A Learning Methodology for Robotic Manipulation of Deformable Objects

The majority of manipulation systems are designed with the assumption that the objects being handled are rigid and do not deform when grasped. This paper address the problem of robotic grasping and manipulation of 3D deformable objects, such as rubber balls or bags filled with sand. Specifically, we have developed a generalized learning algorithm for handling of 3-D deformable objects in which prior knowledge of object attributes is not required and thus it can be applied to a large class of object types. A description of our learning methodology will be given in this paper. We outline our methodology for modeling the object deformation and learning the required minimum forces for grasping. Evaluation of the overall algorithm demonstrates that we can achieve an error level of 14% with respect to the minimum physical lifting force.