Digital twin-based assembly data management and process traceability for complex products
Introduction
Along with the accelerating development of aerospace and aviation technologies, in recent years, the number of production tasks for aerospace and aviation products has also increased dramatically, resulting in rapid growth in the amount of product data. The collecting and sorting of this massive amount of data in real-time, and structurally managing and analyzing these data to fulfill the process traceability and continuous improvement of product design and process, have become key data management topics in the aerospace and aviation sector.
Product Data Management (PDM) systems [1] and Product Lifecycle Management (PLM) systems [[2], [3], [4]] have been adopted to achieve product assembly data management and process traceability. The PDM system mainly implements the data management and process management in the product development phase. The PLM system, which is product-centric and an extension of PDM, integrates and manages the product lifecycle data on a unified platform. It enables users to collaboratively design, manufacture, and manage products at all stages of the lifecycle [5]. In terms of PLM-based data management, Sudarsan built a product information modeling framework, so that the product information could be acquired, stored, used, and reused at all phases of the product lifecycle [6]. By using an open standardized product metadata model and Service-Oriented Architecture (SOA), Srinivasan built a framework that integrated business and engineering processes to support PLM. The feasibility and effectiveness of the presented framework was provided by two cases in the enterprise [7]. Lentes constructed an ontology-based platform for manufacturing engineering and PLM, which also can promote the reuse of knowledge [8].
The research described above manages distributed heterogeneous data by constructing a single bill of materials (BOM) that acts as a single product data source or an integrated information framework throughout the product lifecycle. However, most of the research in this area is still at the theoretical level, and little of it focuses on managing the large amount of dynamic data generated in the reverse production processes. It is difficult for research to meet the high standards and requirements of dynamic data management for complex products. There are two reasons for this.
- (1)
The assembly process of complex products involves high complexity, strong dynamics, and many uncertainties, especially in the model development stage. The assembly cycle can be as long as 10 months and sometimes requires repeated adjustments. In addition, the instability of the design process often leads to frequent engineering changes and rework. This is why managing and tracing the design, process, and execution data that is dynamically generated during the assembly process has always been a challenge for product assembly enterprises.
- (2)
The requirements of assembly quality control for complex products are stringent, and great value is placed on process traceability, quality status analysis, and continuous improvement of the product design. Achieving rapid assembly process tracing and data value mining based on assembly data management is another important topic.
In order to better express and utilize product lifecycle data and information, Grieves introduced the concept of “virtual digital representation equivalent to physical products” in a PLM course in 2003 [9]. Later, this concept was named Digital Twin (DT) in 2011 and has since attracted the attention of researchers and companies worldwide in various fields [10]. To better understand the concept of DT, NASA and the U.S. Air Force Research Laboratory (UAFRL) pointed out that a DT was an integrated multi-physics, multi-scale, probabilistic simulation of an as-built system that utilizes the best available physical models, updated sensor data, and historical data, to reflect the condition of the corresponding flying twin [11]. Zhuang and Liu [12] reviewed the background of DT and systematically expounded the connotation of Product Digital Twin. On this basis, the concept of DT technology was proposed, which refers to the process of using digital technology to describe and model the physical entities. A digital twin refers to a virtual mapping model that is consistent with the corresponding physical entity and can simulate and mirror its behavior and performance. It is also known as the Digital Twin Model (DTM). To date, researchers have put forth much effort to apply DT technology in the PLM and product lifecycle [[13], [14], [15], [16]]. In the PLM, Tao et al. proposed a DT-driven product design, manufacturing and service framework, and illustrated the application methods and cases [17]. Grieves presented a DT-based fault prediction and elimination approach for complex systems, which was verified in NASA-related systems [18]. To promote the applications of DT in the future, Qi et al. investigated and summarized the key enabling technologies and tools for DT from the perspective of 5-dimensional DT model [19]. In the product design phase, Tao et al. presented a DT-based product design framework that enables the designers to customize, assess, and accelerate the product development cycle [20]. In the product manufacturing stage, Tao et al. proposed the implementation of Digital Twin Shop-floor and clarified its system composition, operation mechanism, characteristics, and key technologies. This provided a theoretical reference for the realization of cyber-physical fusion on the manufacturing shop-floor [21]. Zhang et al. introduced DT technology to enhance dynamic scheduling and explored the DT-based machine availability prediction, disturbance detection and performance evaluation methods [22]. Söderberg et al. discussed the DT application in the real-time geometry assurance in individualized production, based on an example of a sheet metal assembly station [23]. Zhuang et al. proposed a smart management and control approach for complex products based on DT, constructed its framework and described the core techniques and implementation process in detail [24]. Zhang et al. proposed a rapid custom design and optimization method for the automated production line of the hollow glass driven by DT technology, and built a corresponding DT system through the synchronization between the virtual digital manufacturing system model and physical equipment [25]. Kong et al. presented a data construction method for the applications of shop-floor digital twin system [26]. In the product service stage, Tuegel et al. sought to build a DT for each space vehicle, so as to predict accurately the life of a spacecraft structure [27]. GE implemented real-time monitoring, timely inspection, and predictive maintenance of engines based on DT in its Predix cloud service platform [28].
The above research show the application potential of DT technology in PLM, and production is one of the most popular applied fields. However, there is no corresponding approach to adopt DT in dynamic product data management across multiple stages and process traceability for complex products. As the extension of PLM, product DT emphasizes the data integration throughout the product lifecycle through the product digital model, which provides a single data source for product design, manufacturing, service, and engineering changes. It can record all historical segments and processes of the product lifecycle, thereby providing a new solution for process traceability to meet the strict requirements of quality control and data utilization in enterprises performing complex product assembly. Therefore, to solve the issues that these enterprises face, a DT-based assembly data management and process traceability approach for complex products is proposed in this paper.
The rest of this paper is organized as follows: Section 2 analyzes the evolutionary process of the complex product assembly data. Section 3 presents the framework of DT-based assembly data management and process traceability for complex products. On this basis, Section 4 discusses several key technologies. Section 5 develops a system and verifies the proposed approach by a case study of an assembly enterprise. Section 6 presents the authors’ conclusions and discusses future research issues.
Section snippets
Analysis of evolutionary process for the complex product assembly data
The entire process of generation-to-archive of complex product assembly data is undergoing a series of changes. These changes can be considered granularity, period, and version, which respectively correspond to product structure, product lifecycle, and product data version, as shown in Fig. 1.
A complex product is composed of many assemblies, and an assembly is made up of many parts. Hence, product assembly data is a collection of data from the various components that make up the product in the
Framework of DT-based assembly data management and process traceability for complex products
Product assembly data is composed of data of different versions (version dimension), data of different hierarchies (granularity dimension), and data of different stages (period dimension), including product design, process planning, and assembly execution. To achieve complete and accurate assembly data management and process traceability for complex products, DT technology is introduced and then combined with workflow technology in a framework, as shown in Fig. 4.
First, to achieve the product
Product assembly data organization and version management based on workflow
Complex product assembly data includes data from three stages: product design, process planning, and assembly execution. Achieving the association, organization, and version management of data at each of these stages is the premise of product assembly data management. Hence, we introduce workflow technology to build three models: an organizational model of product design data, an integrated organizational model of assembly process and assembly execution data, and a version association model of
Case study
Based on the above research results, we developed the Digital Twin-based Assembly Process Management and Control System (DT-APMCS) using Microsoft’s. Net Framework 3.5 and Visual Studio 2008 to verify the proposed approach. This system consists of key functions such as management of product assembly BOM, synchronous mapping of the product assembly process, integrated management of product assembly data, generation of product assembly data package, and integration with other systems.
Conclusions and future work
DT technology has attracted the attention of academic researchers and industrial practitioners because of its ability to achieve the fusion of virtual and physical space, and because it shows extensive application prospects. However, due to the characteristics of high complexity, strong randomness, process instability, and extremely strict data management and process traceability for complex product assembly processes, achieving the assembly data management and process traceability of complex
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors would like to express our sincere gratitude to the anonymous reviewers for the invaluable comments and suggestions that have improved the quality of the paper. We also thank LetPub for its linguistic assistance during the preparation of this manuscript. This research is financially supported in part by the National Natural Science Foundation, China (No. 51935003), and in part by the National Defense Fundamental Research Foundation, China (No. JCKY2018210C005, No. JCKY2016204A502).
References (29)
- et al.
Systems integration and collaboration in architecture, engineering, construction, and facilities management: a review
Int J Adv Sci Eng Inf
(2010) - et al.
Data-driven smart manufacturing
J Manuf Syst
(2018) - et al.
A product information modeling framework for product lifecycle management
Comput Aided Des
(2005) An integration framework for product lifecycle management
Comput Aided Des
(2011)- et al.
Blockchain-based data management for digital twin of product
J Manuf Syst
(2020) - et al.
Toward a digital twin for real-time geometry assurance in individualized production
CIRP Ann-Manuf Tech
(2017) - et al.
Developing an agent-based PDM/ERP collaboration system
Int J Adv Manuf Technol
(2006) - et al.
An integrated closed-loop product lifecycle management approach for reverse logistics design
Prod Plan Control
(2016) - et al.
Overview of product lifecycle management
Comput Integr Manuf Syst
(2004) - et al.
amePLM: a platform providing information provision in engineering
Int J Prod Res
(2017)