Robust Assembly of Complex Shaped Planar Parts Using Vision and Force

The paper addresses the fusion of vision and force-torque information for robotic assembly tasks. The torque-controlled DLR light-weight robot is used, which is specially designed for fast and safe interaction with a changing environment. The image processing is based on standard hardware components and uses color segmentation and affine invariant feature classification. It provides a position estimation within the region of attraction (ROA) of a compliance based assembly strategy. The assembly planning toolbox is based on a theoretical analysis and the maximization of the ROA. This guarantees the local convergence of the assembly process under consideration of the part geometry. The convergence analysis uses the passivity properties of the robot and the environment. The method is validated through extensive experiments