Una Plataforma de Control Basado en Visión para la Rehabilitación de Robots Manipuladores de Tipo Industrial

In this paper a development of a visually based control platform for industrial robots is described. One of the objectives of a vision based control is to enable robotic manipulators to execute diverse complex tasks in non structured environments. Additionally to this aim, this platform presents two important characteristics: a) replaces the original controller, thus it is not required that the manipulator has a functional controller, and b) the applied control strategy is robust to errors in the characterization of the kinematics model of the robot, errors caused either by inaccuracy in the parameters of the model or due to the moderate deterioration of the mechanical parts. These advantages enable robotic manipulators to accomplish complex tasks and/or extending their useful life. In sum, the implementation of this platform allows a reduction of costs and an improvement of benefits in the commercial use of robotic manipulators for the accomplishment of industrial processes. The developed control platform was validated experimentally with a PUMA 761 robot.