Accented visualization in digital industry applications

Abstract

The paper proposes a new approach of accented visualization useful to develop system architectures implementing interactive user interfaces in digital industry applications. The proposed solution is suitable for image data processing, analysis, virtualization and presentation based on Augmented Reality and the Internet of Things. Accentuated visualization is based on adaptive construction and virtual consideration of the content of the current real scene in the field of view of a person, as well as the viewer’s experience that contains perceptions, points of view and expected behavior. The proposed approach was implemented in a specialized intelligent system for manual operation control. Such a system implements the ideas of Industry 4.0 for smart manufacturing by introduction of cyber-physical decision-making support. The overall solution is used to identify gaps and failures of operator in real time, predict possible operating mistakes and suggest better procedures based on comparing the sequence of actions to an experience of highly qualified operators captured in knowledge base. There are presented the results of solution industrial implementation using neural networks and AR accented visualization in practice.