Files

Abstract

A traditional paper guide for industrial maintenance tasks has been transformed into a digital one composed of three augmented reality supports: augmented reality glasses, spatial projection, and video stream superimposition. For the last two, the recording is done by a 6 degree of freedom (trasnlation and rotations) tracking algorithms by deep learning on synthetically generated data. Tests in an industrial environment showed very good results and also that each support has its own advantages depending on the to-do task.

Details

Actions

Preview