R4F-Reports of Area 1: Simulation Platform for Railway Systems

Below you will find a selection of the scientific reports produced for the Comet project Rail4Future - Resilient Digital Railway Systems - from Area 1 for download.

The Rail4Future Virtual Validation and Simulation Platform will enable for the first time: A fully virtual assessment of the behavior of the railway infrastructure system by deploying large­scale simulations based on hybrid-models enhanced with Al capabilities, real-time data input from the railway operations and visual-analytics methods.

D1.1.1 Report of requirements

This report presents the requirements of the R4F platform, based on which we designed a conceptual model-based digital twin for the holistic large-scale railway infrastructure. R4F Platform, denoted as the Rail for Future Platform, is proposed to combine different elements of the holistic railway infrastructure system in a layered model, in which complex interrelations can be broken down into smaller and simpler clusters.

D1.1.1 One Pager - Report of requirements.pdf

D1.1.1 Report of requirements (M12).pdf

D1.1.2 Blueprint architecture with simulation workflow (M14)

In this report, a conceptual model-based digital twin platform that aims to integrate measurement data, numerical models, and machine learning models from different railway subsystems into a holistic large-scale railway infrastructure platform is proposed for the first time. The R4F Platform may ensure that reliable and valuable data can continuously flow throughout the whole life span of the holistic system and its subsystems, which may help build a fully connected and digitized railway infrastructure system. The R4F Platform will be exploited for problem diagnosis, predictive maintenance, condition monitoring, and some other cases based on user demand.

D1.1.2 OnePager - Blueprint architecture with simulation workflow.pdf

D1.1.2 Report - Blueprint architecture with simulation workflow (M14).pdf

D1.1.3 Multi Domain Model Description

In this deliverable, a multi domain model is shown and described, which connects all simulations (multiphysics, machine learning) with its relevant data and the required semantic to interconnect all characteristics of the model. As the name suggests, it includes more than one domain, which are used to integrate and interoperate different assets, belonging to different subsystems such as track, vehicle, turnout, bridge, with each other in a DT platform called Rail4Future.

D1.1.3 OnePager - Multi Domain Model Description.pdf

D1.1.3 Report - Multi Domain Model Description.pdf

D1.1.4 Interface and Adapter Specifications

In this deliverable, interfaces and adapters are shown and described, which are implemented into the multi domain model, mentioned in the previous Deliverable 1.1.3 (M18), to improve and optimize the model and data integration process in the Rail4Future (R4F) platform for the digitalization of the holistic large-scale railway infrastructure system.

D1.1.4 OnePager - Interface and Adapter Specifications.pdf

D1.1.4 Report - Interface and Adapter Specifications.pdf

D1.1.5 Implemented and Tested Sub Model Environment

This deliverable provides comprehensive insights about the use cases that are considered and which of them were implemented, tested, validated and finally released for the R4F project by using the adapters and interfaces mentioned in Deliverable D1.1.4. Besides, the use cases, which were already implemented, are briefly described in this report. The description of the use case implementation methodology can be found in Deliverable D1.3.5.

D1.1.5 OnePager - Implemented and Tested Sub Model Environment.pdf

D1.1.5 Report - Implemented and Tested Sub Model Environment.pdf

D1.1.6 Simulation Deployment Results

In this deliverable, we present and analyze simulation results derived from use case deployments within the R4F Platform. These results consist of validation curves (FMI-based vs. raw model simulation), CSV- and JSON-datasets, which are valuable for visualization, analysis and exchange among all stakeholders. This result acquisition has an important part in quality assurance of the use case implementation in the platform (see Deliverable D1.1.5 & D1.3.5). This process enables us to confirm the proper functionality of the integration, delivery, and deployment of assets associated with various railway use cases.

D1.1.6 OnePager - Simulation Deployment Results.pdf

D1.1.6 Report - Simulation Deployment Results.pdf

D1.2.1 Report of Requirement Analysis

This document details the requirements of the visual analysis framework for the Rail4Future platform. The visual analysis framework should be designed in a way that facilitates adding new use cases and assets. A flexible user interface in the shape of an interactive dashboard should be provided. Point cloud visualisation, maps, charts and simulation planning should be available and linked where the data allows it. Simulation results should be visualised in a spatiotemporal context.

D1.2.1 OnePager - Visualization Requirements.pdf

D1.2.1 Report of Requirement Analysis.pdf

D1.2.2 Specification of Visual Analysis Framework

This document describes the system architecture of the visualization layer of the Rail4Future platform.  The visualization layer consists of modules dedicated to the different visualisation modalities. The dashboard module combines the modules for point cloud visualisation, maps, charts and simulations. It also includes a workspace module for asset management.

D1.2.2 OnePager - Specification of Visual Analysis Framwork,pdf

D1.2.2 Report - Specification of Visual Analysis Framwork.pdf

D1.2.3 User Manual Of Visual Analysis Framwork

This deliverable contains the User Manual of the visualization prototype for the Rail4Future project. It describes the functionality of the prototype divided into a user manual for standard users and one for expert users who might want to change the data or more advanced settings. The user manual describes the generally applicable part of the visualization prototype, and then each implemented use case.

D1.2.3 OnePager - User Manual Of Visual Analysis Framworkpdf

D1.2.3 User Manual Of Visual Analysis Framwork.pdf (German)

D1.2.4 Evaluation Report of Visual Analysis Concept (M36)

Evaluation in visualization research is critical as it ensures that the developed prototypes effectively communicate the intended data insights and meet the needs of the target users. Evaluation helps in assessing the prototype’s usability, accuracy, and efficiency in supporting users in gaining insights into the visualized data. We performed a qualitative user study to gain in-depth understanding of the participant’s experiences, perceptions, and behaviors.

D1.2.4 OnePager - Evaluation Report of Visual Analysis Concept (M36).pdf

D1.2.4 Evaluation Report of Visual Analysis Concept (M36).pdf

D1.3.1 Simulation Reliability Assessment Indices

Simulations can only be helpful if appropriate processes and metrics are available to assess their quality and determine their significance. If possible, these evaluations should be automated in order to seamlessly integrate the evaluation into processes so as not to create additional hurdles for developers, but still create value and support the evaluation of the available simulation results and data. These criteria are therefore also necessary to enable virtual certification.

D1.3.1 OnePager - Simulation Reliability Assessment Indices.pdf

D1.3.1 Report - Simulation Reliability Assessment Indices.pdf

D1.3.2 Generic Model definition for multi-level optimization (M14)

This report presents a generic model definition for multi-level optimization in railway systems, which encompasses different model types and their corresponding use cases relevant to railways. The methodology for developing and implementing this model is outlined, alongside a modular and scalable architecture - Rail4Future (R4F) platform that integrates various components and subsystems of the railway system, such as the track, trains, and other systems. Mathematical, MBD, and ML models are developed to represent the railway system at different levels of abstraction and granularity.

D1.3.2 OnePager - Generic Model definition for multi-level optimization (M14).pdf

D1.3.2 Report - Generic Model definition for multi-level optimization (M14).pdf

D1.3.3 Report - Enabling large-scale simulations (M18)

The holistic railway infrastructure digital twin (DT) platform is sophisticated and consists of a series of submodels (e.g., turnouts, tracks, vehicles, etc.) that are built through various methodologies and software. However, integrating these submodels i nto the DT platform is tremendously challenging due to considerable computational complexity, software and interface restrictions. In this report , we present a machine learning (ML) based surrogate modelling methodology for the submodel integration in the holistic railway infrastructure DT platform to enable the large scale simulations and illustrate the methodology through a case study.

D1.3.3 OnePager - Enabling large-scale simulations (M18).pdf

D1.3.3 Report - Enabling large-scale simulations (M18).pdf

D1.3.4 Report - Procedures eliminating complexity (M24)

In order to build a large-scale computation platform which enables simulation as a service, performance is a key feature. In this project, mathematical functions, methods combining AI and analytical methods required for fast computing, but credible large-scale simulations are being researched and developed to make the R4F Platform viable for usage. Since this topic consists of different layers in the system, we gather the different approaches from the different layers of simulation deployment and execution in this report.

D1.3.4 OnePager - Procedures eliminating complexity (M24).pdf

D1.3.4 Report - Procedures eliminating complexity (M24).pdf

D1.3.5 Report – Implementation Strategies

In this deliverable, we show and describe strategies we followed to implement different railway models and data into the R4F Platform. During this implementation, we mention different open-source or commercial software tools, packages, libraries, file formats and interface standards, which we propose to use to standardize the simulation units of these models, and then run their simulation in the platform (see Deliverable D1.3.4).

D1.3.5 OnePager – Implementation Strategies.pdf

D1.3.5 Report – Implementation Strategies.pdf

D1.3.6 Report – Automated Model Integration

In this deliverable, we show and describe how we designed and developed the methodology to automate and manage the entire asset integration and processing, based on different interface standards (e.g., FMI, SysML), by using different key technologies such as graph DB, CI/CD pipeline, artifactory repository management, VCS, CAT and CLI. This is enormously important to gain insights about automation techniques helping to reduce time effort and resource consumption.

D1.3.6 OnePager – Automated Model Integration.pdf

D1.3.6 Report – Automated Model Integration.pdf