Conceiving a Digital Twin for a Flexible Manufacturing System

Abstract

Digitization and virtualization represent key factors in the era of Industry 4.0. Digital twins (DT) can certainly contribute to increasing the efficiency of various productive sectors as they can contribute to monitoring, managing, and improvement of a product or process throughout its life cycle. Although several works deal with DTs, there are gaps regarding the use of this technology when a Flexible Manufacturing System (FMS) is used. Existing work, for the most part, is concerned with simulating the progress of manufacturing without providing key production data in real-time. Still, most of the solutions presented in the literature are relatively expensive and may be difficult to implement in most companies, due to their complexity. In this work, the digital twin of an FMS is conceived. The specific module of an ERP (Enterprise Resources Planning) system is used to digitize the physical entity. Production data is entered according to tryouts performed in the FMS. Sensors installed in the main components of the FMS, CNC (computer numerical control) lathe, robotic arm, and pallet conveyor send information in real-time to the digital entity. The results show that simulations using the digital twin present very satisfactory results compared to the physical entity. In time, information such as production rate, queue management, feedstock, equipment, and pallet status can be easily accessed by operators and managers at any time during the production process, confirming the MES (manufacture execution system) efficiency. The low-cost hardware and software used in this work showed its feasibility. The DT created represents the initial step towards designing a metaverse solution for the manufacturing unit in question, which should operate in the near future as a smart and autonomous factory model.

Digital Transformation in Spanish Port System

“…They expose information on different aspects at a physical level such as performance, technical characteristics, repair history, services, etc. This data provides us with the information to carry out simulations, analysis of improvements, creation or increase of business value, and above all to have technical support for decisionmaking [13].…”

Literature Review And Comparison Of Digital Twin Frameworks In Manufacturing

“…Among them, some frameworks reinterpret this structure, showing explicitly the (Kritzinger et al, 2018) distinction of a "Digital Shadow", a "Digital Model" and a full DTw (Cimino et al, 2019;Negri et al, 2020). Therefore, in general, this cluster collects basic architectures where the DTw is in parallel with the real system, discussing interfaces and interoperability with operations management systems such as MES, ERP and optimization modules (Cimino et al, 2019;Magalhães et al, 2022). Looking at the result within this cluster, we can assume that (Barbieri et al, 2021;Negri et al, 2020;Novák & Vyskočil, 2022;Ragazzini et al, 2021) present similar architectures, revisited with the specific inclusion of optimization modules or intelligent layers.…”

Generalised Performance Estimation in Novel Hybrid MPC Architectures: Modeling the CONWIP Flow-Shop System

“…The ability to supply more individualized market demand in a short period of time and at a lower cost is a vital premise for industrialized countries' competitive recovery against emerging countries with lower technological development and lower social and labor expenses [1][2][3]. The Fourth Industrial Revolution (or Industry 4.0) was born in response to this challenge, with the goal of integrating cutting-edge information technologies, artificial intelligence techniques, and distributed production control approaches [4][5][6][7][8][9].…”

“…Practically, the line throughput for each of the 10 average processing times in {5, 10, 15, 20, 25, 30, 35, 40, 45, 50} was determined by executing 40 runs for each combination of α in {0.5, 1, 1.5, 2, 2.5, 3} and WIP in the range [1,30]. Then, the resulting throughput values were normalized with respect to the related r b (i.e., the inverse of the average processing time under consideration).…”