TOUCH-DRIVEN OBJECT IDENTIFICATION AND MANIPULATION

Abstract

Being in human environments, the robots need to know how to manipulate and interact with humans and different objects. To successfully implement this task, the prior knowledge of an object’s properties must be known. We designed a new type of tactile sensor based on the active vibro-feedback for object stiffness classification and integrated them into a gripping system of the Shadow Robotic Hand for fine manipulation. The project was logically divided into two parts. First, we designed an object classification system to categorize plastic and silicone cubes based on their stiffness properties using active tactile sensing. Then, having the prior information about the testing objects’ characteristics, the Shadow Hand was programmed to perform a rolling motion with the cubes, showing dexterity and precise control of the robotic system. The project is intended to combine signal processing and robot control to realize stiffness classification for adaptive grasp and tactile servoing for fine motion manipulation.