Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019

IFAC PapersOnLine 58-8 (2024) 1–6 ScienceDirect ScienceDirect Available online at www.sciencedirect.com 2405-8963 Copyright © 2024 The Authors. This is an open access article under the CC BY-NC-ND license . Peer review under responsibility of International Federation of Automatic Control. 10.1016/j.ifacol.2024.08.041 Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: [email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: [email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: [email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and 10.1016/j.ifacol.2024.08.041 2405-8963 Copyright © 2024 The Authors. This is an open access article under the CC BY-NC-ND license ( https://creativecommons.org/licenses/by-nc-nd/4.0/ ) Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: ajguil[email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: [email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and Application of Digital Twin Technology for the Digitization of Railway Maintenance Services in compliance with European Regulation EU 779/2019. Guillén-López, A.*, Sanchez-Herguedas, A.**, Rodríguez M.** * Department of Management, Complutense University of Madrid, Spain. (e-mail: [email protected]). ** Department of Industrial Management, School of Engineering, University of Seville, Spain. (e-mail: [email protected], [email protected]) Abstract: This paper delves into the application of Digital Twin Technology (DTT) in the digitalization of assets, processes, and services to improve maintenance performance in railway applications in the context of European Regulation EU 779/2019. Digitalization plays a pivotal role in certifying the suitability of rolling stock in the railway sector. The primary objective of this study is to meticulously examine the fundamental tasks required to build a digital twin in this specific context. The process involves the contributions of key entities fulfilling specific roles: the maintenance and asset certification service provider ensures the integrity and compliance of rolling stock; specialists in IoT and process automation, along with experts in software platforms and services, are tasked with implementing the technological aspects of the digital twin. As a result, a methodology is proposed for the integration of digital design that combines three approaches: Service approach, platform architecture, and model-based asset digitalization. This paper delineates critical stages, from data acquisition and management to model creation and information integration, highlighting the crucial role of DTT in this process. Furthermore, the challenge of digitalization in situations where assets and processes are in an initial state of low or non-existent digitalization is underscored. A profound understanding of digital assets, information layers, and underlying models becomes crucial in the context of the current challenges of railway digitalization. Keywords: Digital Twin, Digital Twin Technology, Railway digitalization, Digital Maintenance, Rolling Stock, Regulation EU 779/2019 1. INTRODUCTION This document presents a research and development proposal on the digitalization of highly specialized technical services. The aim is to investigate the design, development, and utilization of Digital Twin Technology (DTT) in the digitalization of maintenance services, specifically within the railway sector. The research is grounded in the DF-MAS project, addressing the challenge of digitally transforming systems with an initial level of low or nonexistent digitalization, as is the case with many railway applications. This paper sheds light on a digitalization process that leverages DTT, emphasizes the specific responsibilities of key roles, and addresses the challenge of transforming assets and processes. It also establishes the foundation for future development related to the skills necessary to successfully create digital twin solutions. This work is part of the digital transformation of the railway transport sector, critical due to security reasons, and pivotal in the challenge of energy transition. The main objective is to digitize the maintenance service for a fleet of machinery used in railway infrastructure conservation. The fleet includes various types of equipment: ballast regulators, profilers, unloading machines, stabilizers, track inspection cars, and wagons. These are heavy machines with high technical complexity, incorporating different systems and subsystems (operation, traction, and auxiliary) subjected to demanding working conditions, leading to high wear and failure rates. This results in elevated maintenance and operational costs (OPEX), providing an opportunity for optimization and management through digitalization. Fleet maintenance occurs in a workshop, making digitalization involve both machine analysis and an examination of the facilities and technical and human resources within the workshop. This service adheres to European safety regulations for rolling stock. This regulation proposes a maintenance management structure that identifies four maintenance functions or ECM (Entity in Charge of Maintenance) (see Figure 1): ECM1 (Global Management), ECM2 (Engineering), ECM3 (Fleet Management), and ECM4 (Workshop) (Regulation (EU) 2016/798, 2016). Different decisions are made in each of these functions. To digitize these decisions, input data must be processed according to the models used (ML, simulation, optimization, etc.) in each function. Decision-making allocation becomes key in digitalization design. The main advantage of DTT, as proposed in the project, is providing a suitable framework for addressing advanced digitalization solutions for assets, processes, and services. It includes designing a common software architecture for integrating various data sources (sensors, information systems, IoT networks, etc.) and the specific design and treatment within this architecture of models that digitally define these assets, processes, and services. The architecture also integrates functions such as data processing, data storage, predictive analytics, task scheduling, simulation, optimization, and 2 A. Guillén-López et al. / IFAC PapersOnLine 58-8 (2024) 1–6 representation. An integration approach with three elements: functions, models, and architecture, shapes the digital twin environment. Figure 1. DT conception, service-oriented and linked EU 779/2019 The digitalization process is complex and must be structured in phases. In this case, three one-year phases have been employed, incorporating technical and service aspects progressively. From the analysis of the four safety functions (Figure 1) and their management needs, the contents to include in each phase are established. The first phase addresses the basic design of the architecture, the IoT solution for data entry into the system, and the Fitness For Service Certification (FFSC) of rolling stock (Regulation (EU) 2019/779, 2019). This service is part of function ECM4 and ensures the safety of rolling stock. In the first phase, the systems of a track maintenance machine (ballast regulator) have been analyzed, and a proof of concept has been designed for one of the critical systems (Axles and wheels system). The second phase includes the remaining safety systems of the machine (13 systems), and in the third phase, the remaining machines will be included. The modeling of the workshop and its resources progresses similarly and in parallel to the needs of each phase. 2. BACKGROUND Digital Twin Technology (DTT) introduces new concepts and tools for addressing challenges within workshops. Digital models are employed to monitor, simulate, predict, and control maintenance management in a virtual environment. Furthermore, they enhance the capabilities of physical assets through interaction and feedback between the virtual and real environments, data fusion and analysis, decision optimization, and more. Digital technology acts as a bridge between the real world and the information world, forming the basis for data tracking of product, process, and service quality, fostering continuous improvement (Hu et al., 2020). Digital twins extend the simulation of physical assets throughout their entire lifecycle, supporting decision-making across the product and process lifecycle. Advances in hardware and software have allowed for more sophisticated machine design processes, enabling precise models that save costs and time (Mendes and Ferreira, 2021). The ability to use various sensors to monitor multiple parameters, coupled with advancements in the Internet of Things (IoT) and Big Data, has facilitated the development of these new technologies. Given these technologies, the digital twin emerges as a natural advancement and is expected to become a necessary element for the industry. The two macro elements to connect are the virtual space and the real space. This involves collecting data, information, and processes from a physical element and implementing them in a virtual version. This allows for examination, study, simulations, and predictive analyses. Future digital twins are expected to be interconnected, creating twins of larger systems. The main idea of intelligent products is to incorporate self-management capabilities. The digital twin should respond immediately to external changes (Monsone, 2021), anticipate and prevent issues, promptly resolve problems, make necessary changes in the virtual world to ensure the physical asset operates as expected, save time and resources in simulation phases, improve digital twin performance, and conduct durability tests. Data plays a crucial role in DTT, particularly in constructing and integrating digital models of assets (machines and workshops). The literature emphasizes data as a central element of digitalization. However, studying data requires structuring this data using models to avoid recording errors and enable statistical control of information. Challenges in data modeling include integration with digital twin models and synchronous updates between them. The Industrial Internet Consortium (IIC) (Malakuti et al., 2020) defines data, models, and services as elements of digital twins. Various data modeling methods have been proposed for workshops and equipment to be maintained. In general, all models represent a digital system framework with three layers: the physical layer, the virtual layer, and the application layer. The physical layer is a significant part of the digital system, serving as the source of real data for management processes, tools, personnel, spare parts, equipment, and the environment. The virtual layer is the core of the digital system, digitally reconstructing workshop or equipment activities in the virtual space, supporting data analysis, mechanism modeling, and control model construction, and completing the integration of data loaded from the physical layer. The application layer is based on presenting global data in the virtual layer, combined with real-time information from the workshop's production process, to achieve global optimization of the workshop process driven by the digital twin (Qian et al., 2021). Data is key to the operation of the digital workshop and supports applications. However, data has coupling characteristics and, due to its large quantity, leads to inefficient and inaccurate application operations. To provide stable and efficient data support for digital twin applications, (Kong et al., 2021) propose a data construction method. The data construction framework is designed based on functional requirements and analyzed according to manufacturing data characteristics. Maintenance is one of the activities where digital twins are applied. Many cases propose improving maintenance as one of the main justifications for digital twin development, and there are digital twins developed solely for this purpose. Maintenance has specific data needs to define assets to be maintained and their tasks. This is linked to activities such as A. Guillén-López et al. / IFAC PapersOnLine 58-8 (2024) 1–6 3 controlling the degradation of failure modes, scheduling and controlling task execution, developing and optimizing maintenance plans, etc. (Abbate et al., 2022) propose a methodology to assess the behavior of maintenance systems through the analysis of real vibration data collected in a workshop. They develop a digital twin-based approach and integrate it with the statistical method of control charts. The methodology is applied to a real case study of an electric motor. Implementing configurable digital twins that can flexibly adapt to changing industrial configurations remains a challenge due to the need to integrate multiple fragmented data sources. (Christou et al., 2022) present the architecture, design, practical implementation, and evaluation of a comprehensive platform addressing these challenges. The platform provides the means to collect, manage, and route data streams from heterogeneous cyber-physical systems, in a configurable and interoperable manner. They present a data model scheme and a model processor scheme. (Kim et al., 2022) propose a method to use the Asset Administration Shell along with ISO 15926 to facilitate maintenance data exchange. They define the operation and maintenance (O&M) system framework and information requirements. Based on this, they design the maintenance data structure and apply reference data to identify equipment types and properties for each equipment type using ISO 15926. Following the pre-designed data structure, they develop a Shell-based neutral format for exchanging maintenance data between O&M software systems. Although the workshop scheduling process can be observed, analyzed, and controlled in real-time by the digital twin model, there are uncertain events that cause abnormal disturbances in the physical workshop. This leads to deviations between the digital twin's performance and the physical workshop, affecting the accuracy of predictions. (Qian et al., 2023) indicate the need for a synchronous evolution of data provided to the digital twin to ensure its accuracy and the workshop's performance. In their paper, they outline how to effectively construct synchronous updating methods of the digital twin model based on dynamically generated sample data in workshops with discrete scheduling. The design of the digital twin requires defining the functional structure of the equipment. From there, components to be digitized can be considered, which, when aggregated, lead to the digital twin of the machine. Additionally, interventions of maintenance and operation, and the characteristics of each, as well as the workshop resources needed for interventions, must be added. To address modeling, simulation, and verification challenges in digital twins of workshops, (Zhang et al., 2022) propose an aggregate modeling method for digital twin workshops. The modeling method consists of adding several digital twins: digital twins of equipment, materials, and operators make up the digital twin of the production line. This, together with the warehouse twin, forms the digital twin of the workshop, establishing a multi-level digital twin model. Optimal policies for predictive, preventive, corrective maintenance, and even opportunistic maintenance play a significant role in the success of maintenance operations. (Bányai, 2021) analyzes a new method for optimizing maintenance policies related to energy efficiency. The method uses failure data and state information from both the physical system and the digital twin-based discrete event simulation. The study presents a functional model for optimizing maintenance policy, a mathematical model, and an algorithm for finding optimal solutions. (Peng and Zheng, 2022) investigate key technologies for a digital twin-based management and control system for a workshop. They introduce a digital twin between the enterprise management layer and the workshop management and control layer through a fuzzy rule neural network. The study's results confirmed that integrating the digital twin into the conventional workshop management and control system brought changes in system composition, processes, and information integration. The use of Building Information Modeling (BIM) for asset management in the architecture, engineering, construction, and owner-operated sectors has become popular. Asset management based on BIM has increasingly attracted researchers' and professionals' attention. (Lu et al., 2020) provide a comprehensive review and analysis of recent research and the development of industry standards affecting BIM and asset management in the operations and maintenance (O&M) phase. However, BIM is not always sufficient for asset management throughout the entire lifecycle, especially in the O&M phase. For this reason, they propose a framework for the future development of intelligent asset management, integrating the concept of the digital twin. (Heaton and Parlikad, 2020) present a BIM-based approach for designing and developing a digital twin. The approach uses the objectoriented aspect of a BIM model to create an Asset Information Model (AIM), supporting the creation of the digital twin. Based on a case study of the West Cambridge campus, the article illustrates the development of a single 3D model for integrating multiple BIM models with abundant metadata. (Daniotti et al., 2022) develop a BIM-connected management system with a set of operational and multifunctional tools for various actors in architecture, engineering, and construction. They integrate these tools to improve BIM adoption in building renovation environments based on an information flow between activities. This close relationship between BIM and real-time data brings the project development closer to the context of the digital twin. (Wei et al., 2022) develop a holistic model of digital twins for off-site construction and an evaluation framework to help off-site construction companies systematically approach digital twins. By applying this holistic model to off-site construction processes, they aim to improve the exploration and application of technological advances. 3. METHODOLOGY FOR DIGITAL INTEGRATION In this study, a methodology for integration has been applied, combining three complementary approaches: (i) business/service digitization approach, (ii) technological approach to architecture development and platform proposal & (iii) asset digitalization based on models. Business/Service Digitization Approach, A business/service digitization approach has been employed, aligning with the division of safety functions as illustrated in Figure 1. This approach ensures a detailed focus on critical functions identified in the EU 779/2019 regulation. Following this framework, starting from business requirements, a study has 4 A. Guillén-López et al. / IFAC PapersOnLine 58-8 (2024) 1–6 been conducted on the necessary and available data, including the technology to be deployed to manage data input. Additionally, a software architecture vision has been formulated to support the functionalities of the digital twin. Technological Approach to Architecture Development and Platform Proposal: This technological approach involves the development of an architecture and the proposal of a platform, ensuring the capability for the integration of systems and models both presently and in the future. A comprehensive approach to integrating technologies and digital capabilities has been designed. Reference architectures for the development of advanced digitalization solutions, such as RAMI 4.0 (Reference Architectural Model Industrie 4.0) and IIRA (Industrial Internet Reference Architecture), have been studied and adapted. These architectures are widely employed in the literature for the development of Digital Twins (DTs). In terms of data management, the proposal works along two lines (Figure 2). The first line has focused on identifying the technologies used for data generation and supply, along with their integration strategy through an IoT Smart Bus Microservice. An application has been developed to handle various types of transactions, including monitoring, GPS, tags (RFID, NFC, BLE, QR code, etc.), CMMS (Computerized Maintenance Management System), and manual data entries by technicians, among others. The utilization of a commercial CMMS (Computerized Maintenance Management System) as an entry point for the basic digitization of maintenance is a key contribution in this phase. It has enabled the definition of the hierarchical (technical) structure, asset identification, and introduction of maintenance protocols in a language natural to maintenance personnel. This strategy implies that the CMMS will not be replaced by the Digital Twin but will be integrated as a part of the DT. This completes an IoT/IoE solution that enables full connectivity of the Digital Twin (data input and output). The second line involves modeling transactional events. The decision has been made to disaggregate data transactions towards the platform at the most granular level possible, using individual data transactions. The goal of this model based on individual transactions is to exploit the full potential of the collected basic data and generate new data (electronic data) from them. This approach allows for eventdriven data handling, enhancing data update efficiency and subsequent utilization in subsequent derived calculations. The central part of Figure 2 illustrates how input-output transactions are modeled as events. To manage data storage and host the Digital Twin, a data management platform based on the "Lakehouse" concept has been designed. It allows for storing structured, semi-structured, and unstructured data, similar to a Data Lake, while also providing data quality, performance, security, and governance similar to a Data Warehouse, all within a single platform. It is a layered infrastructure composed of multiple technologies spanning data ingestion, storage, processing, and consumption. Technologies and tools have been determined for the various layers, ensuring sufficient flexibility and capability to support the evolutionary development of the DF-MAS Digital Twin concept. This architecture will facilitate the development of intelligence and representation/interface functionalities as part of the twin. Figure 2. Design of the SMART BUS for individual data input/output by events Asset digitalization based on models. This methodology has been completed by incorporating a specific framework (Crespo Márquez, 2022) as a reference for the sequential development of content and models for the digitization of assets and their maintenance (Figure 3). The asset, as a key element to digitize, is conceived as a set of models. Each model represents specific, complementary aspects of the asset and its behavior. This asset-centric view aligns with the concept of Asset Administration Shell (AAS). The employed framework organizes the development process of the necessary models. Not all assets require the same level of model development, and certain models need to be utilized before others. Within the models, there is, on one hand, the basic data model of the asset. This involves the use of various approaches and/or description tools for assets (class models, technical structure models, operational context models, graphical models, etc.). On the other hand, there are decision support models (interpretation rules, predictive analytics, optimization, simulation, etc.). The latter are related to the generation of new capabilities that enhance the development level of the Digital Twin's intelligence, improving performance in problem-solving, such as task scheduling in the workshop for the vehicles in a fleet, for example. The framework mandates the establishment of a roadmap in determining the various models that can be used, providing an overarching view to understand the scalability of the digitization strategy, both horizontally (new contents) and vertically (application to more assets and/or services within the business). A. Guillén-López et al. / IFAC PapersOnLine 58-8 (2024) 1–6 5 4. USE CASE The use case examines the maintenance of large railway infrastructure conservation machines. These are railway rolling stock and are subject to all European safety regulations. The aim is to apply digital twin techniques to digitize the process of obtaining the FFSC as part of the ECM4 function. This will enable the digitalization of the machines, the personnel working in the workshop, and its resources. It will provide full real-time traceability of a critical process, given its safety implications. Additionally, it facilitates the introduction of data capture techniques to automate part of the process and the development of predictive techniques for monitoring the degradation and expiration of machine elements. Finally, it involves developing planning and scheduling models for maintenance tasks that leverage all available data and digital capabilities. The FFSC is the document that authorizes railway rolling stock to operate on the RFIG (General Railway Network). Obtaining the FFSC requires carrying out a series of safety inspections, verifying compliance with technical limits established by regulations, and controlling the traceability of documentation and the process. It includes FFSC status and expiration control, as well as planning and execution of inspections. The choice of the FFSC case has two motivations: (i) it is a critical service due to its impact on safety and the company's financial results, and (ii) the complexity and richness of the case have been evaluated, particularly considering the involvement of the three defined asset types (machine, workshop equipment, and personnel) and includes a very comprehensive typology of data and data inputs, allowing for the analysis of asset definition data, maintenance tasks, and meters (both offline and online). In the use case, we have focused on a tamper-type machine, analyzing the maintenance procedures included in the FFCS processes of a critical system within these machines, Axles and wheels system. The FFCS mandates control over 13 machine systems that influence circulation safety. Initially, the methodology outlined in section 3 has been proposed for implementation, albeit for just one of these systems in the initial phase. For this system, the elements that compose it, the tasks to be verified, human resources, the tools needed to carry them out, and the spare parts to be replaced with their expiration dates data models are determined. IoT monitoring and data acquisition solutions for the machine, system, and human and technical resources are designed and integrated into a single platform and process, based on individual digital transactions. A computing edge structure is achieved, enabling the integration of any type of data input and interfacing with information systems, particularly traditional CMMS. This is crucial as the solution aims to generate new capabilities without replacing CMMS, which will continue to serve as the interface for workshop technicians with the system. This is complemented by the development of a cloud architecture that collects unstructured information in a lakehouse and reorganizes and reprocesses it according to the asset digitalization and maintenance models approach outlined in section 3, thereby generating a solution oriented towards the digital representation of assets and maintenance. These data are hierarchically structured based on a triple hierarchical encoding (element identification, element location, and quantified variable value). In a second phase, the model development is extended to the rest of the tamper's systems. The intervention scheduling models for the fleet of equipment in the workshops and the optimization models of preventive interventions are proposed and validated. The input and output data of these models are determined: the optimal scheduling of tasks and their frequency of execution. The programming and optimization models are executed in it, so that the digital twins can be represented with the values that are online transferred, and their evolution can be predicted from the data coming from the execution of the programming and optimization models. 5. CONCLUSIONS The DF-MAS project has involved a detailed and multidisciplinary analysis, resulting in the formulation of a basic concept of Digital Twin Technology (DTT). This analysis covered various aspects necessary for the application of digital twins in the case study, including: • Evaluation of technical elements and procedures of the machine maintenance service for track maintenance. • Study of data and information needs in the context of a digitized and partially automated process. • Analysis of available technologies for data generation. • Review of prediction and optimization models applied to maintenance. • Consideration of elements of advanced technology architecture to support integration. This synthesis process, based on the analysis, has led to the determination of a project vision regarding "what" the technology of digital twins is and "how" it can be applied, highlighting the potential improvement that this concept introduces compared to conventional digitization strategies. In the short term, this improvement focuses on incorporating advanced digital capabilities for local service improvement. However, in the long term, and with a much more ambitious vision, it opens up the possibility of much more powerful developments based on the ability to interact with other twins in global twin environments (in the case of railways, interacting with other digital twins of the railway system agents, particularly with the digital twins of railway infrastructure). Figure 3. Framework for sequential model development for asset digitization and asset maintenance 6 A. Guillén-López et al. / IFAC PapersOnLine 58-8 (2024) 1–6 REFERENCES Abbate, R., Caterino, M., Fera, M. and Caputo, F. (2022). Maintenance Digital Twin using vibration data. Procedia Computer Science, 200(2019), 546–555. https://doi.org/10.1016/j.procs.2022.01.252 Bányai, Á. (2021). Energy consumption-based maintenance policy optimization. Energies, 14(18). https://doi.org/10.3390/en14185674 Christou, I. T., Kefalakis, N., Soldatos, J. K. and Despotopoulou, A. M. (2022). End-to-end industrial IoT platform for Quality 4.0 applications. Computers in Industry, 137, 103591. https://doi.org/10.1016/j.compind.2021.103591 Crespo Márquez, A. (2022). Digital Maintenance Management Guiding Digital Transformation in Maintenance (2022nd ed.). Springer Nature Switzerland AG. Daniotti, B., Masera, G., Bolognesi, C. M., Spagnolo, S. L., Pavan, A., Iannaccone, G., Signorini, M., Ciuffreda, S., Mirarchi, C., Lucky, M. and Cucuzza, M. (2022). The Development of a BIM-Based Interoperable Toolkit for Efficient Renovation in Buildings: From BIM to Digital Twin. Buildings, 12(2). https://doi.org/10.3390/buildings12020231 Heaton, J. and Parlikad, A. K. (2020). Asset Information Model to support the adoption of a digital twin: West Cambridge case study. IFAC-PapersOnLine, 53(3), 366–371. https://doi.org/10.1016/j.ifacol.2020.11.059 Hu, C., Gao, W., Xu, C. and Ben, K. (2020). Study on the Application of Digital Twin Technology in Complex Electronic Equipment. In B. Duan, K. Umeda, & W. Hwang (Eds.), Proceedings of the Seventh Asia International Symposium on Mechatronics. Vol II (pp. 123–137). Springer. https://doi.org/10.1007/978-981- 32-9441-7 Kim, B., Kim, S., Teijgeler, H., Lee, J., Lee, J. Y., Lim, D., Suh, H. W. and Mun, D. (2022). Use of Asset Administration Shell Coupled with ISO 15926 to Facilitate the Exchange of Equipment Condition and Health Status Data of a Process Plant. Processes, 10(10). https://doi.org/10.3390/pr10102155 Kong, T., Hu, T., Zhou, T. and Ye, Y. (2021). Data Construction Method for the Applications of Workshop Digital Twin System. Journal of Manufacturing Systems, 58(PB), 323–328. https://doi.org/10.1016/j.jmsy.2020.02.003 Lu, Q., Xie, X., Heaton, J., Parlikad, A. K. and Schooling, J. (2020). From BIM Towards Digital Twin: Strategy and Future Development for Smart Asset Management. In T. Borangiu, E. Oltean, S. Răileanu, F. Anton, S. Anton, & I. Iacob (Eds.), Service Oriented, Holonic and Multi-agent Manufacturing Systems for Industry of the Future (pp. 392–404). https://doi.org/10.1007/978- 3-030-27477-1_30 Malakuti, S., van Schalkwyk, P., Boss, B., Ram Sastry, C., Runkana, V., Lin, S.-W., Rix, S., Green, G., Baechle, K. and Varan Nath, C. (2020). Digital twins for industrial Applications. IIC Journal of Innovation, February, 1–19. https://www.iiconsortium.org/pdf/IIC_Digital_Twins_I ndustrial_Apps_White_Paper_2020-02-18.pdf Mendes, G. and Ferreira, A. (2021). Extending the Multiphysics Modelling of Electric Machines in a Digital Twin Concept. Proceedings of the 11th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2021, 2, 689– 693. https://doi.org/10.1109/IDAACS53288.2021.9660964 Monsone, C. (2021). Holistic Approach to Smart Factory. In E. Mercier-Laurent, M. Ö. Kayalica, & M. L. OWoc (Eds.), Artificial Intelligence and Knowledge Management (pp. 160–176). Springer. https://doi.org/10.1108/978-1-80117-348-320210009 Peng, F. and Zheng, L. (2022). Fuzzy rule-based neural network for high-speed train manufacturing system scheduling problem. Neural Computing and Applications, 9. https://doi.org/10.1007/s00521-022- 07190-9 Qian, W., Guo, Y., Cui, K., Wu, P., Fang, W. and Liu, D. (2021). Multidimensional data modeling and model validation for digital twin workshop. Journal of Computing and Information Science in Engineering, 21(3). https://doi.org/10.1115/1.4049634/1094193 Qian, W., Guo, Y., Zhang, H., Huang, S., Zhang, L., Zhou, H., Fang, W. and Zha, S. (2023). Digital twin driven production progress prediction for discrete manufacturing workshop. Robotics and Computer- Integrated Manufacturing, 80(April 2022), 102456. https://doi.org/10.1016/j.rcim.2022.102456 Regulation (EU) 2016/798. (2016). Official Journal of the European Union. Regulation (EU) 2019/779. (2019). In Official Journal of the European Union. Wei, Y., Lei, Z. and Altaf, S. (2022). An Off-Site Construction Digital Twin Assessment Framework Using Wood Panelized Construction as a Case Study. Buildings, 12(5). https://doi.org/10.3390/buildings12050566 Zhang, L., Guo, Y., Qian, W., Wang, W., Liu, D. and Liu, S. (2022). Modelling and online training method for digital twin workshop. International Journal of Production Research. https://doi.org/10.1080/00207543.2022.2051088