Visual servoing and force control fusion for complex insertion tasks

Abstract

To have access to force sensing is indispensable element for applications in which robots interact directly with objects in external settings. The very nature and the limited accuracy of the vision systems used for position control, implies that these types of systems are not adequate for controlling the interaction of the robot with its setting. The application that is carried out here is an insertion process in which the object in which the insertion is to be made is in motion. To do so, a system that fuses the information from different vision and force systems to improve the performance of classical visual servoing systems are presented.