Touch sensing for humanoid robots

Abstract

A new generation of humanoid robots is emerging to work together with, or even replace, human operators performing complex dextrous manipulation operations in a variety of applications such as health and elder care, hazardous or high-risk environments, telemedicine, or manufacturing. To meet the challenging operational requirements of such applications, this new generation of humanoid robots should not only look as humans, but should also behave like them, being able to sense and perceive the external world and perform tasks as humans do. Touch sensing and perception is essential when handling objects while working on such complex activities in unstructured environments. The major challenges encountered when replicating the human touch sensing mechanisms are due to the inherently low resolution of the tactile images produced by the artificial sensors, to the complexity of interpreting the sensor data, and to the fact that robot hand technology is still clumsy when compared with the nimble dexterity of the human hand and fingers. This paper presents practical touch sensing solutions for humanoid robots (Fig. 1) that mimic the complex sensing mechanisms occurring in a human hand while exploring by touch 3D objects.