Lead Screws & Linear Actuators: The Precision Backbone

Ball Screws vs Roller Screws

Two distinct types of linear screw mechanism are used in humanoid robots, each with different performance and cost profiles. Ball screws use recirculating steel balls to transmit load between the screw shaft and nut — they are mature, well-understood, and widely available in miniaturized form. Planetary roller screws replace the balls with small threaded rollers, achieving higher load capacity, longer life, and better precision at higher cost. Both are critical to humanoid robot design.

A typical full-size humanoid robot uses approximately 14 planetary roller screws (for high-load joints like knees and hips) and 34 miniature ball screws (for lower-load joints, fingers, and wrist drives). The engineering tradeoff is straightforward: planetary roller screws handle 3–10x the load of equivalent ball screws and survive shock loads far better, but cost roughly 4x more — approximately ¥1,500 per planetary roller screw versus ¥400 per miniature ball screw at current market rates. Robot designers allocate each type based on joint requirements: high-force, weight-bearing joints get planetary roller screws; high-count, moderate-load joints get ball screws.

Planetary Roller Screws — High-Performance Joints

Planetary roller screws are used in the most demanding joints — the knees, hips, and ankles — where robots must bear the full weight of their body and payload while maintaining precise position control. Manufacturing these screws in miniaturized form factors for robot-scale applications is extremely challenging: the threaded rollers must be machined to micron-level tolerances, heat-treated for hardness without warping, and assembled with precision that ensures smooth, low-friction operation across millions of cycles. Foreign leaders include Rollvis (Switzerland), Ewellix (Sweden, formerly SKF Motion Technologies), and Schaeffler (Germany). Chinese companies like Beite Technology (贝特科技, a confirmed Tesla Optimus supplier) and Henggong Precision (衡工精密) are making significant progress in domestic production.

Miniature Ball Screws — High-Volume Joints

Miniature ball screws handle the majority of joints by count — fingers, wrists, elbows, and other moderate-load applications. Because robots use so many of them (averaging 34 per robot), cost reduction here has an outsized impact on overall robot economics. The miniaturization challenge is significant: ball diameters of 1–3mm require sub-micron raceway tolerances, and the recirculation mechanisms must function reliably at scales where even small debris can cause jamming. Japanese leaders NSK and THK dominate the high-precision end of this market, but Chinese companies including Beite Technology (北特科技, Shanghai-listed) — which is investing ¥1.85B RMB in planetary roller screw capacity — and Best Precision (贝思特) have developed competitive miniature ball screw product lines for humanoid applications. For dexterous hand applications, even smaller lead screws are required for finger actuation, pushing manufacturing precision to its limits.

Market Size & Projections

The lead screw market for humanoid robots is one of the fastest-growing segments of the entire robotics supply chain. Starting from an estimated ¥6.2 billion RMB in 2025, the market is projected to reach ¥95 billion by 2030 and over ¥1.1 trillion by 2035 — a compound annual growth rate of 68% over the decade. These projections assume continued price erosion (5% per year for roller screws, 3% per year for ball screws) offset by explosive volume growth as humanoid robot production scales.

These projections rest on industry assumptions of 163,000 humanoid robots per year by 2030 and 2 million per year by 2035 — ambitious targets that depend on both cost reduction and commercial adoption. If adoption is slower, the 2030 figure could be half this; if faster (driven by government procurement or a breakthrough in manufacturing cost), it could exceed projections. For context, the lead screw market is comparable in scale to the reducer market (¥88.5B by 2030) and significantly larger than the force sensor market ($4.27B by 2031) — reflecting its outsized share of total robot component cost.

Competitive Landscape

The global lead screw market has historically been dominated by Japanese and German precision engineering companies — NSK, THK, Bosch Rexroth, and Ewellix among them. For industrial automation applications, these brands remain preferred. But China's humanoid robot boom is creating a new market segment with different requirements: smaller diameters, lighter materials, and volume pricing that Western suppliers are not optimized to serve.

Chinese companies most actively targeting this segment include Beite Technology (贝特科技, confirmed Tesla Optimus supplier), Best Precision (贝思特, specializing in miniaturized ball screws), and Henggong Precision (衡工精密). Their pricing typically runs 30–50% below foreign equivalents, with lead times that are often weeks shorter due to proximity to robot assembly plants in Shenzhen and Shanghai. The segments most vulnerable to domestic substitution in the near term are miniature ball screws (where Chinese quality is already competitive) and mid-range planetary roller screws. The high-precision roller screw segment — where tolerances below 5 microns are required — remains more challenging for domestic suppliers.

• Japanese leaders (NSK, THK): dominant in industrial precision applications; slow to optimize for humanoid-specific geometries
• Chinese challengers: rapidly developing miniaturized roller screw capability; some already qualified by leading robot makers
• Price trajectory: industry expects 5%/year price decline as volumes scale and manufacturing yields improve
• Bottleneck: precision grinding equipment — still largely imported from Japan, Germany, and Switzerland

Manufacturing Process

Making a miniaturized lead screw for a humanoid robot joint is an exercise in extreme precision. The thread profile must be accurate to microns; the surface finish must minimize friction while maximizing life; and the entire assembly must survive tens of millions of cycles under dynamic load.

The key manufacturing steps — raw material selection, thread grinding, heat treatment, surface treatment, and assembly — each contain potential bottlenecks. Thread grinding is the most critical: it requires CNC grinding machines from suppliers like Japan's Nachi and Germany's Reishauer, which remain largely imported. Heat treatment must harden the screw surface without introducing warpage — a particular challenge at miniature scales where even microns of distortion render the part unusable. Quality control at these scales requires expensive metrology equipment (coordinate measuring machines, laser interferometers) that adds to the capital cost of production. Chinese manufacturers are closing the gap by investing in imported grinding equipment, developing proprietary heat treatment processes, and implementing statistical process control methods adapted from the semiconductor industry.