Within the framework of the H2020 European Project 'Bots2ReC' (roBots to ReConstruction) ("Robots to Re-Construction", 2017), an automated system for the removal of asbestos from a real world rehabilitation site will be developed. The system will consist of multiple robotic units, each one is composed of a mobile platform and a robotic arm with a rotative abrasive tool. In this paper, only the arm control is considered, and the smooth position-force controller presented in (Mohy El Dine, 2017) is validated on a Kuka LWR lightweight robotic arm. The arm is equipped with force/torque sensor, camera and spindle with a grinding tool. The mentioned controller reconsiders the position, force and impedance control strategies together from a practical point of view to achieve the real grinding task. The experimental validation shows the efficiency, advantages and drawbacks of the proposed controller and it draws conclusions about the main factors that affect the wall grinding operation and its quality.