# Data repository

## Content

This repository consists of data from a survey conducted with a terrestrial laser scanner ([Riegl VZ600i](https://www.riegl.com/de-deutschland/products/detail/riegl-vz-600i)).

The 3D digital model was generated on the basis of the 3D point cloud data from the scan and integrated into [Unreal Engine](https://www.unrealengine.com/en-US/unreal-engine-5), 
the available project contains additional custom functionality for sensor data integration from multiple sources:

* WIM sensor system integration via OSC client (B02, ISAC, RWTH Aachen University)
* Artificial Loading data from sensor mat (CPM Lab) via OSC (B03, i11, RWTH Aachen University)
* Vehicle position data from the CPM labs perception system via MQTT (B03, i11, RWTH Aachen University)
* Pavement response simulation results from (B05, ifam, RWTH Aachen University) via OSC client

Some digital assets from the Fab digital asset library were used, the particular assets are listed here: 

* Vehicle Variety Pack - [Switchboard Studios](https://www.fab.com/listings/dc1ada50-2523-44b1-b0e2-a72d14076fb4)
* Megascans Meadow Pack - [Megascans](https://www.unrealengine.com/marketplace/en-US/product/megascans-meadow-pack)
* Industry Props Pack 6 - [SilverTm](https://www.fab.com/de/listings/b5603e44-e1b0-4346-9c3d-04887aa9f87d)
* Eurpean Beech - [Quixel](https://www.fab.com/listings/d11cc01d-9422-41b7-950f-416c9ce79caf)


The output of the laser scanner is stored in Hierarchical Data Format ([HDF5](https://www.hdfgroup.org/solutions/hdf5)). 
[HDF5](https://www.hdfgroup.org/solutions/hdf5/) files can be accessed by software tools like [HDFView](https://www.hdfgroup.org/download-hdfview) or coding languages, e.g. Python using package [`h5py`](https://docs.h5py.org/en/stable/index.html).

The point cloud data can be visualized with point cloud inspection software like [CloudCompare](https://www.cloudcompare.org) 

### file ``VLS.rar``

This file contains a packaged Unreal Engine project containing the 3D model of the WIM Research Site an map and the data connection handlers detailed above, which hook into data streaming topics via ports of a local network in our project.
[Unreal Engine](https://www.unrealengine.com/en-US/download) can be downloaded for free.
The Version used is 5.2 and should be used to exclude potential compatibility issues, the directory includes a simple setup instruction manual. 

### file ``WIM_Research_Site.h5``

This file contains the point cloud data (points location, coloring, intensity and classification) recorded by the laser scanner including their metadata.


### file ``WIM_Research_Site_Survey_Segmented_PC.zip`` 

This file contains the full TLS scan data as LAS segments per class from the annotation process.
The data serves as a foundation for different use cases, such as the use case of semantic segmentation. The derivative for semantic segmentation is stored in WIM_Research_Site.h5. This file provides segments with higher density, allowing for object-specific analyses. 

## References

The data generation as well as the underlying methods, models and software will be described in a future publication on the: "Case Study of a Digital Shadow System for Pavement Loading Monitoring"
