Online Payload Identification of a Franka Emika Robot for Medical Applications

Abstract

Torque-control approaches implemented in collaborative robots include the compensation of static and dynamic effects, such as the gravity compensation approach. Therefore, it is required to know the inertial parameters of each robot’s link. Robot controllers generally allow to get access to these inertial parameters and even directly provide the compensation torques. However, these torques must be recalculated when attaching a new tool to the end-effector, by providing the controller with the inertial parameters of the added payload. In this paper, we briefly mention four methods for the identification of the inertial parameters and present one in depth to identify the inertial parameters of a payload held by the end effector of a collaborative robot, i.e. Franka Emika robot, in order to improve its dynamic compensation in the context of medical applications.