Tactile Sensing, Skill Learning, and Robotic Dexterous Manipulation

Chapter 8: Learning form-closure grasping with attractive region in environment

Rui Li^a; Zhenshan Bing^b; Qi Qi^a    aSchool of Automation, Chongqing University, Chongqing, China^bDepartment of Informatics, Technical University of Munich, Munich, Germany

Abstract

Robotic grasping has long been a difficult yet fascinating research topic in robotics. For the last 30 years, synthetic methods have achieved encouraging results in building a firm and stable grasp. Nevertheless, such methods have experienced significant challenges when friction, uncertainty, and generalization are considered during the grasping process. Learning methods, on the other hand, have shown promising outcomes in teaching robots to grasp a large variety of unseen objects.

Usually, ...