The need
Tunnel examinations are normally carried out in the middle of the night by a team of tunnel inspectors, with lines requiring closure so that inspections can take place. A team of inspectors walk through the tunnel, manually inspecting it with torches, hammers and sticks. This process is slow, hazardous and antisocial, and produces vague results.
The solution
NPL’s DIFCAM project (digital imaging for condition asset management) is a step towards automating this process. DIFCAM collects highly detailed images and height data from the inside of the railway tunnel which is then used to automatically detect defects in the brickwork and track changes in the tunnel structure and shape over time. Network Rail have invested an estimated £1million with NPL in this technology over the last decade or so.
In November 2024, NPL used NMS funding to significantly progress the technology. Previously, NPL was only able to collect partial datasets due to the time constraints associated with accessing tunnels on the live rail network. The NMS funding resulted in a week of scanning the entire inner surface of the Sharpthorne Tunnel at the Bluebell Railway in West Sussex. This is a Victorian brick-built tunnel which is identical to almost all railway tunnels on Britain’s rail network This is the first ever dataset of its kind to be collected in full. NPL aims to return to the Bluebell Railway in December 2025 to repeat the exercise, so that its Digital Image Correlation technology (DIC) can be used to show how the tunnel has changed over a year.
As part of the same exercise, NPL collaborated with Quantum Structures, a company that carries out large-scale ground penetrating radar (GPR) scans of structures including railway tunnels, with their TSIR machine. NPL asked Quantum Structures to scan
Sharpthorne Tunnel at the same time as their own inspections . The aim was to combine both datasets so that NPL not only knew what was happening on the surface of the tunnel, but also what might be happening behind the brickwork structure, up to several metres into the geology. This is an extremely powerful tool and something which had never before been attempted.
GPR data can be processed in the same way as DIC, which means that the underlying geology of the tunnel can also be tracked over time. NPL is hoping that Quantum Structures will repeat their scans in a year’s time so that their data can also be tracked over a 12-month period.
The impact
This exercise generated significant industry interest as it marks the beginning of a new generation of tunnel examination. Network Rail sent a Principle Tunnels Engineer to observe the activities along with a project manager from their Civils & Sustainability Programme. The rail regulator, Office of Rail and Road, also heard about the investigations and sent their Chief Civil Engineer to observe.
At the time of writing, Quantum Structures have taken advantage of NPL’s M4B programme to attempt to combine the two datasets for the first time. This has never previously been attempted, either at this scale or with such differing datasets. This includes significant challenges, including the handling and processing large datasets (up to 10TB in size), the combination and layering of differing datasets (both in type and size), and the integration of the resulting dataset into a useful viewer for the end-user, such as geographic information system (GIS) software as used by Network Rail and other civil engineering organisations.
Looking forward, NPL now has sound experience in large-scale data collection and can provide solutions for many of the unanswered questions that organisations will encounter as they transition structural health monitoring towards automation.