skip to content

Construction Information Technology Laboratory

Automating the process of creating digital twins of rail infrastructure

PhD Candidate Mahendrini Ariyachandra is researching methods for automated generation of geometric digital twins of existing and under construction railways, specifically focusing on the rail assets of track structure. The cost and effort of modelling existing rail infrastructure from point clouds currently outweigh the perceived benefits of the resulting model. Studies show that the time required for generating a geometric railway information model, i.e. a holistic data model which has recently become known as a "Digital Twin", of an existing railway from point cloud data is roughly ten times greater than laser scanning it. Hence, there is a persistent need to automate the process of creating digital twins of rail infrastructure.

The preliminary step of Ms. Ariyachandra’s research is to automatically detect masts from air-borne LiDAR data, as their position and function (separating substructure from superstructure) is critical to the subsequent detection of other rail assets. To accomplish this, a novel method has been proposed which tackles the challenge above by leveraging the highly regulated and standardised nature of railways. Railway infrastructure geometric relations remain roughly unchanged within established regions. This method starts with cleaning the point cloud data through reducing its arbitrary positioning and orientation. Datasets are then processed to restrict the search for masts relative to the distance from the track centreline. A prototype was developed and implemented with the Point Cloud Library (PCL) C++ on Visual Studio. The solution was tested and verified with three railway point cloud datasets with a cumulative length of 18km. The experimental results culminated in an overall detection rate of 93.8% for three datasets.

For further information, contact Mahendrini Ariyachandra at


© Mahendrini Ariyachandra

Latest news

PhD studentships advertised as part of the Cloud-based Building Information Modelling (CBIM) training network

13 November 2019

The following PhD studentships are available at the University of Cambridge on the European Commission-funded, Marie Sklodowska-Curie "CBIM" European...

Infrastructure Computer Vision – Book available for pre-order

28 October 2019

Pre-order the soon-to-be-released Infrastructure Computer Vision to explore advanced information and sensing technologies, and how to apply them to individual projects.