A vision guided robotic system for flexible gluing process in the footwear industry

his paper presents a novel gluing machine comprising a Cartesian robot and a vision system. The vision system enables the location and reconstruction of the shape of objects to be glued; the detected information is then used to plan the trajectory of the robot whose end-effector is a glue gun and to move the robot with an error suitable to industrial gluing operations. A calibration procedure that enables transforming coordinates between the robot frame and the vision system frame is described. The calibration considers several mechanical inaccuracies and its effectiveness was evaluated using error maps. In particular, the paper examines objects to be glued along their edges, as frequently occurs for fabrics, leathers, and shoe soles. For this, two procedures to plan the trajectories of the robot are proposed: the first is for objects that can be treated as flat 2D objects, that is, their height variation is negligible; the second procedure is for 3D objects, that is, those with significant height variation. Several applicative examples are reported to highlight the flexibility of the gluing process.