Touch sensing is helpful and important for a robot to understand and have bidirectional physical contact with its environment, particularly objects it interacts with. Sensors provide information to robot systems about their surrounding environment in order to help them to achieve dynamic interaction tasks, as an example in bionic prostheses, in teleoperation or in robots that should autonomously take decisions according to the situation. To this end, several tactile and vision-based solutions have been adopted in robotics over the past few decades enabled by the improvements of the functional and metrological performances of state of the art sensors, which have been widely reported in the literature. In an effort to gather and summarize a selection of scientific and technological advances in this field, this brief review proposes a categorization based on addressed tasks with tactile sensors. Specifically, it outlines a selection of current trends in material and shape recognition, and encoding of textural and compliance properties, by sensorized robots. Selected tactile sensor technologies are discussed with their applications to achieve specific tasks, also in combination with other sensors. Finally, touch exploration strategies are discussed briefly. This review identifies unresolved challenges and suggests future trends for the application of tactile sensors in various jobs.

Keywords: tactile sensors, material recognition, robots, touch, computer vision, robotics.

Authors: Syed Ali Hassan; Calogero Maria Oddo.

Abstract presented during the 17th edition of IEEE International Symposium on Medical Measurements and Applications – MEMEA 2022, 22-24 June 2022, Messina (Italy).

Paper published by IEEE. Read the accepted publication here: https://zenodo.org/record/7648834#.Y-9Te3bMI2w