Precision Placement of DLOs with Flexibility via Shape Manipulation

In this paper, we present an open-loop method designed to enhance the precision of placing flexible linear objects on both flat and curved surfaces. Unlike rigid objects, where the motion can be accurately estimated from the configuration of a single point, the motion of soft objects is influenced by non-negligible internal forces, resulting in inconsistent and unpredictable behavior across different segments. We leverage the relationship between object shape and reaction forces, as explored through Lagrangian mechanics, to determine the desired shape that the object should be manipulated during placement, thereby mitigating unexpected displacements. A sequence of gripper motions is generated based on practical applications. We conducted a series of experiments utilizing a robotic arm to align and place deformable linear objects (DLOs) with different flexibility onto surfaces with diverse curvatures, demonstrating the effectiveness of our approach.