Slip-Aware Robotic Handling: Tactile Object Characterization for Safe Manipulation

Abstract

Slip is a significant challenge in manipulative tasks involving object gripping. Most research has focused on hard or deformable objects rather than simultaneously addressing both. This study develops and implements a robust slip detection mechanism that effectively identifies slips across various types of objects. The study also introduces a method for classifying objects based on tactile sensor images to determine whether they are hard or deformable. A gripper control method that utilizes this classification to enhance the handling of different object types is proposed.