Norwegian researchers are testing a prototype for a lightweight, autonomous rail inspection system, set to slash maintenance costs. Part of the EU-funded IAM4RAIL project, the AI-powered prototype, equipped with LiDAR and high-resolution cameras, has been mounted on a service train and is now operational on Norway’s Sørland Line.
Currently, Norway relies on a single, highly specialised inspection train, named Roger1000, to inspect its 4,000 kilometres of track. But deploying this heavy, slow-moving diagnostic train requires halting or rerouting traffic, delaying schedules and driving up costs.
In the Europe’s Rail project IAM4RAIL, co-funded by the European Union, researchers of NORCE Research have developed a lightweight prototype for inspection of railway infrastructure. As a prototype, an inspection system has now been mounted on a Bane NOR service train for trial runs at 40–50 km/h, the next step is to develop a lightweight inspection vehicle. “Our solution is to build inspection vehicles that are small and light enough to be lifted onto the tracks by two people,” explains Atle Aalerud, Senior Researcher at NORCE Research. “The vehicle can be pulled from the tracks to let trains pass before it is put back to continue its work.”
The inspection system
• LiDAR and camera sensors to collect data on the condition of the catenary, tunnel profiles and clearances—key parameters for operational safety
• Autonomous data collection and analysis, enabling real-time deviation reports and reducing reliance on manual inspections.
• The end goal: compact, lightweight vehicles that two people can deploy without disrupting operations.
Impact for rail will be ‘tremendous’
The prototype has now been installed on a service train running test cycles on the Sørland Line, the researchers announced on 20 January, marking the first real-world application of the technology. “Using known technology, we are constructing units that will significantly reduce maintenance costs throughout Europe. The impact will be tremendous,” Aalerud stated.
Unlike Roger1000, which requires dedicated track time and logistical coordination, the new system will be able to be easily moved to avoid delays. The long-term goal is full automation, where the system self-reports deviations and eventually replaces large maintenance trains entirely, according to NORCE Research.
Read more: