Development of a multi-modal tactile force sensing system for deep-sea applications

Abstract

With the increasing demand for autonomy in robotic systems, there is a rising need for sensory data sensed via different modalities. In this way system states and the aspects of unstructured environments can be assessed in the most detailed fashion possible, thus providing a basis for making decisions regarding the robot s task. Com- pared to other sensing modalities, the sense of touch is underrepresented in today s robots. That is where this thesis comes in. A tactile sensing system is developed that combines several modalities of contact sensing. The use of the tactile sense in robotic grippers is of great relevance especially for robotic systems in the deep sea. Up to now manipulation systems in master-slave control mode have been used in this area of application. An operator performing the manipulation task has to rely on visual feedback coming from cameras. Working on the ocean s seafloor means having to cope with conditions of limited visibility caused by swirled-up sediment.