Fast Vision-based Grasp and Delivery Planning for unknown Objects

Fast manipulation of unknown objects is a basic ability on the way to the robotic co-worker, because it is mostly impractical to have a model of each object in the environment. This paper addresses the problem of grasping and delivering unknown objects by an industrial robot from a table using one single camera image. The delivery planning for each object is done with respect to the packing problem, with the additional constraint that the objects must be regraspable after their delivery. The aim is to build up a system, which has for example the ability to tidy up a table. For this aim, we present two planning algorithms, one for the grasp planning and one of the delivery planning. The performance and robustness of the system is validated by test runs with several objects.