Twins on Track: Maximizing the Potential of Digital Rail Asset Management

ANA DE LUNA: Hello, everyone. Thank you so much for joining today's presentation. We will be discussing the topic of Twins on Track. How can we maximize the potential of digital rail for asset management? My name is Ana De Luna, and I am a BIM consultant for Latin America for Deutsche Bahn Engineering & Consulting. I am an architect with a master's in project management and a master in BIM, and here with me today is Julio Palma, who I will let him introduce himself.

JULIO PALMA: Thank you, Ana. My name is Julio Palma. I'm a civil engineer. I'm part of the technical sales team in Latin America in Autodesk. I'm a sustainability leader. Thank you for having us today.

This is going to be our agenda today. First of all, we're going to discuss about digital twins in rail system and how Deutsche Bahn deals with this infrastructure. We're going to take a deep dive into the Deutsche Bahn use of digital twins in their projects. Later, we're going to discuss about how asset management is a key factor in digital twins for Deutsche Bahn projects, and later, we're going to have a brief case study regarding a project in Latin America about digital twins.

First of all, we're going to start with a brief about challenges in rail infrastructure around the world. First, we have centennial infrastructure that is aging. This infrastructure is in need to re-enter operation in the coming years. There are a lot of Brownfield projects where existing infrastructure needs to be updated to meet new operational requirements. In most of the cases, there is no record of primary information that is needed to take this into an operation project, and there is a big uncertainty regarding this infrastructure that is already in place.

And later, there is also a fast-- these are fast projects, fast-track projects that are needed to be developed in short time in design and to be placed into operation in the short term. Beyond that, there are new business models and market needs.

There is a big need for risk management, and that is represented in BPP's projects where the different stakeholders need to perform value engineering processes in order to optimize the CapEx of the project, and there is a big interest in sustainability KPIs where both the operational and embodied carbon is optimized in order to comply with regulatory alignments. And finally, there is a big focus on efficient operation that is actually one of our big messages here in this class today.

Let's discuss a little bit about a specific case. This is Colombia's case. In Colombia, for example, there is a "30 years behind schedule" infrastructure. There is a lot of work in progress of getting over that behind of schedule, and there are big projects in order to update that transportation infrastructure.

There are investments. There are growing investments in more that $24 billion in different strategies around transportation. This is regarding rail renovation, roads, and port concessions. And there is a national BIM standard going live in 2026. Ana?

ANA DE LUNA: Yes, thank you. So let's start with the topic of digital twins in railway system and connecting with the challenges that Julio just said to us. But I want to start this presentation by this analogy. So imagine walking into a kitchen where the pantry is a mess, items are scattered, labels are unclear, and you spend too much time searching for basic ingredients. So this organization makes meal prep frustrating and can be time consuming as well.

So now, in contrast, there's this pantry where everything is needed, everything is arranged, clearly labeled, and accessible, so preparing meals can become more efficient and enjoyable. So how would you describe the difference between these two pantries? Do you think we need to go from the left pantry to the right pantry? And how do we go from left to right?

Why did I bring this analogy into the topic of asset management? Now let's apply this scenario into asset management. Just like the messy pantry, inefficient asset management systems force employees to spend valuable hours gathering data instead of doing their job. So according to McKinsey reports, employees spend an average of 9.3 hours per week searching for information, which is equivalent to losing an entire worker in inefficiency. This data is supported by IDC, showing that knowledge workers spend roughly 2.5 hours per day, which is around 30% of their workday, searching for information.

The lack of organization and the access of critical asset information leads to inefficiency, leads to wasted time, and without a clear structure, managing assets can be chaotic and can be unpredictable. But just as the organized pantry, it can simplify the meal prep, having in a structure up-to-date data allows teams to work efficiently, to anticipate maintenance needs, and to allocate resources effectively.

So how the digital twin comes into the topic? Sorry. So digital twins, like well-ordered pantry, provide a clear and intuitive way to manage assets, empowering teams to make faster and more informed decisions. But what is a digital twin? This is a virtual representation of physical objects system or processes that use real-time data.

It is not only using our industry, the AEC industry, but you see on the right that it's also used in agriculture, in healthcare, in aerospace, so different industries are working with digital twins. They are valuing the real-time data and simulations to optimize operations and reduce costs. You see on the left that it is a growing market. It is valued for about $13 billion, and it's projected to go up to $259 billion by 2032. These are rapid-growing market because they see this, they recognize this power of data.

In the context of real asset management, digital twins are revolutionary in the way the entire rail networks are managed. With these data-rich models, operators can achieve and can improve the asset life cycle management. They do predictive maintenance, and they have more efficient operations.

But what is an asset in railway? In the world of railway, an asset refers to any physical or digital resource that contributes to the operation of the overall network, such as the rolling stock, which are the locomotives, the passenger cars, the freight cars. We also have the linear assets, which are the tracks, the signals, the bridges, the tunnels, the power.

We also have facilities or the buildings, the stations, the depots, other buildings that are for repairing. We have a support fleet, which are the cars, the trucks, any specialized equipment for repairs. And finally, we also have IT assets, the data centers, the network, the hardware, everything that is connecting.

And you can see in the next picture how these different assets are connected to have an entire network. Every single one of these assets is performing in a certain way and is giving signals of their state. And the true power comes from integrating different assets from different data from different sources, and it's providing an entire view or a holistic view on the overall system.

This data, where this data is coming from, it can be classified into three different categories. We have primary data, which comes from rolling stock data or infrastructure data. We also have secondary data, which comes from auxiliary sources, such as the asset data, BIM models, geographic data, operational data, maintenance data. And finally, we have external factors that contribute to the overall operations, such as the weather data.

Everything is gathered, and the holistic way of seeing the railway system enables to be more cost-efficient, to improve the processes, and to boost availability and reliability of the entire system. By evaluating not just the individual components but understanding the overall and comprehending the overall system can increase how the assets are performing between each other.

So now let's deep-dive into Deutsche Bahn use of digital rail. To give you an overview, Deutsche Bahn group, it is the largest rail network in Europe. We are the owners of 33,000 kilometers in Germany. Part of these assets are also the stations, like I said.

There's 5,700 stations, 25,000 bridges, 750 tunnels over this in Germany. And we are transporting passengers and goods. And you see the numbers there. It can give you an idea of the challenges and also the opportunities that we have as the owners of these kilometers, this railway network.

For this, Deutsche Bahn has a strategy, which it calls Strong Rail, and it's built on the idea that the railway system is strongest when it is digital. So the digital transformation is essential to address the current and the future challenges to ensure operations that are more efficient, reliable, and customer-focused. This Digital Network is key to optimize operational processes.

With this strategy and in order to overcome the challenges and the new demands, we have three objectives-- having more capacity in the overall system, have a higher quality in our operations and being more stable and reliable service, and finally, have modern working environments, which gives more efficient results.

There are different focus and projects to achieve these objectives, such as the ones that you see on the screen, and the one that is more related with the topic today is about advancing digital infrastructure. But as you can see, they're interconnected to the overall results of digital transformation. To put these technologies into practice, there are different pilot development or infrastructure projects all over Germany, and today, we will show you later a project outside of Germany, in Latin America.

This way, a digital twin is created by an overall model of the infrastructure. From the design and construction phase, we create a virtual image that is then enriched with information obtained, for example, by sensors, by drones, or even the person, which is provided in real time via IoT platform during the operation and maintenance phase. But it is exactly to have digitalized the entire rail system.

And while BIM provides the static data, digital twins are using these sensors in order for obtaining the asset. So how is it related? Imagine having a bustling rail station. Trains are arriving, departing. Passengers are rushing to their destinations. And behind the scenes, there's rail operators that are grappling with aging infrastructure, increasing demands, and need to be very efficient with their service.

So for the railway industry, to connect it with the analogy of the pantry, any disorganization comes with an enormous cost. The maintenance of railway infrastructure costs to the sector approximately 40,000 euros per kilometer of track with maintenance activities representing 38% of total operation cost. These numbers underline the urgent need to be more optimized and streamlined in our system.

This is where digital twins come into play. They allow us to organize the pantry of the railway assets and ensuring that every piece of the data is accessible, is clearly labeled, is optimized for use, and, at the end of the day, is reducing time spent, searching information so they drive productivity and improve the planning of maintenance to overall reduce the cost across the system.

So how does the rail asset maintenance work? There are different types of maintenance, and it goes from reactive maintenance, which is when it happens something urgent that it needs urgent intervention when a system is failing, so we maintain after a failure, of course. Then we have preventive maintenance, which is we maintain regularly to prevent failures.

We have condition-based maintenance, so we maintain or repair assets when condition degrades. Or we have predictive maintenance. It's scheduled and prioritized maintenance based on the state of our assets. So the idea and the goal that you see there-- the goal is to go into predictive maintenance, and it's just not about implementation. It's a chance to improve maintenance, and this is enabled by data. We need to collect data, understand our data, determine what is the condition, [INAUDIBLE] statistics, and then do predictions.

Now, going from reactive to predictive, this graph that you're seeing is illustrating the exponential cost of reactive and/or predictive maintenance. The navy line shows that planning and investing in setting up digital twins with the sensors and combine data science with analytical techniques for this predictive maintenance will, in the end, in the long term, cause savings.

The upfront investment in digital twins will also result in increased reliability as for the user and for the owner, and you can see it around three years implementing the results of this. But it is a long-term-- it is a long-term base of understanding how the digital twins and moving from reactive to predictive benefits in the long run.

So from the BIM perspective, how can we jump in on the digital asset management initiative? We know that BIM provides aesthetic data, but in order to be digital twins, we need to connect and to integrate real-time data with asset management systems. So this integration is crucial for effective maintenance and operations, which allows real-time data analysis and decision-making. So to get BIM into the asset management, there's three things, the strategy, the digital twins, and the training.

According to McKinsey, asset managers are investing more in technology but not necessarily with an optimal strategy, so we need to start with the strategy in mind, one that is aligned with the overall asset management strategy, with data analysis, and with sustainability. So we need to design data-driven strategies for rail systems that require stakeholder collaboration.

So to break down this digital strategy and develop a BIM strategy for digital asset management, we need to or we are proposing having these four pillars, people, policies, processes, and technology, and acting in three different phases from development to implementation to acceleration. I will go deeper into each one of those.

So for example, for people, we have an initial training sessions to make everyone familiarize-- every employee to familiarize what this new digital asset management is about, the technologies and methodologies, and these can also be restructuring in roles and responsibility that aligns to these new processes and technology. So there's an initial training and an organizational change.

In the meantime, in the implementation phase, there's a specialized training, so any special skills that need to be handled, and finally, in the acceleration phase, we have training expansion. So how can we ensure that we have all the necessary competencies to operate? And finally, the adaptation of organizational change-- so there's a continuous monitoring and supporting the organizational change of having the digital in asset management.

Then we have the policies, and for the policies we need first to create or review internal policies. And we need to ensure how can we are supporting the digital asset management, so any internal aspect. But we also need to connect with the external regulations, so any applicable laws and regulations that we need to comply with legal, with operational risk and also with the standards, such as ISO 19650 for BIM or the ISO 55001 for asset management.

Also important is to define the classification system to ensure that we are talking about the same data. In the mid-term, we also do any policy adjustments based on the feedback, and at the end of the day, we do an ongoing and regular monitoring and reviewing of the policies and adapt based on needs and based on technological advances.

Then we have the processes, and first, we need to understand in depth the current processes and identify opportunities of improvement, so having the needs, the challenges faced. Then we suggest to start with a pilot, so execute pilot processes to test the processes in a controlled environment to later on do a large scale. In this pilot, do evaluation. Do adjustment. Do an analysis of the results. Do any process optimization, refinement of these processes.

And finally, you go into the improve processes by implementing any automation, continuous monitoring, and any process documentation that needs to be formalized, it's needed to have it documented in order to understand how is it the pilot phase into a more scalable phase.

And finally, but not least, the technology-- so first, we need to evaluate and do the technology selection, what tools, what suppliers, and the best solutions for this digitalization of asset data. Then we move into the deployment of these case studies for the BIM uses and do a technology monitoring, so continuous evaluation about how these technologies are being implemented. And then we move into a large-scale deployment and technological adjustments based on the review and the continuous improvement of the rest of the processes from the people, the policies, and the processes.

From all these strategies, once we have these strategies, we'll then need to start the creation of the digital twin, and for that we need to collect, model, connect, and manage data. In the collection, we need to collect the necessary information based on a strategy, and we have two aspects of collection, so the geometry and information.

From the geometry, it depends on the case of each project. For example, if it's a greenfield project, you can have an as-built model that is coming from the design of the construction. If you have a brownfield that doesn't have a model, you can do scannings and then do a scan to BIM, which allows you to have accurate and rapid evaluation. It eliminates the need to rely on outdated information.

Then you have the-- collect the necessary information based on the strategy, so any metrics, any KPIs, and other company data. And then you have external documentaries, like technical sheets, warranties, manuals, anything that it can be contained in the model as form of information. Then we start with the BIM modeling of the digital twin from the development of the BIM execution plan specific for the use of operation and maintenance phase for asset management.

We understand the model segregation strategy, the model authorizing tool. What is the common data environment that is going to be used, the naming convention, the level of detail, the level of information? Everything like this is defined in the strategy and collected in the BIM execution plan.

And then you start developing the BIM model your facilities. Fostering these processes, you can import the point cloud into Revit, for example, and improve accuracy. You can modify the level of detail depending on the need, and you can use VR/AR applications to fully understand your facilities.

Then you will go into the connection, the digital twin. This follows a connection strategy, the tool selection of the different data. So you have maintenance data, such as warranties, suppliers, tasks. You have operations data, IoT, asset data, classification, manufacture quantities, everything connected.

And at the end, you do the management. So the goal with proper data management is to make information usable and transferable, according to project requirements. So identify what information responds to operation and maintenance needs, such as what information I need in order to make an informed decision, what information I need in order to carry out key operation maintenance activities, or what information is necessary to answer critical or urgent questions in order to-- when an event or a disaster happen or for safety.

So based on this information, responding this kind of questions, and by analyzing data of the sensors and historical data for maintenance records, the idea is to prevent any likelihood of something happening and suggesting any scheduling for maintenance solutions.

Of course, there are different challenges that they're facing with this methodology. Those include, for example, data standardization, data management, data security, as well as the need to update all IT infrastructure, the challenges of connectivity, privacy, security, the lack of standardized modeling approach, and the significant challenge of the market include the cost of deployment, the demand of power and storage.

So implementing digital solutions can be costly, but they require significant investment on the sensors, on the software, of the infrastructure, all the data quality control, the security solutions. But furthermore, maintaining the digital twin infrastructure requires an investment to operations. But the digitalization strategy is not necessarily a walk in the park, but it is a necessary park we need to walk in. So even if we start with small steps, we are going into the right direction.

And those first steps-- I want to talk about this case study in LATAM. This is a pilot project that we did with first metro system in Colombia that has 25 years of operation. And as we said or as I said before, we started with the objectives and the strategy. So we had an initial knowledge transfer and training, so we were all on the same page on the topic.

Then we started evaluating what was the best pilot in terms of what input information do I have access to. Do we need a survey? Do we have operation and maintenance information, asset information? How are we going to process it with BIM? What level of detail do we need? What level of information is necessary? How can we answer the questions that I was talking about previously?

And what is the BIM strategy overall in the system? Is there a BIM strategy for operation and maintenance? And all these questions at the end of the day for the main idea is to select a pilot for operation and maintenance. Based on this, a substation of the metro system was selected, and we also needed to understand the current management process.

You see here the different steps from the system alert to alert communication to do an on-site inspection, and you see the verification, the maintenance execution, the order settlement in SAP. But all of these can have a delay in detecting communication issues, so rely on manual communication often leads to delays and miscommunication. And this has slowed down the response to problems.

And just like I was talking before, like a cluttered pantry with [INAUDIBLE] and outdated items, making this meal prep harder than in the real time, having lack of information causes inefficiency. It can also be time-consuming to do these inspections and manual maintenance orders, and it can have a delay on getting the materials and finishing the work.

So by understanding the current situation, we did a-- once the substation was selected, we started collecting information, and since we decided to do the scan and pictures to understand the area, we created a point cloud using ReCap, and we started requesting information, gathering documentation involving the different stakeholders that had the different information about the same asset.

Once information was collected, we move into creating based on the strategy for operation maintenance, going from scan to BIM, and modeling the different disciplines, architecture with the shell, MEP with the lighting, fire protection, fire alarm, electric equipment, et cetera, and not only the geometry but also the information in the form of attributes or parameters. Then the BIM model was then integrated in Autodesk Tandem and started connecting with the operation maintenance platforms, such as SAP in this case, and having in mind this future connection with IoT.

And here, you see how this overall in Tandem looks like. We have the digital twin. We have the properties, the different elements, the different parameters of the asset. And the advantage of having an inventory available to review, this inventory can also be filtered according to what you are looking for. And by having a dashboard, you see the data in different ways, and it can help you make decision process even more, depending on what the situation is about.

So by having a digital twin for management, you can have a faster detection and communication with the digital twin that is real-time data. It's automatically captured and shared, eliminating the need to do manual communication, so the teams are instantly alerted or depending on what is the strategy for alerts.

You can also streamline inspections. So the digital twin, by providing the real time, you can do remote visibility of how is the asset conditions, reduce on-site inspection, and integrate with other platforms, like I said, SAP, to automatically generate or be more efficient in how you generate the maintenance order, which simplifies the process.

And you have a better material tracking. Unfortunately, the execution right now is you don't do it visually, but you do the maintenance execution, and you start tracking, offering up-to-date inventory and maintenance needs, depending the material availability. You understand more your inventory, and you are, at the end of the day, reducing delays, always having organized pantry where everything is in place, easy to find. So the work gets done faster without unnecessary downtime.

And at the end of the day, as I've talked during this whole presentation about predictive maintenance, you can start gathering the historical data on how your assets are performing and then start predicting what is the likelihood of something happening and planning and scheduling your maintenance needs according to that.

So to solve this problem of inefficiency and wasted time, the concept of digital twins offer us this powerful solution. Just like organizing this pantry ensures that everything is in the right place for maximum efficiency, safety, and easy to access, these digital twins provide this structure for railway operations.

So they empower us to predict needs, to prevent issues before they arise and ensure that all the parts of the railway network work together seamlessly by reducing waste, minimizing downtime, and improving the overall performance. So the digital twins enables us to manage assets smarter and more effectively, creating a smoother and more reliable operation. So we'll hand out to Julio.

JULIO PALMA: Thank you, Ana. OK. Thinking about digital twins brings us to the concept of maturity levels, levels based on how digital twins supports business value and empowers digital transformation. Verdantix proposed a maturity level for digital twins comprised of five levels a descriptive twin, which provides a foundation of normalized data for facility assets, spaces, and systems. Leveraging as-built design and construction data creates a digital replica of the facility in a normalized representation created for operations.

Informative twins augment the descriptive twin with operational and sensor data, both normalizing this data and delivering real-time and historical insights. Predictive twins introduce analytics to provide early fault detection and predictive insights for optimizing building operations. Comprehensive twins add simulation to perform what-if scenarios, how much the upgrade of the system affect building performance, or how our reconfiguration of the space would affect occupation and utilization.

And finally, autonomous twins leverage AI models to add on behalf of occupants or self-tune the facility. As you can see in Ana's case, DB is right now performing predictive twins. There is a road to-- there is a roadmap in order to evolve in a comprehensive and autonomous digital twin.

To support twin building during project delivery, Autodesk Tandem builds upon Autodesk Platform Services and integrates with several products. This includes Autodesk Docs, our file-based common data environment. We bring model data into Tandem for design tools like Revit and formats like IFC, and then we can capture metadata using commonly used tools like Microsoft Excel. In the future, we'll be expanding both the model formats we support, and we'll include an integration with asset model-- the assets models of Autodesk Build to support file asset capture during the construction phase.

While twin building is very much a mobile-centric experience, we recognize that operations professionals need a different experience to unlock the operational outcomes. The experience for building operations is start with curated dashboards that provide a broad view of the data that matters.

The dashboards enable users, firstly, to monitor critical data, discover potential issues, examine trends, identify anomalies, and review recommendations. From the dashboards, the user can drill down into an anomaly or recommendation to investigate the potential issue, examining the data in detail, and understanding relationships and navigating complex data with ease. Using those insights leads to an informed decision on a corrected course of action.

And thank you. Thank you, Ana, and thank you for having us in this class today. It was a pleasure, and see you next time.