Background, Introduction and Motivation -- Part I Avoiding Picking Potentially Entangled Objects -- Deep Learning for Classifying Potential Entangled Objects -- Entanglement Map: A Visual Representation for Entangled Objects -- Shape Reconstruction of Entangled Objects -- Part II Disentangling Manipulation Planning for Entangled Objects -- Affordance Maps for Picking or Separating Entangled Objects -- Learning Efficient Policies for Entangled Wire Harnesses -- Dynamic and Bimanual Manipulation with F/T Feedback for Entangled Wire Harnesses -- Conclusions.

This book introduces methods for bin picking in manufacturing. These methods can be used to develop unified, dexterous, and robust bin picking systems for entangled objects. The target objects include both rigid and deformable objects. Robotic bin picking is a valuable task in manufacturing, aiming to automate the assembly process by utilizing robots to pick necessary objects from disorganized bins. Previous studies have addressed various challenges related to bin picking. However, when objects with complex shapes or deformable properties are randomly placed in a bin, they tend to get entangled, making it difficult for the robot to pick up individual items. This poses challenges in perception, as the robot must be capable of distinguishing between isolated objects and potentially tangled ones in a cluttered environment. This book is of interest to students, researchers, and professionals in manufacturing industries.