Pointcloud BIM model railway created from a point cloud

The BIM model for railways obtained from point clouds is a process that utilizes 3D laser scanning technology to capture precise data of railway infrastructures. This data is converted into 3D BIM models, which provide a detailed and accurate digital representation of existing infrastructures, improving visualization, asset management, and railway project planning. The conversion from point clouds to BIM models reduces the need for manual measurements and minimizes the risk of errors, bringing particular benefits in predictive and corrective maintenance of railway infrastructures.

In this article, you will discover the technologies used in this particular process and the advantages that follow. Additionally, a video will be available to show you the final result of the integration between GIS, BIM, and point clouds, also using a platform for managing IFC models, point clouds, textured meshes, etc. geolocated and integrated with GIS.

Inserting a survey of a railway station with point cloud

What is the point cloud BIM model railway?

To obtain a digital, intelligent, and manageable railway BIM model, the use of point clouds is essential. This approach allows for an accurate representation of the real state of complex infrastructures such as an existing railway network.

The point cloud BIM model railway is a three-dimensional digital representation of the railway infrastructure, obtained from laser scanner or photogrammetric surveys, used for design, management, and maintenance.

Data from the survey is imported into BIM software and used as a reference base for digital modeling. Within the BIM environment, the point cloud is treated as an informational element of the model: the same operations intended for other BIM objects, such as walls, tracks, terrain, bridges, tunnels, and stations, can be applied to it, thus facilitating the faithful reconstruction of the infrastructure.

Point clouds are generated through 3D laser scanning or photogrammetry, technologies capable of capturing millions of points in three-dimensional space. The result is an extremely detailed representation of the real conditions of railway infrastructures, essential for developing reliable and up-to-date BIM models.

How does Point Cloud technology work?

The point cloud is a collection of three-dimensional spatial points that represent the outer surface of an object or space. These points are obtained through 3D laser scans that emit beams of light and measure the time it takes for them to return, thus creating an accurate digital image of the existing structure. This technique is particularly useful for complex environments, such as underground tunnels and railway bridges, where irregular geometry makes traditional measurement methods difficult​.

Integration of the point cloud into the railway BIM model

After acquisition, the point cloud data is imported into BIM software. Here, it is used as a basis for creating detailed three-dimensional models of railway infrastructure. The BIM not only offers an accurate visual representation but also integrates functional and technical information, such as materials, dimensions, and properties of objects, useful for managing the project lifecycle​.

For these two topics, I recommend reading the article “The point cloud at the service of BIM” to better understand how this technology works and how it could be useful for the world of infrastructure.

Data Acquisition and BIM Modeling Phases

Integrating point cloud data into a BIM model is a complex but highly efficient process that allows for the creation of accurate and detailed three-dimensional models of railway infrastructure. This digital approach is essential for managing and planning existing infrastructure and for implementing new projects.

The entire process can be divided into the following 4 phases:

• data acquisition through 3D laser scanning of the site;
• data processing and cleaning to eliminate noise and imperfections;
• BIM import and modeling for both the design and project management phases;
• model validation.

Data Acquisition

The first phase of the process involves data acquisition using 3D laser scanners or photogrammetry techniques. Laser scanners emit light pulses that, reflecting off the surfaces of structures, allow for the creation of a three-dimensional point cloud. This point cloud accurately represents the geometries of the surveyed environment, capturing every detail of the railway infrastructure, such as tracks, bridges, and tunnels. This phase is crucial, as the accuracy of the scan will influence the quality of the final model.

The start of modeling based on a 3D survey provides the technician with an updated reference of the existing state, allowing for the replacement or integration of potentially outdated or incorrect 2D or 3D representations.

Data Processing and Cleaning

Once acquired, the raw point cloud data requires preliminary processing to remove errors, unnecessary points, and noise that may have been captured during scanning. During this phase, filtering is performed to reduce the amount of unnecessary data and to optimize the accuracy of the model. The data cleaning ensures that the digital representation remains true to reality, reducing the risk of inconsistencies that could negatively impact subsequent modeling​. Additionally, there is also assembly/registration, which involves combining the various point clouds captured on-site, aligning them with each other. There are various methods to perform this operation:

• alignment – point clouds are aligned based on the number of common points, overlap speed, tilt, and point density;
• registration via planes – clouds are aggregated according to the number of common vertical and horizontal planes;
• registration with targets and spheres – this method, usually the most accurate, uses targets and spheres placed on the ground to register the clouds semi-automatically.

BIM Import and Modeling

After cleaning, the point cloud data is imported into BIM software. This phase includes converting the collected points into usable three-dimensional geometries. Depending on the project’s needs, one can proceed with creating a detailed 2D or 3D model that accurately represents all detected structural elements, such as tracks, bridges, and mechanical, electrical, and plumbing (MEP) systems. The most common BIM software for this type of operation allows for easy handling of large volumes of point cloud data.

There are various techniques that can be used, which also vary depending on the software employed. For this reason, I want to share the in-depth article “Scan to BIM, from laser scanner survey to the BIM model” that clarifies the ideas.

Phases of the Scan to BIM process

Deep learning and artificial intelligence in point cloud segmentation

The segmentation of point clouds is a fundamental process for 3D analysis and modeling, especially in the context of infrastructures that involves the “decomposition” of a point cloud into multiple elements. The difficulty in applying this technique lies in recognizing an element: how does software distinguish a specific set of points and define them as tracks?

For this, deep learning techniques, such as the so-called DGCNN (Dynamic Graph Convolutional Neural Networks), come to the rescue, which are used to improve the accuracy of semantic segmentation.

The semantic segmentation of 3D point clouds with deep learning techniques can indeed help to automatically recognize the technical elements of a railway at an appropriate level of detail, allowing for an accelerated modeling process starting from survey data.

This approach can naturally also be used in other fields and disciplines, such as in the modeling of HBIM (Historical Building Information Modeling) models.

This is a very interesting and complex topic, so I recommend reading “The transformative impact of Artificial Intelligence on structural analysis” which specifies in more detail all the necessary aspects to understand the use of artificial intelligence in the construction world.

Integration of the point cloud in BIM for collaborative design

Collaborative design is the key to the BIM approach, and the integration of the point cloud into BIM is essential to make it possible.

To enable the shared management of point clouds, appropriate cloud platforms are used, such as usBIM, through which you can use a point cloud viewer for free and manage point clouds online.

With a point cloud viewer, you can:

• visualize the point cloud;
• generate infinite sections;
• manage the survey from any internet-connected device;
• share the file directly with a link that leads to the cloud folder;
• take measurements directly on the model;
• federate multiple models simultaneously in the same view;
• add annotations and external links to the model to create a true informational container to share with colleagues.

Starting from the point cloud, or from portions of it, you can proceed with BIM modeling, obtaining a digital model consistent with the point cloud, which can then be shared in the cloud again for data management and comparison.

From laser scanning to BIM: automating infrastructure modeling

The workflow necessary to draft a project for the construction or renovation of a railway infrastructure is complex and involves many disciplines, including architecture, engineering, and plant engineering.

To bypass the typical slowness of the design process, laser scanner technology comes to the rescue, which has now established itself as a method capable of ensuring speed and precision, as well as allowing for the production of particularly accurate BIM models to support subsequent design phases.

Obtaining a BIM model of a railway through a point cloud with the scan to BIM process guarantees unprecedented accuracy in digital models, eliminating the need for manual measurements and reducing design errors.

The efficiency of the process accelerates design and construction, allowing for faster and more accurate planning thanks to access to detailed 3D data, also facilitating collaboration among teams, improving communication and coordination through shared and real-time updated models.

Moreover, the economic aspect should not be overlooked, as the adoption of scan to BIM techniques reduces the overall project costs, optimizing resource management and minimizing waste through more precise planning.

We can summarize everything in the following points:

• detection of conflicts – the combination of point cloud data with BIM allows for identifying conflicts between new constructions and existing infrastructures already during the design phase;
• accurate as-built documentation – the generated 3D models allow for reliably documenting the real conditions, facilitating future operations in maintenance phase;
• cost reduction – thanks to the accuracy of the data and the automation of processes, the integration between point cloud and BIM reduces material waste and construction times;
• simplified collaboration – all project participants can access updated and detailed information, promoting a faster and more accurate decision-making process.

How point cloud data supports digital twins in railways

The ability to visualize the point cloud directly through a web browser, integrated with the 3D BIM model, has revolutionized the monitoring of the conditions of tracks, stations, and signaling devices. This system allows for quick access to all the properties of the elements, making the management process simpler and faster, and obtaining in-depth analysis of objects or environments, thus contributing to improving the understanding and management of real systems.

Thanks to the use of digital twins integrated with point clouds and BIM, it is possible to plan preventive maintenance interventions and foresee future infrastructural improvements: predictive maintenance, based on real-time data analysis, allows for improved management of the life cycle of infrastructures, reducing operational costs and preventing failures and thus damages: hence, greater energy efficiency, lower environmental impact, and increased safety.

The technology allows for real-time monitoring of railway conditions, intervening promptly to reduce maintenance costs and increase operational efficiency​.

The integration of large amounts of territorial data is managed smoothly, allowing for continuous large-scale control, with significant benefits for sustainability and resource optimization.

A concrete example of this approach is shown in the video below “Pointcloud BIM model railway”, which illustrates the study of 50 km of railway modeled from a point cloud and an openBIM model in IFC 4.3 format, with complete reading of the properties of the objects and the use of artificial intelligence for the segmentation of the point cloud and the generation of the model, within a platform for managing Geospatial Digital Twin.

FAQ – Point cloud BIM model railway

What is a BIM model of a railway obtained from a point cloud?

It is a 3D BIM model of the railway infrastructure created from survey data (3D laser scanner or photogrammetry). It serves to reconstruct the real state of the project and to support design, management, and maintenance.

Why is the point cloud important in railway BIM?

Because it provides a very accurate “as-is” basis of the existing infrastructure (tracks, bridges, tunnels, stations). This way, BIM modeling starts from real and updated data, reducing errors and inconsistencies.

What is the difference between 3D laser scanning and photogrammetry for creating point clouds?

Both generate point clouds but with different methods: the laser scanner captures data through laser pulses, while photogrammetry reconstructs geometry from images. The choice depends on context, required precision, and operating conditions.

What are the main phases to convert a point cloud into a BIM model?

Typically, the workflow includes:

• data acquisition (3D survey),
• processing and cleaning,
• BIM import and modeling,
• validation of the final model.

What is the purpose of cleaning and registering point clouds?

It serves to remove noise and unnecessary points and to correctly “merge” multiple scans into a single coherent dataset. This step improves the reliability of the BIM model that will be built on the point cloud.

What advantages does Scan to BIM offer in railway infrastructures?

It allows for obtaining more accurate and faster models to develop, reduces manual measurements and errors, improves planning, and facilitates activities such as as-built surveys, interference checks, and maintenance support.

How is a point cloud managed collaboratively in BIM?

Through cloud platforms and online viewers that allow for visualizing the point cloud, creating sections, measuring, adding annotations, and sharing files via links, making it easier for teams to compare.

How do point clouds and BIM support digital twins in railways?

The integration between point clouds and BIM models makes monitoring the conditions of the infrastructure and managing information more effective, promoting preventive/predictive maintenance and more in-depth analyses of the life cycle.