Eventos Anais de eventos
COBEM 2021
26th International Congress of Mechanical Engineering
MODULAR, SCALABLE AND GENERIC DIGITAL TWIN APPROACH FOR DYNAMIC MANUFACTURING PROCESSES
Submission Author:
Matheus Antonio Nogueira de Andrade , BA
Co-Authors:
Matheus Antonio Nogueira de Andrade, Herman Lepikson, Carlos Alberto Machado
Presenter: Matheus Antonio Nogueira de Andrade
doi://10.26678/ABCM.COBEM2021.COB2021-0728
Abstract
Digital twins are commonly referred as the virtual copy of a physical asset. But there is no common definition regarding this term that is widely adopted. Alongside with other terms in the industry 4.0 context, digital twins are being discussed more and more. The creation and development of a digital twin depend on the physical counterpart that is being mirrored. Products, services, industrial processes and physical phenomena are some of the instances being digitized towards digital twins. Dynamic manufacturing processes are the focus of this work. The potential of digital twins has led to the creation of working groups in associations to develop standards for digital manufacturing environments that can evolve to support the foundation for digital twin implementations. These recent efforts highlight the need for developments of generalizable and scalable solutions for digital twins. This article proposes a modular, scalable and generic structure for the development of digital twins based on the concept of multi-agent systems to support the specificities of different industrial processes in a single configurable constructive structure. Intelligent cooperating agents can perceive the environment in which they are and react appropriately to it, ensuring the modular characteristics of this structure. The ASTM E3012 standard proposes a modular approach for manufacturing processes dividing them into Unit Manufacturing Process (UMP) that can be understood as the smaller part of an industrial process that can perform a significant changing of shape, condition or property to the raw material. The UMP concept was useful in this paper to help understanding how entire industrial processes can be divided into smaller pieces. UMP and multi-agent systems concepts can also help to achieve a scalable objective. As a case study, the digital twin dynamic model of an essential oil steam distillation plant was created to demonstrate the proposed methodology. Essential oils are extracted from plants through several methods. In the steam distillation process, the steam acts as a mean of transportation to the oil. Steam distillation extraction are performed in batches using distillers with discrete quantities of plant material. As a batch process, it presents both continuous and discrete dynamics. This way, the essential oil extraction process is a good demonstration of a generic digital twin. The proposed concept of digital twin can be the basis for the creation of several others.
Keywords
Digital Twin, manufacturing processes, multi-agent systems, essential oil
