Physical Human Robot Interaction for a Wall Mounting Robot - External Force Estimation

Abstract

The use of collaborative robots enhances human capabilities, leading to better working conditions and increased productivity. In building construction, such robots are needed, among other tasks, to install large glass panels, where the robot takes care of the heavy lifting part of the job while the operator directly manipulates the panel to install it. In order to facilitate physical human robot interaction, an
algorithm for estimating the operator input force is presented. It features a nonlinear observer structure embedded into an extended Kalman filter that also considers measurements from load cells. In this way the designed algorithm is more robust to
modeling errors. The derivation of the needed kinematic and dynamic model is also presented and the performance of the estimation is evaluated on on a high fidelity simulation model of the wall mounting robot called WallMo.