External Force Estimation on a Robotic Surgical Instrument

In this paper, a novel force/torque estimation algorithm for the in-house developed instrument in the robotic-assisted arthroscopic surgery system is proposed. This surgical robot system consists of two parts with 7 degree-of-freedom (DOF) Franka Emika robot for providing 4-DOF Remote Centre of Motion (RCM) about the incision-trocar and an instrument performing bone grinding. The method utilizes Neural Networks (NN) in the Cartesian space to estimate external forces acting on the instrument. The instrument is a rigid-link mechanism attached to the end of the Franka robot by a 6-DOF wrist force sensor. With this proposed method it is possible to obtain force and torque estimation in Cartesian space for any rigid-link wrist mechanism under RCM constraints. Several experiments are performed on an actual robotic system prototype and results show the efficacy of the proposed method.