ARTC carry out regular fortnightly corridor track patrols across its 8,500+km network. This is conducted using hi-rail vehicles and trained specialist rail inspectors. The process is costly, dangerous and prone to human errors due to high cognitive load on operators and the access requirements on the network. ARTC is scoping available technologies to augment and or replace entirely manual track patrols by using forward-facing revenue train mounted high resolution cameras and machine learning analytics.
Cordel.ai has been working with ARTC since early 2014 and in 2017 we began working to provide an alternative program or solution to their manual hi-rail track patrol program. Cordel.ai utilises forward-facing high-resolution video cameras, GNSS and IMU positioning as well as LiDAR to create a digital twin of the network. Once data is captured it is automatically analysed by our Convolutional Deep Learning Neural Networks (CDLNN) or simply machine learning, platform to automatically detect, classify and grade the condition of the assets, defects and safety incursions inside the railway corridor.
Through successful collaboration and extensive testing, ARTC and Cordel are continuing to work on the system. As of Jan 2020 and as part of the current AK Car scanning program further investigative works have been undertaken to assess the viability of this technology across the entire Australia wide network.
"Results from trials have been encouraging and we are continuing to embed this new capability as the preferred method of inspections for transit space monitoring across the entire ARTC Interstate Network."
Damon Goulding, Asset Planning Engineer
Transform your infrastructure inspections
Visit our contact page to get started today or book a demo to give us the opportunity to show you first hand how we enable deeper insights into critical infrastructure.