Sensor Fusion Techniques Used for Learning/Identifying Features of Robot's Part Assembly Tasks

Abstract

In this paper, sensor fusion techniques, fused vision/optical sensors and tactile/vision sensors are introduced, which are used for learning/identifying features of tasks related to a robot's part-bringing procedure from a part position to a position over a hole for the purpose of a micro-part insertion in partially unstructured environments. An entropy function, which is a useful measure of the variability and the information in terms of uncertainty, including learning, is introduced to measure the overall performance of a task execution related to a part assembly. By employing a learning approach, the uncertainty associated with the part-bringing task is reduced. The above tools are necessary for a robot manipulator to perform complex assembly, material handling, manufacturing or machining tasks, for example remote maintenance and hazardous material handling, in an unstructured environment.