Robotic Automation for Grasping and Insertion Connectors: A Case Study

Abstract

Despite advances in robotic automation, the grasping and insertion of connectors remains problematic. This can mainly be attributed to the fact that most connectors have only a few image features by which they can be identified and the wires lack sufficient rigidity to support the connector in a stationary position. Many robots use RGB-D structured light systems to identify the pose of an object and grasp it; however, this type of camera is too expensive for most manufacturing applications. In this study, we developed an inexpensive system in which a pair of two-dimensional industrial cameras provide top and side views to facilitate the grasping and insertion of connectors. The objective functions were to maximize reliability while minimizing cost. In experiments involving 32 trials, the proposed scheme achieved a success rate of 90%.