A digital factory model is a virtual representation of a real factory that can be used for various purposes such as planning, simulation, optimization or documentation. To create a digital factory model, you first need to capture the real factory in 3D. This can be done using various 3D scanning methods, which are presented in this newsletter. Tips are also given for selecting the appropriate hardware, software and techniques to create a high-quality and realistic digital factory model.
3D scanning methods
3D scanning is a process in which the geometry, color and texture of real objects or environments are captured using sensors. The captured data is processed into a point cloud consisting of millions of points, each with an x, y and z coordinate and an RGB color value. A point cloud can be seen as a kind of digital imprint of the real structure, which can be further processed in suitable software.Es gibt verschiedene Methoden des 3D-Scannings, die sich in der Art der verwendeten Sensoren, der Genauigkeit, der Geschwindigkeit, der Reichweite und dem Preis unterscheiden. Die wichtigsten Methoden sind:
Laser scanning: This method uses a laser beam that scans the surface of the object or environment and measures the distance and angle to each point. Laser scanners can be divided into two categories: terrestrial and mobile. Terrestrial laser scanners are set up stationary on a tripod and can achieve very high accuracies (up to 2 mm) and ranges (up to 500 m). However, they are relatively expensive and require multiple scanning positions to ensure complete coverage.
Mobile laser scanners are carried by hand or on a vehicle and can continuously capture data as they move through the environment. They are faster and more flexible than terrestrial scanners, but have a lower accuracy (up to 6 mm) and range (up to 130 m).
LiDAR scanning: This method is a variant of laser scanning that uses the time of flight or phase shift of reflected laser pulses or waves to determine the distance to any point. LiDAR scanners can be used in various platforms, such as airplanes, drones, vehicles or robots. They can cover large areas quickly and efficiently, but have a lower resolution and color quality than laser scanners.
Photogrammetry: This method uses multiple images taken from different angles by a camera to reconstruct the 3D structure of the object or environment. The images are analyzed using an algorithm that identifies matching points and calculates the camera position and orientation. A point cloud is then generated from the image points, which can be enriched with color information. Photogrammetry can be carried out with various cameras, such as the Apple iPhone 12 Pro and newer. It is inexpensive and easy to use, but has lower accuracy and reliability than laser or LiDAR scanning.
Tips for choosing the right hardware and software
Selecting the appropriate hardware and software for 3D scanning depends on various factors, such as the goal of the planning task, the scale and complexity of the factory, the available budget and the required accuracy and quality of the digital factory model. Here are some general tips that can help with the selection process:
- Define the goal of the planning task: Before you start scanning, you should know what you want to achieve with the digital factory model. For example, whether you need a rough overview of the factory structure or a detailed representation of the machines and systems. This will help you to determine the necessary level of detail and the data quality requirements.
- Evaluate the size and complexity of the factory: The size and shape of the factory will influence the choice of scanning method and the number of scan positions. Large and open areas can be captured quickly with LiDAR or photogrammetry methods, while small and angled areas can be captured more accurately with laser scanning methods. Factory complexity refers to the number and type of objects to be scanned, such as machines, pipes, cables, racks, etc. Complex objects require higher resolution and color quality, which can be achieved with laser or photogrammetry methods.
- Consider the available budget: The budget is an important factor that limits the choice of hardware and software. The cost of 3D scanning varies depending on the method chosen, the hardware and software required, the duration of scanning and the processing of the data. In general, laser and LiDAR scanners are more expensive than cameras, and specialized software for processing and modeling point clouds is more expensive than general software. You should therefore weigh up the cost-benefit ratio of the different options and choose the one that best suits your requirements.
- Compare the accuracy and quality of the different methods: The accuracy and quality of the 3D data depend on the scanning method you choose, the hardware and software used, the scan settings and the environmental conditions. You should therefore compare the advantages and disadvantages of the different methods and choose the one that offers the best balance between accuracy, quality and speed. You should also check the compatibility of the data formats with the software you intend to use for further processing and modeling.
Procedure for creating a digital factory model
Once you have captured the 3D data of the real factory, you can start creating a digital factory model. This step requires the use of specialized software that is able to process and model the point clouds. Here are some tips that can help you achieve a good result:
- Choose the right software for point cloud processing: There are various software solutions for point cloud processing that offer different functions and services. Some examples are: Faro Scene, Leica Cyclone, Autodesk ReCap, Bentley ContextCapture, Trimble RealWorks, PointCab, etc. You should choose the software that best fits your data and requirements, and that offers high compatibility with the data formats of the scanning hardware and modeling software. You should also consider the ease of use, speed, quality and price of the software.
- Mesh the point clouds carefully: To convert the point clouds into a digital factory model, you need to mesh them, i.e. create a surface of triangles that approximates the shape of the objects or areas. This step requires a lot of computing power and can take a long time depending on the size and complexity of the point clouds. You should therefore carry out the meshing carefully to ensure the quality and accuracy of the model. This includes adjusting the meshing parameters, such as the resolution, smoothing, simplification, hole filling, etc., checking the meshing quality, such as the number, size, shape and orientation of the triangles, the consistency, completeness and topology of the surface, etc., and correcting the meshing errors, such as the gaps, overlaps, spikes, artifacts, etc.
How artificial intelligence can help
Creating a digital factory model using 3D scanning is a complex and demanding process that requires a lot of time, resources and expertise. To make this process easier and faster, artificial intelligence (AI) can be used to analyze, understand and optimize the 3D data. Here are some examples of how AI can help:
- AI can improve the meshing of point clouds by determining the optimal resolution, smoothing, simplification, hole filling, etc. for each point cloud or segment, checking the meshing quality and correcting the meshing errors. This can reduce meshing time and data volume and increase model accuracy and consistency. One example of this is the PointFuse software.
- AI can simplify the modeling of the digital factory model by replacing the meshed point clouds or the segments with predefined or generated 3D models of objects or areas that better represent the real geometries and textures. This can reduce modeling time and manual editing and improve model realism and aesthetics. This is an approach taken by Riiico, for example.
Conclusion
3D scanning and modeling of real structures such as buildings and machines is a powerful technique that offers many possible applications in the field of digital factory planning. 3D scanning can be used to create detailed and accurate models of the real environment that can be used for planning, documentation, analysis or simulation. The models can also be connected to the metaverse to create virtual worlds that offer users an authentic experience. However, 3D scanning also requires appropriate hardware, software and methodology to ensure the quality of the data and reduce the amount of data. 3D scanning and modelling of real structures is a promising technology that will be further developed and optimized in the future to open up new possibilities for factory planning and other areas.
