Variable-impedance and force control for robust learning of contact-rich manipulation tasks from user demonstration

This paper proposes a Cartesian variable-impedance and force controller that enables manipulators to track position and force references demonstrated by a user through kinesthetic teaching. The proposed approach deploys the variability of user demonstrations to adapt the compliance profile of the manipulator to uncertainties and utilizes interaction force measurements during task reproduction to enhance force tracking performance. A passivity analysis is provided to demonstrate the stability of the system and a simulation exemplifies how passivity is achieved in the presence of variable impedance and force feedback. Furthermore, using a 7-DOF manipulator equipped with a force sensor, two experiments were conducted to highlight the ability of the proposed approach in successfully reproducing tasks with disturbances, where the state-of-the-art methods fall short