EYYA: Enhancing train maintenance with AI-powered condition reporting

This case study presents EYYA, a company applying artificial intelligence and computer vision to automate train condition monitoring. Addressing the limitations of manual inspection processes, EYYA’s TRACK system captures real-time image and video data to detect cleanliness issues and surface defects such as graffiti and scratches, enabling more efficient and objective maintenance decisions. Through support from the BridgeAI Independent Scientific Advisor (ISA) offer, delivered by the Alan Turing Institute, EYYA strengthened the performance and robustness of its AI models. The collaboration introduced key improvements including data augmentation strategies, refined labelling processes, and model optimisation techniques to address challenges such as imbalanced datasets and overfitting. These advances enabled the transition from early-stage models to deployment-ready systems in operational rail environments. In parallel, alignment with emerging standards such as ISO/IEC 42001 supported the development of responsible AI practices, enhancing trust and credibility with clients. This case demonstrates how combined technical and governance support can accelerate AI adoption, enabling scalable, trustworthy solutions with both commercial and societal impact. Learn more: https://zenodo.org/records/17781656