Bin Picking Pneumatic-Mechanical Gripper for Industrial Manipulators

Abstract

The human body is an advanced and complex machine. A simple task like grasping an object, which is done intuitively, has puzzled researchers for decades. For each object, humans have a suitable grip to grab it optimally and safely. Even for objects which we have never seen before, we can choose the best possible grip and can adapt as the environment changes to get a better grip based on cognition and experience. One of the main capabilities of humans is their ability to learn about unfamiliar things and processes. This capability helps us adapt to different situations and still be able to solve a problem. This works focuses on developing a self-learning robotic system which can replicate the human learning capabilities in a handing-over task. The proposed system consists of two submodules: 1) Vision analysis and environment monitoring, which provides accurate global and local information about the area in which the robot has to hand over the specific object; 2) Safe and flexible bin-picking gripper, which handles various objects with complex geometries. The work is a consortium of 3 partners: IWU Chemnitz (Germany), Novosibirsk State Technical University (Russia) and Technical University of Sofia (Bulgaria) under European program ERA.Net RUS PLUS 2017