TYO:6324
Every humanoid robot demo has the same hidden problem: the joints. A leg or wrist needs enormous torque reduction inside a space the size of a coffee cup, with no slack in the gear train. For more than fifty years the answer has been strain wave gearing, and one Japanese company has owned that answer almost outright.
An American invention, a Japanese business
Strain wave gearing was invented in the United States in the mid-1950s by C. Walton Musser. The mechanism is elegantly strange: a rigid outer ring with internal teeth, a thin flexible cup with slightly fewer external teeth, and an elliptical wave generator inside that deforms the cup so its teeth engage the ring at two opposing points. Rotating the wave generator walks the flexible cup around by the tooth difference, producing reduction ratios of 30:1 to 160:1 in a single stage — with essentially zero backlash.
The design was licensed into Japan and, in 1970, Harmonic Drive Systems was established to commercialise it. What began as licensed technology became a Japanese manufacturing franchise, because the value migrated from the patent to the ability to actually make the flexible spline reliably, millions of cycles deep, at production cost.
Why nobody has displaced it
Strain wave gears are conceptually public — the mechanism has been in textbooks for decades. Yet the company retains a share of the global market usually estimated well above 70 percent. The reasons are manufacturing, not legal:
- The flexspline is a fatigue part. A thin steel cup that is elastically deformed on every rotation must survive hundreds of millions of cycles. Alloy selection, forming, heat treatment and surface finish decide whether it lasts years or months.
- Tooth geometry is proprietary in practice. The company’s tooth profiles have been iterated across generations to raise torque capacity and reduce ratcheting, and each generation is validated on real customer duty cycles.
- Design-in is sticky. Robot arms are dimensioned around a specific gear unit. Changing supplier means redesigning the joint housing, re-tuning the controller, and repeating life testing.
| Application | Why strain wave gearing |
|---|---|
| Robot wrists and small arms | Highest reduction per unit volume; zero backlash for repeatable positioning |
| Semiconductor and FPD equipment | Wafer handling demands sub-millimetre repeatability in vacuum-compatible mechanisms |
| Machine tools | Rotary tables and indexing heads requiring rigidity under cutting load |
| Space | Mass-critical, non-serviceable actuators on rovers, arms and satellite mechanisms |
| Humanoids | High torque density in joints that must fit inside a human-scale limb |
The humanoid question
The recent wave of humanoid robot programmes — from US developers, Chinese manufacturers and Japanese incumbents alike — has made joint actuators the most watched component category in robotics. Strain wave gearing is the incumbent answer for compact, high-ratio joints, which puts Harmonic Drive Systems structurally in the path of that demand.
Two caveats deserve stating plainly. First, humanoid volumes remain small relative to the industrial robot and semiconductor equipment markets that actually drive the company’s revenue today; announcements are not shipments. Second, humanoid designers are actively exploring alternatives — planetary reducers, cycloidal units, and quasi-direct-drive actuators with low gear ratios and high-torque motors — precisely because they want backdrivability and shock tolerance that a stiff, high-ratio strain wave gear does not naturally provide.
The realistic outcome is a split: strain wave gearing keeps the precision-critical joints, while direct-drive-style actuators take the legs and other impact-loaded axes. For heavier industrial joints, cycloidal reducers from Nabtesco remain the standard.
A cyclical business with a structural position
The financial profile is worth understanding before treating the company as a pure growth story. Revenue is tightly coupled to capital spending in two volatile end markets — industrial robots and semiconductor production equipment — so results swing hard between boom and digestion years. Operating margins expand and contract sharply with utilisation because the manufacturing base is fixed and capital-intensive.
The group also operates internationally rather than exporting from Japan alone, with long-established affiliates in Germany and the United States serving European and North American robot and machine builders.
What it means for investors and partners
- Highest-purity exposure to robot joints. Few listed companies map as directly onto “number of precision joints built worldwide.” That is the appeal and the volatility in one sentence.
- Watch the actuator architecture debate, not the humanoid headlines. The investable question is which gear topology wins in volume humanoid designs, and that is being decided in engineering trade-offs now.
- For hardware developers, the constraint is lead time. Precision gear capacity does not expand quickly. Robotics startups routinely discover that the gearbox, not the motor or the software, sets their production schedule.
Our overview of how strain wave and cycloidal gearing divide the robot market is here: The Gears Inside Every Robot Joint.
The takeaway
Harmonic Drive Systems is what happens when a company treats a public mechanism as a manufacturing problem for fifty years. It does not own the idea of strain wave gearing. It owns the ability to make it survive a hundred million cycles — which, in robotics, turns out to be the same thing.
Interested in Japanese business opportunities?
Whether you're looking for technology partners, engineering talent, or market insights — we can help connect you with the right Japanese organizations.
Get in Touch →
