ABSTRACT
This chapter focuses on the development of a novel large-area soft artificial skin with integrated Fiber Bragg Grating (FBG) sensors for the robotic perception of physical interaction. As interactions are becoming inevitably closer between robots and humans within these tasks, it is logic to equip them with a sort of tactile sensory experience. The localization and force prediction task based on the cross talk of adjacent sensors is a highly-challenging problem that is hard to model. The authors had designed and developed a polymeric artificial sensing skin, based on the FBG technology, with the aim to mimic the human sense of touch. The experimental platform and datasets used to assess the capability of the artificial skin in perceiving the applied loads, and the neural network structures evaluated during the experiments. To determine the neural network structure that suitably tackles a given problem requires some iteration.
