Schedule

Welcome to Session 1 & Housekeeping

Talk 1 – Virtual Homologation of European Rolling Stock

The talk provides the digital taxonomy of the key rolling stock types and covers the key engineering principles for virtual homologation. There is a concentration on the aspects of the rolling stock that make it sustainable in terms of passenger transportation, these aspects are primarily safety and crashworthiness. Virtual homologations examples are provided that contribute to make the rolling stock and hence rail system more sustainable. It shows how passengers can be confident that they are travelling in safe rolling stock and the vehicle consist, manufacturing technique or materials used should not compromise the crashworthiness aspects and this is demonstrated virtually.

Professor John Roberts (Kasetsart University)

Talk 2 - Digital Twin-Driven Methodology For Predicting Wheel Degradation and Optimising Wheel Maintenance

Wheels are key components of high-speed trains. Their degradation has a direct impact on the safety, reliability, and quality of operations. High-speed train wheels require frequent inspection, reprofiling, and renewal, resulting in huge maintenance costs. Advances in wheel-rail dynamics theory, measurement technology, and data science make it possible to develop a digital twin for the degradation of wheels over their life cycles. In this presentation, we will introduce how wheel degradation can be modelled and predicted, taking into account wheel wear, reprofiling, and defects. We will then present how wheel reprofiling strategies can be optimised with the objective of minimising life cycle costs. Furthermore, we will show the application of the proposed methodology in China, which not only enhances the safety and reliability of train operations but also yields significant economic benefits.

Professor Weihua Zhang (Southwest Jiaotong University)

and Dr Yuanchen Zheng

Talk 3 – Integrating Crowdsourcing and Digital Twins for Service Innovation Platform Design

In the context of Industry 4.0 technological development and digital transformation in industry and our society, this talk will first discuss the trends and challenges of shifting from product design to product-Service ecosystem (PSS) Design along the product lifecycle under the manufacturing servitisation. Secondly, it will discuss and demonstrate how digital twins and crowdsourcing technologies are integrated for smart product service ecosystem design.

Professor Sheng-feng Qin (Northumbria University)

Talk 4 - Creating a Digital Twin For Rail Infrastructure

This talk will introduce the concepts behind a digital twin and how digital assets management, reality capture and BIM contribute to digital twins of railway systems. The use of GeoAI for creation and utilisation of a digital twin will be presented followed by a discussion of real-world projects and state of the art approaches and technologies. Finally, we will look at how digital asset management can benefit from the digital twin.

Dr Uwe Jasnoch

Talk 5 - Prometheus – An Autonomous Reconfigurable Drone for Building Digital Twins of Subterranean Mines

There are a significant number of subterranean mines below critical rail infrastructure which need to be constantly monitored to ensure the safety of the rail network. Traditionally this is done through in-person inspections or laser scans from surface boreholes. To increase safety and mapping coverage, the Prometheus reconfigurable drone was developed which could be deployed through a 150 mm borehole and autonomously explore a mine network to develop an environmental digital twin. The drone was deployed in several mines in the UK in 2021 demonstrating the core capabilities.

Dr Simon Watson (The University of Manchester)

Talk 6 - Digital twin-based Discrete Manufacturing System Solutions

This talk has three parts. Part 1 will introduce the Professor Ding’s research team and industrial software development company. Part 2 will first introduce what is Discrete Manufacturing System (DMS) and its challenges, then presents concepts, technological layers and utilities of digital twins, and discusses digital twin-based DMS solution frameworks including digital models, digital shadows and digital twins and their current state-of-the-art research. Part 3 will discuss key techniques for creating digital models implementing digital shadow and realizing digital twins in applications, and finally demonstrate several case studies.

Professor Guofu Ding (Southwest Jiaotong University)

Panel Discussion

Welcome to Session 2 & Housekeeping

Talk 7 - Smart Energy Networks: Living Labs, Digital Twin and AI Applications.

The talk will introduce one of Europe’s largest smart energy network demonstrators on our campus, and discuss our research on digital twins for smart energy systems and AI applications.

Professor Zhong Fan (Keele University)

Talk 8 - Digital Twin for Product Design

With the advent of new generation information technologies in industry and product design, the big data-driven product design era has arrived. However, the big data-driven product design mainly places emphasis on the analysis of physical data rather than the virtual models, in other words, the convergence between product physical and virtual space is usually absent. Digital twin, a new emerging and fast growing technology which connects the physical and virtual world, has attracted much attention worldwide recently. This talk presents a new method for product design based on the digital twin approach. The development of product design is briefly introduced first. The framework of digital twin-driven product design (DTPD) is then proposed and analysed. A case is presented to illustrate the application of the proposed DTPD method.

Professor Ang Liu (University of New South Wales)

Talk 9 - A New Differentiable Architecture Search Method for Optimizing Convolutional Neural Networks in the Digital Twin of Intelligent Robotic Grasping

Convolutional neural networks (CNNs) have been widely used for object recognition and grasping posture planning in intelligent robotic grasping (IRG). Compared with the traditional usage of CNNs in image recognition, IRGs require high recognition accuracy and computational efficiency. However, the existing methodologies for CNN architecture design often rely on human experience and numerous trial-and-error attempts, which make it a very challenging task to obtain an optimal CNN for IRGs. To tackle this challenge, this talk presents a new differentiable architecture search (DARTS) method considering the floating-point operations (FLOPs) of CNNs, named the DARTS-F method for constructing a digital twin (DT) of IRG, which converts the discrete CNN architecture search to a gradient-based continuous optimization problem and considers both the prediction accuracy and the computational cost of the CNN during the optimization. To efficiently identify the optimal neural network, this talk further presents a bilevel optimization method, which first trains the neural network weights in the inner level and then optimizes the neural network architecture by fine-tuning the operational variables in the outer level. In addition, the proposed DT of IRG considers the physics of realistic robotic grasping in the DT’s virtual space, which could not only improve the IRG accuracy but also avoid the expensive training time. In the experiments, the proposed DARTS-F method could generate an optimized CNN with higher prediction accuracy and lower FLOPs than those obtained by the original DARTS method. The DT framework improves the accuracy of real robotic grasping from 61% to 71%.

Professor Weifei Hu (Zhejiang University)

Talk 10 - Key Technologies of Shape-Performance Integrated Digital Twin for Major Equipment

Accurate prediction and reliable analysis of morphology and performance of major equipment is one of the key technologies to realize its intelligence and independent innovation. As a link connecting the physical world and the digital world, digital twin can realize true mirror of the whole life for material design of physical entity, structure design, manufacturing, and operation and maintenance management in digital space. Facing the geometric morphology and mechanical performance of major equipment, and analyzing the difficulties in establishing its digital twin, a solution of “computation -measurement combination” based on measurement data and mechanism model is proposed. Considering the timeliness and accuracy requirements of the twin model, a shape-performance integration digital twin (SPI-DT) framework for major equipment is built. Additionally, six specific problems faced by building the digital twin of major equipment are discussed in detail, including “unrealizable calculation”, “inaccurate calculation”, “delayed calculation”, “unmeasurable data”, “incomplete measurement” and “inaccurate measurement”, and the relevant solutions and key technologies are given. The feasibility and validity of the proposed framework and key technologies are described by combining typical case, which provides a theoretical and methodical reference for the further application of digital twin in major equipment. Finally, the future development trend and further challenges of digital twin of major equipment are discussed.

Professor Xueguan Song (Dalian University of Technology)

Keynote - Spatial Digital Twins for Smart Cities

Contemporary decision-making for Smart Cities and sustainable urban modelling is highly dependent on geospatial references. In the last years, many spatial Digital Twins have been developed to analyse, simulate and forecast different urban phenomena for Smart Cities. All of them have used a range of 2D or 3D spatial data obtained from different institutions, agencies and companies. Fusion of information has become a necessity and a critical aspect of creating spatial Digital Twins. Through the years many topics related to integration of geospatial data have been discussed. Research on Spatial data infrastructures was the first attempt to harmonise GIS data. More recently the term Digital Engineering came to reflect the effort to establish standards in construction (BIM) industry. The spatially enabled Digital Twins for Smart Cities brought yet another dimension, i.e. the fusion of geospatial data and real-time sensor streams with the intention to model the complex interaction of environmental and physical conditions. For example, indoor air quality recordings, building occupancy numbers, outside temperature, time of day and 3D built structure, allows for adaption of designs, regulations, and policies towards improving health and human well-being. While it is acknowledged that every domain has its own specificities, the need to fuse, re-use and share geospatial data is becoming increasingly important for spatial Digital Twins. Critical to realise is that real-world consist of unique artefacts, but they are perceived and modelled from different perspectives, which leads to diversity representations and data sets. Therefore, agreements must be found at different levels (semantic, geometric, topological), which corresponds to the textual description provided to the real-world object (e.g. name, metadata, etc.) and would allow for seamless fusion of data. This talk will discuss the specific of spatial Digital Twins and its importance for Smart Cities, illustrated with examples from a couple of projects. The challenges in geospatial data fusion will be briefly addressed as well as envisaged directions for research.

Professor Sisi Zlatanova, (University of New South Wales)

Panel Discussion