Project Summary
This project aims to build a kernel of a smart rolling stock maintenance solution, using digitalisation, to move away from industrial manual processes. Digitalisation via Digital Twins (DT) is a promising solution and will include three main parts in this project: 1) virtualisation of the physical world (Digital Twin (DT) of the depot and rolling stock), 2) analysis of communication architecture and connections (and their requirements) between the virtual and physical entities, but also between various DTs, in the context of DT for maintenance. In addition, we will review available technologies and their capabilities against the identified requirements, and finally 3) Data storage, management, and analysis to enable mart decision-making.
Project Achievements
The activities undertaken during this project includes improving rolling stock and depot virtualisations. Analysis of the Information Exchange Standard and creation of a rolling stock model using maintenance manuals. Building anomaly detection models trained with simulation data. The following bullets show the main activities performed during the research: • Defining requirements: This includes evaluating each task’s needs and specifying tools, like selecting Simpack and Vampire software for their simulation capabilities. It also involves choosing appropriate methods for defect detection, data collection, and responding based on the data collected. • Designing the DT model and defining the required inputs. • Performing numerical simulations and experimental measurements , using several high-tech sensors and techniques, to create useful data to be tested by the proposed DT model. • Testing the proposed model made any necessary adjustments for optimal performance. • Implementing revisions based on test results to ensure that the DT model correctly validated. • Adapting the scheduling tool for informed decision-making support and preparing a journal paper for submission.
Conclusions
The project successfully demonstrated the transformative potential of digital twins in optimising manual rolling stock maintenance processes. By enabling health monitoring, predictive maintenance and informed decision-making, the solution enhances operational efficiency, reduces downtime, and mitigates unexpected failures. Economically, it lowers maintenance costs and extends asset lifespan. Socially, it enhances the reliability and safety of train operations, leading to greater passenger satisfaction. Environmentally, it supports sustainability through more efficient resource use. The concretisation and deployment of the project extend its impact beyond immediate maintenance improvements, setting a new benchmark for digital innovation in transportation and paving the way for future advancements.
Next Steps
Further tests and validation for the digital twin model: Although we have tested individual use cases, we have not yet implemented a fully integrated digital twin. Conducting small-scale live tests at the IRR facility will allow us to refine the system within a controlled environment before transitioning to broader industrial applications. The findings will be shared through various dissemination channels, including presentations and publications. Applying for further funding to expand the scope of the project: We plan to apply for further funding to expand the scope of the project after submitting another paper detailing the project’s approach and findings. This includes enhancing system capabilities and scaling up its deployment. Engaging with industry stakeholders will be crucial in this phase. We aim to strengthen partnerships with more rail operators and maintenance providers to tailor the system to diverse operational needs. We will also conduct market analysis to explore commercialization opportunities and ensure the system meets industry demands.
