The digitalization of railroad infrastructure is aimed at the improvement of maintenance and construction activities. Currently, inspections are done manually, with a domain expert classifying objects. This is a challenging task, considering the Netherlands has more than 3,400 km of railways that need to be inspected and maintained.
Ambient Intelligence collaborates with Strukton Rail to work with point clouds, which are sets of spatial data points captured by 3D scanning techniques such as lidar. These point clouds contain many million points of data, resulting in 3D representations of the railway environment. In this project, we focus on the application of deep learning to automate the segmentation of the point clouds in separate parts of the catenary system, such as the poles, bars, conductors and resistors. This automation is intended to support and replace the manual inspection of the catenary system.
Topic
Point clouds, deep learning
Program objectives
Key research challenges are how to cope with large point cloud datasets efficiently and automatically, how to apply state-of-the-art deep learning algorithms to this use case, and how to communicate segmentation results clearly. The main outcome will be a software prototype which ingests a given point cloud dataset to give feedback on the discovered catenary subsystems in the dataset.
Partners
Strukton Rail
Duration
November 2020 – October 2021
More information
Financing
This project is financed by TFF