Engineers working to stabilise the Toowoomba Range in Queensland are using cutting-edge artificial intelligence technology to conduct 3D virtual inspections of the railway, allowing them to identify changes and potential hazards in the land.
Transport and Roads minister Mark Bailey said the technology was allowing engineers to view difficult-to-access areas of the range remotely, helping them plan required upgrades.
“Through this trial with software company Skand, our engineers are able to see the Toowoomba Range in a way they’ve never seen it before,” he said.
“This state-of-the-art technology transforms the range into a 3D map which engineers can ‘fly’ through, helping them assess the impacts of rainfall, rockfalls or landslips on our infrastructure.
“Engineers can also zoom in on objects like rock formations and see elevation and angle data, providing them with a level of visibility that traditional geotechnical inspections can’t provide.
“The State Government is committed to delivering better services right across Queensland, which is exactly why we’re championing innovative technology like this to continue improving our rail network.
“It forms part of our record four-year $29.7 billion transport and road plan for Queensland, supporting 25,200 good, secure jobs throughout the state.”
Queensland Rail Head of Regional Jim Benstead said the Toowoomba Range was an important transport corridor for the Darling Downs coal and agricultural sector.
“Innovative trials, like the one with Skand, are helping us learn more about how we can better monitor the Toowoomba Range and help avoid disruptions in the future,” he said.
“It’s vitally important we maintain the condition of this rail line to support local industry and economy through reliable transport solutions.”
Brisbane-based company Aerial Drone Solutions captured both photogrammetry and LiDAR for the 24km project.
Director Roy Moser said drones took over 40,000 images for the 3D model and millions of laser pulses captured the formation of the range through vegetation for LiDAR and that data was then fed into Skand’s software.
“We flew multiple drones over part of the range using a LiDAR Scanner and RGB cameras. The scanner fires out millions of laser pulses and then collects them to reconstruct the terrain in high detail to create a Point Cloud. The Point Cloud is then fed into Skand’s software to visualise the 3D reconstruction. This gives QR a full virtual model of the entire corridor,” he said.
Skand co-founder Brett Chilton said the power of the technology could be seen when two different data sets from different times were matched together.
“When you get multiple data sets and layer them in the one portal you can really get insights into what is changing on the range, and we hope to explore this as the trial continues,” he said.
“In other case studies, the tech has developed machine learning algorithms that can identify and classify defects, reducing the need for physical inspections, decreasing costs and safety risks.
“The technology allows Queensland Rail engineers to see through the vegetation of the trees which is really important because they can see the underlying features on the ground such as boulders and areas of interest.”
Stage 2 of the Toowoomba Slope Stabilisation project continues on the range, with the construction tender for remediation works of the slope at 141km released in the first week of September.
The works include soil nailing, shotcreting, scour protection and improved drainage.