Deformable Object Recognition Using Proprioceptive and Exteroceptive Tactile Sensing

Abstract

The somatosensory sense is a combination of several perception systems. It plays an important role in the exploration and recognition of the environment. In this work, we show how proprioceptive (internal) and exteroceptive (external) sensing can help to distinguish deformable and rigid objects. First, we describe our planar robot arm, which serves as the experimental testbed for the integration of internal and external sensors. The internal sensing system consists of the motor current sensors and the joint encoders of the planar robot arm. The external sensing system is a pressure sensing tactile array with five sensors. We use Robot Operating System (ROS) in order to integrate these systems and control our planar robot. We also present our pipeline for recognition of object deformability based on these sensor measurements. The performance of our approach is demonstrated using a set of experiments in which we distinguish soft and rigid of objects of different sizes.