S-DISG: Position and Torque Sensing for Driving Integrated Spherical Gear

Dexterous joint actuators significantly enhance the motion range and flexibility of robots, endowing them with human-like capabilities, which has garnered substantial interest from both academia and industry. The driving integrated spherical gear (DISG) pair, consisting of two magnetically coupled spherical gears, facilitates dexterous joint motion through the internal driving trolley. This article presents a targeted approach to position encoding and payload identification within the DISG system. We propose the spherical encoding method that arranges an array of ArUco tags on the inner wall of the spherical gear, combined with a camera and a visual recognition algorithm that fully leverages redundant tag information, achieving a configuration estimation with an average error of 3.95<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^\circ$</tex-math></inline-formula> and a root-mean-square error of 2.15<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^\circ$</tex-math></inline-formula>. In addition, we integrate spherical encoding data with inertial measurement unit data, taking advantage of the geometric relationship between the driving mechanism and the spherical gear to enable load estimation. Experiments conducted under loaded conditions demonstrated a load identification error of less than 8%. Building on these results, we implemented DISG joint position control for a robotic hand application. The results validate the proposed method’s effectiveness for position encoding and payload identification in real-world scenarios, providing technical support for advancing and applying DISG as a dexterous joint actuator.