Every gram in a humanoid joint is carried by the joint below it, and torque density sets what a robot can do. Convoloid has not yet been analyzed or tested at the fine pitches robot joints use. We are seeking a robotics partner for a fine-pitch validation program.
A heavier knee actuator needs a stronger hip, which needs a bigger battery, which adds mass everywhere. Gear mass at the distal joints is the most expensive mass in the robot.
Lifting, climbing and recovering from a stumble all demand peak joint torque. Actuators sized for those peaks are heavy for the rest of the duty cycle.
Fleets of robots running shifts accumulate gear cycles fast. Pitting and backlash growth show up as precision loss, service calls and downtime.
| Involute | Convoloid® | |
|---|---|---|
| Tooth contact | Convex on convex | Convex on concave (conformal) |
| Small-pinion undercut limit | Yes (base circle) | No base circle |
| Smallest module analyzed to date | — | 3.0 mm |
| Fine-pitch (module < 3) performance | — | Not yet validated: the purpose of the program |
With no base circle there is no undercut limit. At module 3 mm and above, analysis shows single-stage ratio increases of 5–40% and more. Whether that carries to fine pitch is what validation will show.
Sun–planet meshes have been fully designed and analyzed. A conjugate internal-ring algorithm exists, and ring-mesh optimization is a defined work package within a validation program.
Robotics engineers ask for evidence, not adjectives. Here is the record Convoloid has today, from larger gearing. A validation program extends it to robot joints, with first-mover licensing rights for the partner.
Under a U.S. DOE/NREL wind-energy research contract (2005–2008), a Convoloid retrofit of a 108 kW two-stage gearbox ran back-to-back against an optimized involute at up to 240% of rated load. Convoloid surfaces stayed stable; the involute’s micro-pitting progressed.
See the source →Published: 20–35% higher torque capability than equivalent involute designs, rated per AGMA/ISO with the same derating factors applied to both (Berlinger & Colbourne, 2011).
See the source →US 10,527,149 in force to ~2036. European patent EP 3 387 294 granted, with Unitary Patent and UK validation, to 2036.
See the source →Barney Berlinger, Convoloid's inventor, explains what makes the tooth form different and why it matters for humanoid robotics.
A validation program tests Convoloid at your module, loads and duty cycle, with first-mover licensing rights for the partner.