A singularity-restrained kinesthetic teaching method for collaborative robots

Collaborative robots play an indispensable role in both industrial and service sectors . Kinesthetic teaching technology allows users to program robots through intuitive hand-guiding actions, endowing collaborative robots with convenient deployment capabilities. However, such manual guidance may bring the robot close to singularities , as users primarily focus on the motion of the end-effector, which leads to performance degradation and poses a challenge to task reliability. This paper proposes a kinesthetic teaching method that avoids singularities in Cartesian space. The advantage of this method is that it only restrains the robot from approaching singularities while having no effects in the rest of the workspace by asymmetrically adjusting the mapping relationship between hand-guiding force and end-effector velocity. Experiments have been conducted on all possible singularities of a 6-DOF robot. The results indicate that the proposed method effectively mitigates uncoordinated motion, and the operational performance of the robot has been enhanced. • A method is proposed to avoid singularities by adjusting force–velocity mapping. • The method ensures inverse solution accuracy and avoids energy injection issues. • The method restrains motion towards singularities without affecting motion away.