Registered

Intelligence series: How can Digital Twins aid predictive maintenance?

Digital Twins are virtual representations of a physical asset, process or system. Sensor data is fed into a 3D model allowing for the asset to be simulated and monitored remotely. Digital Twins have been closely related to Building Information Modelling (BIM) systems but are slightly different.

Latest update: March 2019

What are Digital Twins?

Digital twins are virtual, real-time representations of physical assets, processes and systems. In comparison to other conventional simulations, which only represent purely virtual scenarios, digital twins are intrinsically linked to a physical asset and aim to represent the asset in real-time. 

Machine learning techniques, statistical and physics-based models, are used to analyse the physical asset's operational data, and operational and maintenance history. The digital twin then acts as a real-time simulation, allowing the asset to be monitored remotely, enabling predictive maintenance to optimise asset performance. 

Recent developments in the technology incorporate cloud technology and augmented/virtual reality to allow users to interact with the digital twin intuitively. 

Once a digital twin is constructed, the twin could be used for the basis of conventional simulations. As digital twins closely resemble their physical counterpart, testing can be accurately simulated at a lower cost than physical testing. This has been described by some as a predictive twin.

What industries use Digital Twins?

In the aviation industry, GE Aviation has created digital twins of all prototype engines involved in the development of the GE9X engine, allowing the designers and engineers to analyse the variation in performance between test cycles, and highlighting the effects of ageing components on engine performance. By using the digital twins from existing prototypes, the specific effects of each design variation can be assessed, therefore allowing the best elements of each design to be used to improve the reliability and durability of the final GE9X engine. 

In the energy industry, digital twin tools have been developed to provide wind farm owners and operators with an insight into turbine conditions and performance. These tools utilise real-time data from the turbines and meteorological measurement equipment to estimate turbine life-span and enhance their asset management capabilities. Predictive analysis of the turbine drivetrain and structural integrity monitoring can be used to adjust wind turbine variables to increase energy production. 

In the heavy industries sector, Aluminium of Greece have worked with GE Power to enhance their aluminium smelting process and increase operational efficiency and productivity through the use of digital twins. The technology has enabled Aluminium of Greece to reduce their energy consumption and use of raw materials, and also improve the overall plant analysis through analytics.

How will it impact the rail industry?

By analysing sensor data, digital twins can model the lifetime performance of various assets such as rolling stock, therefore better predicting when faults and failures could occur. This can reduce maintenance and operational costs by reducing unplanned downtime. Furthermore, applying digital twins can streamline processes such as manufacturing. For example, tracking how rolling stock is designed, configured, built, operated and serviced can identify issues during assembly and result in targeted actions to optimise the amount of material used and improve fuel efficiency. 

GIS can be used to provide information on the location of assets and create a digital twin of the rail network. Creating synergies between the two technologies can highlight infrastructure faults. In addition, GIS can efficiently determine the location of rolling stock faults, therefore allowing faster repair times and minimising network disruption. It will also allow for better coordination of train movements/timetable planning. Digital twins of infrastructure systems can prevent delays and improve maintenance and operations.

What is the current state of R&D?

Many rail and software companies have invested in cloud-based software systems in order to harness big data for predictive maintenance. Chinese rolling stock company CRRC have developed a prognostics and health management system for critical components of high-speed trains. Siemens have developed Railigent - a suite of applications, based on the MindSphere IoT platform. Railigent allows operators to manage rail data and optimise maintenance and operations. These systems lack the full 3D model characteristic of most digital twins but provide the necessary cloud architecture and integration to enable digital twin development. 

For example, the MindSphere platform can be used to gather performance data for Building Information Modelling (BIM). BIM is a process for integrating and managing data on a construction project across the product lifecycle, through CAD representation and standardisation. This allows AEC (Architecture, Engineering and Construction) teams to work in parallel. Crossrail is currently developing a BIM environment. Data integrated with BIM systems could then be used to create digital twins. Siemens Mobility is collaborating with Bentley Systems to mature BIM systems, using the gathered data to create digital twins for design and construction of rail infrastructure projects. Other digital twin platforms for rail projects have been released by companies such as Willow Rail, General Electric and Virtalis

Rete Ferroviaria Italiana is currently mapping parts of the Italian railway network to a 3D digital model. Also, Alstom have created a basic digital twin of the West Coast Main Line network. This allows Alstom to identify maintenance bottlenecks and smooth maintenance peaks, hence improving availability for the customers.

What uncertainties remain?

Creating digital twins can be challenging. For example, each twin must be tailored to each individual asset/system which can be time-consuming to develop. Due to the complexity of modelling the physical asset/system, digital twins require expertise, which can be costly. If the asset is not sufficiently modelled, e.g. missing sensor data, a variation between the behaviour of the physical asset and the digital twin could result in inaccurate assessments. Furthermore, cyber-attacks could become a growing threat if digital twins on the cloud are vulnerable to hackers.

What should the rail industry do?

Infrastructure managers could invest in IoT platforms and BIM systems to begin building a platform for data integration. Stakeholders could also ensure legacy and new data systems are interoperable with these platforms. This would lay down the digital framework required to begin building digital twins. Stakeholders could engage with groups such as BIM4Rail to facilitate this. The National Infrastructure Commission has set out a roadmap for a national digital twin, a combined digital model of the entirety of the UK infrastructure. The rail industry could collaborate with other transport and infrastructure organisations to ensure interoperability between systems. Stakeholders could also engage with research institutes with expertise within artificial intelligence, to build complex digital twin models of rolling stock.

Log in or register to keep reading
Register for free individual access
  • Unlock research, articles and more
  • Get updates on RSSB’s activities

 

 

 

 

Need some help?
To talk to us about accessing RSSB content or corporate membership:
X
Cookies help us improve your website experience.
By using our website, you agree to our use of cookies.
Confirm