Schaeffler Validates Formed Strain-Wave Gearboxes as Humanoid Robotics Pushes Toward Scale

Schaeffler has completed validation testing of a new Schaeffler strain-wave gearbox manufacturing process designed for humanoid robots, with mass production scheduled to begin in Germany in 2027. The company says the forming technology could significantly reduce the time, material and cost required to manufacture precision gear reducers for robot joints.

The important part of the announcement may not be the gearbox itself. Strain-wave gearing is already well established in robotics and other precision-motion applications. The potentially consequential development is how Schaeffler proposes to manufacture it.

Instead of relying primarily on time- and capital-intensive precision machining to produce critical gearbox geometry, Schaeffler says it has developed a forming process that shapes components with high pressing forces. According to the company, manufacturing steps that previously required minutes can be completed in seconds, manufacturing cost can fall by more than 25%, and material consumption can decline by more than 75%.

If those improvements survive high-volume production without sacrificing accuracy, stiffness or fatigue life, the implications could extend beyond one humanoid-robot program. Precision gear manufacturing itself could become an important battleground in the effort to industrialize robotics.

But validation testing is not the same as long-term field validation.

That distinction matters because the flexible component at the center of a strain-wave gearbox operates under repeated cyclic deformation, making fatigue performance one of the key engineering questions behind any new manufacturing method.

Key Takeaways

Schaeffler says it has completed validation testing of formed strain-wave gearboxes developed for humanoid robots and plans mass production beginning in 2027.
The company claims the forming process can cut manufacturing cost by more than 25% and material consumption by more than 75%.
Key manufacturing operations can reportedly be completed in seconds rather than minutes.
Schaeffler says formed units achieve torque and efficiency comparable to conventionally manufactured versions, but detailed independent performance data has not yet been published.
The key reliability question is whether high-volume forming can preserve dimensional accuracy and fatigue life across millions of robot-joint cycles.
Humanoid robotics itself remains an emerging market; industry organizations caution that widespread adoption is not yet assured.

Quick Take

Schaeffler’s announcement is less about inventing another robot gearbox than about attacking a potential manufacturing bottleneck. If high-precision strain-wave components can be formed rapidly instead of extensively machined while retaining dimensional accuracy and fatigue resistance, gearbox production may scale much more economically. The unresolved question is whether laboratory validation will translate into consistent long-life performance under demanding robotic duty cycles.

What Changed in Schaeffler Strain-Wave Gearbox Manufacturing

Conventional strain-wave gearbox production typically depends on extremely accurate manufacturing because a small, lightweight mechanism must simultaneously provide high reduction ratios, torque capacity, stiffness and positioning accuracy.

Schaeffler says its new process changes the manufacturing equation by using high pressing forces to create the desired component geometry rather than generating all critical geometry through material removal.

The company reports that this approach provides high dimensional accuracy while also improving process stability.

Schaeffler says the resulting gearboxes deliver torque and efficiency comparable with conventionally produced components. It has completed what it describes as extensive validation testing and plans to begin mass production in 2027, initially in Germany and later in additional regions.

The company is not approaching forming technology without prior experience. Schaeffler says it has supplied more than 2 million formed strain-wave gearboxes to automotive markets during the past decade and is now transferring that manufacturing knowledge to humanoid robotics.

That automotive history strengthens the industrialization argument, but it should not be interpreted as direct evidence that a humanoid robot gearbox will achieve identical life or performance. Load spectra, motion profiles, reversals, stiffness requirements, duty cycles and positioning demands can differ significantly between applications.

By the Numbers

Schaeffler-reported figures:

25%+ reduction in manufacturing cost
75%+ reduction in material consumption
Seconds instead of minutes for some key manufacturing steps
2027 planned start of mass manufacturing
2 million+ formed strain-wave gearboxes previously supplied into automotive markets over ten years
Several hundred humanoids Schaeffler separately says it plans to deploy in its global manufacturing network over five years

Why the Schaeffler Strain-Wave Gearbox Matters to Robotics

A strain-wave gearbox achieves a large reduction ratio within a very small mechanical envelope.

The basic mechanism uses three central elements: a wave generator, a flexible spline and a rigid circular spline. The wave generator elastically deforms the flexspline so its external teeth engage the internally toothed circular spline at opposing regions. A small difference in tooth count produces a large speed reduction as the wave generator rotates.

The architecture is attractive for robotics because compact joints often require a combination of high torque density, low weight, high positioning accuracy and high reduction ratios.

Established strain-wave products are also known for very low or effectively zero gear-mesh backlash, another important characteristic when a servo system must repeatedly position a robot joint with high accuracy.

These characteristics explain why strain-wave reducers appear in industrial robots, collaborative robots, precision equipment and increasingly humanoid designs.

They also explain why manufacturing quality matters so much.

A flexible, thin-walled precision component that repeatedly changes shape cannot simply be produced faster if that acceleration introduces dimensional variability, residual stresses or geometry changes that compromise its operating life.

Expert Insight

The manufacturing breakthrough will not ultimately be measured by how quickly a component leaves the press. It will be measured by whether millions of formed components can leave production with repeatable geometry and then retain accuracy after millions of load reversals in service.

The Reliability Question: What Happens to the Flexspline?

Recent research highlights why the reliability question deserves as much attention as manufacturing cost.

A July 2026 review in Sensors describes the flexspline as a thin-walled structure exposed to cyclic stresses during operation. The repeated deformation can create stress concentrations that contribute to crack initiation, propagation and eventual fatigue failure. The authors identify tooth-root stress concentration and accumulated cylinder bending stress as important mechanisms influencing fatigue behavior.

That does not mean formed strain-wave gearboxes are inherently less durable.

It means manufacturing claims should be evaluated against the properties that determine real gearbox life.

Schaeffler states that its process delivers high dimensional accuracy and comparable torque and efficiency. The August announcement, however, does not publish detailed data for fatigue cycles, torsional stiffness retention, backlash over life, tooth-profile deviation, noise, wear, failure distributions or accelerated-life testing.

Those metrics will become increasingly important as the technology moves from validation to series production.

What This Means

For gearbox manufacturers, robotics OEMs and precision-motion suppliers, the competitive issue may increasingly shift from simply designing a high-performance reducer to manufacturing one economically at automotive-like volumes.

That creates a difficult engineering requirement: reduce cycle time and material consumption without loosening the tolerances that enable precision motion.

If forming proves capable of doing both, it may change the economics of precision gearing. If dimensional variation or fatigue life becomes difficult to control at scale, machining and post-processing may remain essential in areas where Schaeffler hopes to eliminate or reduce them.

Humanoid Robotics Is Moving From Demonstration to Manufacturing

Schaeffler’s timing is significant because humanoid robotics is beginning to confront questions that mature industrial automation already understands well: supplier capacity, component reliability, production yield, serviceability and cost.

Earlier this year, Schaeffler announced a strategic partnership with British robotics company Humanoid involving development and supply of robot actuators. Schaeffler also said it intends to deploy several hundred humanoid robots across its own production network over five years.

That gives the company an unusual position as both a component supplier and an intended industrial user.

Broader robotics demand also remains strong. International Federation of Robotics data show more than 4.6 million industrial robots were operating globally in 2024, with 542,000 new units installed that year. Preliminary IFR figures indicate U.S. robot installations increased 11% in 2025.

But conventional industrial robotics and humanoid robotics should not be conflated.

The IFR has cautioned that large-scale adoption of humanoids remains uncertain and that humanoids are more likely to complement existing robot technologies than simply replace them. Its humanoid analysis also highlights the importance of scalable component supply chains as countries and manufacturers attempt to move toward production.

“Humanoid robotics may ultimately be limited as much by manufacturing repeatability and component life as by artificial intelligence.”

That makes Schaeffler’s gearbox announcement particularly relevant. Humanoid demonstrations can attract attention with software, vision systems and AI, but every moving joint still depends on mechanical components that must survive actual industrial duty.

What Has Not Yet Been Proven

The August announcement establishes several things: Schaeffler has developed the process, conducted internal validation, intends to commercialize it and has considerable forming experience.

It does not yet establish that the process will reshape the strain-wave gearbox market.

Among the unanswered questions are:

What exact validation cycle and load spectrum were used?
How does flexspline fatigue life compare with precision-machined alternatives?
How consistent are backlash, lost motion and torsional stiffness after extended cycling?
What additional machining, grinding, heat treatment or surface finishing remains necessary after forming?
What are production scrap rates and dimensional capability at true series volumes?
How sensitive is the process to gearbox size and reduction ratio?
How do field failures compare once the gearboxes accumulate significant service hours?

These are not objections to the process. They are the normal questions that separate a promising manufacturing technology from an established industrial production method.

Questions for Robotics and Gearbox Buyers

Organizations evaluating next-generation precision reducers should consider:

What lifecycle testing supports the rated service life?
How are dimensional tolerances controlled from batch to batch?
What performance changes occur after extended reversing duty?
How is flexspline fatigue monitored or validated?
What inspection method verifies formed tooth geometry?
Which manufacturing steps have actually been eliminated, and which precision-finishing operations remain?
What field data will be available before high-volume purchasing commitments are made?

What to Watch as Schaeffler Moves Toward 2027 Production

Several measurable developments will determine the significance of Schaeffler’s announcement.

2027 production launch: The first major milestone is whether series manufacturing begins on schedule in Germany.

Named customer programs: Production commitments from humanoid manufacturers will provide stronger evidence of market acceptance.

Published lifecycle data: Fatigue life, stiffness retention, lost motion, efficiency and wear data would allow more meaningful comparison with conventional alternatives.

Manufacturing yield: Cost reductions matter only if the process produces acceptable parts consistently at high volume.

Real-world robot hours: Schaeffler’s planned use of humanoids within its own factories may eventually produce valuable operational evidence about joint reliability.

Competitor response: Other precision-gear suppliers may respond with their own forming, machining, additive or hybrid manufacturing approaches as robot production volumes increase.

Industrial Gearbox Solutions Editorial Perspective

Schaeffler’s formed strain-wave gearbox is important because it moves the conversation about humanoid robotics away from demonstrations and toward industrialization.

A robot that can perform an impressive task once is an engineering achievement. A manufacturer that can build thousands of robots economically, with joints capable of repeating those movements reliably for years, faces a different problem entirely.

Schaeffler appears to be targeting one part of that problem by reducing the time, material and potentially the capital required to manufacture precision gearing.

The opportunity is substantial, but so is the engineering burden.

Strain-wave gearboxes depend on controlled elastic deformation of a highly stressed precision component. Faster production becomes meaningful only when geometry, stiffness, backlash, fatigue resistance and operating life remain predictable across production volumes.

For the power-transmission industry, that may be the larger story to watch: robotics is beginning to force precision gearbox manufacturing to confront the economics and repeatability expectations of mass production.

And if forming can meet those expectations, the consequences may extend well beyond humanoid robots.

Frequently Asked Questions

What is a strain-wave gearbox?

A strain-wave gearbox is a high-ratio precision gear reducer that uses a wave generator to elastically deform a flexible spline into engagement with a rigid circular spline. Its compact size, high reduction ratio and low backlash make it useful in robotics and precision-motion applications.

What did Schaeffler change in the manufacturing process?

Schaeffler says it developed a forming process that shapes critical components using high pressing forces instead of creating all geometry through conventional material-removal machining. Some manufacturing steps can reportedly be completed in seconds rather than minutes.

How much does Schaeffler say the new process saves?

The company reports manufacturing-cost reductions of more than 25% and material-consumption reductions of more than 75%. These figures are manufacturer claims and have not yet been independently validated in public high-volume humanoid production data.

When will Schaeffler begin mass production?

Schaeffler says series manufacturing is scheduled to begin in Germany in 2027 before expanding to additional regions.

Why is flexspline fatigue important?

The flexspline repeatedly deforms during gearbox operation. Research indicates that cyclic stress, tooth-root stress concentration and cylinder bending stress can contribute to crack initiation and fatigue failure, making lifecycle performance an important measure of strain-wave gearbox reliability.

Sources

Schaeffler — formed strain-wave gearbox announcement, August 13, 2026
Primary source for the manufacturing announcement and company-reported performance figures.
Read the Schaeffler announcement

Schaeffler — Humanoid technology partnership, January 13, 2026
Background on Schaeffler’s actuator strategy and planned internal humanoid deployment.
Schaeffler and Humanoid partnership

International Federation of Robotics — Humanoid Robots: Vision and Reality
Independent industry perspective on adoption, applications, limitations and scalable supply chains.
IFR humanoid robotics analysis

Sensors — Research Progress on Mechanical Properties and Fatigue Failure of Harmonic Drive Flexspline
Recent peer-reviewed technical review of flexspline stresses and fatigue mechanisms.
Read the flexspline fatigue research

Harmonic Drive — High-Precision Strain-Wave Gearing Technical Catalog
Technical reference for strain-wave gearbox architecture and operating principles.
Harmonic Drive technical catalog

International Federation of Robotics — World Robotics 2025
Background data on the scale and continuing expansion of industrial robotics.
IFR World Robotics data