CC Informatics showcases AI-driven structural inspection tool 'AssetScan' at Sellafield demonstration

CC Informatics (CCI) has successfully demonstrated its innovative AssetScan technology to a group of 30 Sellafield staff, highlighting the potential of artificial intelligence to transform visual inspections in the nuclear sector. 

Held at the Centre of Excellence (CoE) Leconfield, the demonstration marked the conclusion of CCI’s proof-of-concept (PoC) project under the Game Changers open innovation programme. AssetScan was originally submitted through the programme’s General Application route, which allows solution providers to propose technologies that may benefit organisations including Sellafield Ltd, NRS Dounreay, and NRS Magnox. 

AssetScan is a patent-pending, trademarked AI product that automates the detection and measurement of defects in structural assets. Initially developed for use in the water industry, where it was applied to inspect masonry structures, the technology has now been adapted to suit the unique materials and conditions found on nuclear sites – including concrete, cladding and exposed steelwork. 

The demonstration outlined how AssetScan integrates high-resolution visual inspection data with photogrammetry and cutting-edge artificial intelligence. The process begins with either site-collected or provided imagery, which must meet specific criteria to enable accurate 3D modelling – namely image overlap and feature consistency across at least three views. 

Using industry-standard software, CCI constructs a photogrammetric 3D model of the structure. AssetScan’s AI engine, built on convolutional neural networks (CNNs), then processes the raw images to detect structural anomalies. These enhanced images are overlaid onto the 3D model, and orthorectified views of key surfaces are exported for post-processing and defect measurement. 

Throughout the PoC, CCI applied AssetScan to a range of Sellafield structures, with a particular focus on concrete and steel elements. The demonstration showcased the results of these trials, sparking insightful discussions among attendees about real-world applications in engineering and asset condition monitoring.

"AssetScan represents a paradigm shift in visual inspection of large infrastructure, because our technology can be used to assess very large infrastructure holistically – which is something that a manual reviewer cannot do. It is a scalable solution that automates defect identification and tracking in huge image datasets – meaning more rapid detection of change, and provision of metrics that can assist human decision makers," said a CCI spokesperson. "We’ve demonstrated that AI, when combined with remotely collected survey data and 3D modelling, can significantly enhance the way visual inspection data is managed and reported on entire infrastructure systems – meaning better decisions, sooner, and resulting in better records to be kept for the future."

As the nuclear sector seeks new ways to improve safety and operational efficiency, AssetScan stands out as a promising tool to support digital transformation in asset management. With its successful adaptation from the water industry to nuclear applications, the technology may soon play a vital role in supporting the long-term integrity of critical infrastructure.