraints are most likely to emerge as humanoid volumes scale.

1. Harmonic and Strain-Wave Drives

Harmonic reducers — compact, zero-backlash gearboxes that lower motor speed while multiplying torque — are the dominant joint reduction technology in humanoid robots. Each robot requires 6 to 20 units depending on the joint count. At a cost that accounts for roughly 36 percent of rotary actuator expense, reducers are simultaneously the largest actuator cost driver and the hardest actuator component to scale.

The global market is concentrated among three players: Harmonic Drive Systems (Japan), which holds 20 to 25 percent of the global harmonic reducer market; Nabtesco (Japan), which holds approximately 60 percent of the cycloidal reducer segment; and Leaderdrive (China, 688017.SH), the primary domestic Chinese alternative. All three are documented suppliers to production humanoid OEMs.

The bottleneck is structural. Producing strain-wave gearing requires precision grinding tolerances that only approximately 12 percent of global machine tool manufacturers can meet. Unlike electronics — where capacity can be added rapidly through semiconductor fab expansion — harmonic gearboxes require dedicated tooling, metrology infrastructure, and long qualification cycles. Demand could outpace the speed at which suppliers add qualified capacity as humanoid volumes rise through the 2026 to 2028 window.

2. Planetary Roller Screws

McKinsey describes planetary roller screws as presenting an "even more acute" bottleneck risk than harmonic drives. These components convert rotary motor motion into precise linear actuation — controlling the extension and retraction of humanoid legs, spines, and arms.

The risk profile is specific: planetary roller screws occupy a high-precision niche with a narrow supplier base, long lead times, and limited substitution options. Ball screws — the commodity alternative — benefit from broad machine tool manufacturing demand that drives cost reduction. Planetary roller screws do not. As humanoid OEMs pursue higher payloads and more dynamic motion profiles, demand for robotics-grade roller screws could exceed existing suppliers' ability to scale capacity.

SKF (Sweden, SKA.ST) is the primary Western-origin supplier of precision planetary roller screws for humanoid applications, alongside select high-precision Asian manufacturers. The segment's structural constraints make early supplier qualification and co-development agreements the primary risk mitigation strategy for OEMs.

3. Force, Tactile, and Perception Sensors

The sensing and perception stack faces a different type of constraint. Unlike reducers and screws — where the bottleneck is manufacturing precision — sensing components face a technology readiness problem. Tactile sensors capable of providing the spatial and force resolution required for dexterous humanoid hand manipulation do not yet exist in a production-ready form at the required cost point.

McKinsey flags robotics-grade force and tactile sensing as facing structural supply constraints specifically because no adjacent industry operates at the required specification. The EV value chain — which provides adjacency for motors, batteries, and electronics — has no equivalent for tactile sensing. This means suppliers cannot leverage existing volume production from adjacent markets to compress costs.

Six-axis force-torque sensors — critical for wrist and ankle joint control — are further along. Keli Sensing (China, 688611.SH) is a documented Tier 1 supplier to Tesla Optimus for six-axis force-torque sensors. But the full tactile sensing stack needed for generalised dexterous manipulation remains a development-stage supply chain.

The EV Adjacency Advantage — and Its Limits

One of McKinsey's more nuanced findings is the role of adjacency to the electric vehicle supply chain. Many humanoid components overlap structurally with EV components: brushless DC motors, NdFeB permanent magnets, battery cells, power electronics, and compute chips all have EV or consumer electronics equivalents operating at high volume.

This adjacency matters because it directly influences how quickly cost curves compress. Components with strong EV spillover — motors, batteries, compute — benefit from existing tooling, supplier depth, and process maturity. Components without such adjacency — precision force sensing, tactile arrays, robotics-grade roller screws — face cost curves that must be established from scratch.

The rare earth magnet supply chain illustrates the strategic dimension of adjacency. NdFeB magnets are required in every brushless DC motor in every humanoid joint. China controls approximately 85 to 90 percent of global rare earth processing capacity. MP Materials (US, NYSE: MP) and Lynas Rare Earths (Australia, ASX: LYC) are the primary non-Chinese alternatives — critical for US and European OEMs seeking to de-risk their magnet supply chains from geopolitical exposure.

Two OEM Responses to the Bottleneck

McKinsey identifies two strategies OEMs are adopting in response to the supply constraint picture.

Vertical integration — building or acquiring component production in-house. This accelerates development and reduces cost at scale but requires capital-intensive manufacturing investment and long build timescales.

Close co-development — working with a limited number of supply partners to jointly develop humanoid-specific component variants, often including investment or exclusivity arrangements. Tuopu Group's relationship with Tesla Optimus (actuator assemblies), AgiBot's investment relationships with Sanhua (thermal management) and Jinyang Shares (battery cells), and Figure AI's relationships across the sensor and reducer stack all reflect this pattern.

Neither path is cost-efficient at small volumes. Both require OEMs to make supply chain bets before volume certainty exists — compressing the scaling dilemma rather than resolving it.

What This Means for the Sub-$20,000 Target

The sub-$20,000 BOM target is achievable — but not through volume alone. McKinsey's analysis implies it requires simultaneous progress on three fronts:

  1. Supplier ecosystem maturation — qualified manufacturers of harmonic drives, planetary roller screws, and force-torque sensors must invest in dedicated humanoid production capacity ahead of demand.
  2. Design standardisation — as long as every OEM uses bespoke component geometries, suppliers cannot achieve the production standardisation needed for cost compression. Modular actuator standards would accelerate this.
  3. EV adjacency leverage — OEMs that design their actuation stacks to maximise overlap with EV-qualified components (motors, magnets, power electronics) will reach cost targets faster than those requiring bespoke specifications throughout.

The companies that move first to secure qualified supply relationships — particularly in the high-bottleneck clusters of reducers, planetary roller screws, and tactile sensors — will have a structural cost advantage that compounds as volumes scale.

Geppetto Supply Chain Coverage

Geppetto indexes 134 component suppliers in the global humanoid robot supply chain, covering 15 component categories across Chinese and Western manufacturers. The suppliers named in this article are all indexed with full profiles, OEM relationship data, and bottleneck risk ratings.

Browse the full supply chain directory →

Key component categories: