Digitalization of Manufacturing Processes: Proposal and Experimental Results

Autor(es)

Rolle, Rodrigo Pita ; Martucci, Vinicius de O. ; Godoy, Eduardo P. ; IEEE

• Tipo de documento: Artigo Científico
• Fonte: 2019 IEEE INTERNATIONAL WORKSHOP ON METROLOGY FOR INDUSTRY 4.0 AND INTERNET OF THINGS (METROIND4.0&IOT); v. N/A, p. 6-pg., 2019-01-01.

Resumo

The Industry 4.0 is a new industrial paradigm that aims to fulfill the needs for more reliable, flexible and efficient industrial processes by implementing digital technology on the shop floor. The development of smart devices, new software tools and communication protocols makes it possible to connect real machines and instruments to the virtual space, enabling more sophisticated control and even future predictions. Digital Twins are an approach for intercommunicating physical and virtual machines or systems, whose main goal is to improve performance on the real system by using information from virtual tools that simulate the physical parts. Typical applications involve performance analysis, bottleneck detection, failure prediction and others. The ambitious aim of replicating the behavior of whole machines or systems also brings a lot of challenges: modeling and simulating complex systems with acceptable computational costs, assuring real-time communication and developing methods for deep analysis are some goals for researchers and vendors. This paper presents an architecture proposal for practical implementation of digital twins, that is based on an open-source tool for process control, lightweight communication protocols and flexible tools for modeling and 3D visualization. The implementation is meant to make the platform as general as possible, so that a myriad of machines and production systems can be modeled and represented on the digital twin architecture. (AU)

• Processo FAPESP: 18/19984-4 - Arquitetura Orientada a Serviços para Aplicações de Internet das Coisas Industrial e Indústria 4.0
• Beneficiário: Eduardo Paciência Godoy
• Modalidade de apoio: Auxílio à Pesquisa - Regular