Cycloidal Quasi-Direct Drive Actuator Designs with Learning-based Torque Estimation for Legged Robotics

Abstract-- This paper presents a novel approach through the design and implementation of Cycloidal Quasi-Direct Drive actuators for legged robotics. The cycloidal gear mechanism, with its inherent high torque density and mechanical robustness, offers significant advantages over conventional designs. Additionally, we develop a torque estimation framework for the actuator using an Actuator Network, which effectively reduces the sim-toreal gap introduced by the cycloidal drive's complex dynamics. However, integrating the gearbox into a confined space with conventional planetary, spur gear, This paper presents a QDD actuator with a 10:1 cycloidal and belt drive mechanisms is challenging without sacrificing gearbox for legged robots. We also present a gated recurrent gear load capacity since they are less resilient to significant unit (GRU) based torque estimation framework to model impulse load, such as those experienced during a fall.