Architecture for Digital Twin implementation focusing on Industry 4.0

Industry 4.0 is a new technological wave that dramatically transformed the industrial environment. It consists in taking to the shop floor solutions and concepts from Internet of Things, such as high connectivity, integration and cooperation between distinct elements. The development of smart devices, the continuous evolution of software tools and the implementation of new communication protocols make it possible to implement sophisticated techniques for control, supervision and optimization of industrial processes. Digital Twins are an approach for integrating physical systems to virtual counterparts that replicate in the virtual space the operation of real instruments and machines. Its main objective is to improve the real system performance by using information generated by the virtual counterpart, for example, by executing performance analysis, bottleneck identification, failure prediction, detection of deviations and others. Some of the biggest challenges on this context are creating trustworthy models with acceptable computational costs, assuring real-time communication and developing deep analysis methods. This paper presents the development of a modular architecture for implementing Digital Twins. The architecture is based on an open-source tool for process control, lightweight protocols and flexible tools for modeling and 3D visualization. Tools were chosen in order to allow diverse machines and productive systems to be modeled and represented on the Digital Twin architecture.