Loughborough University and Project 23rd Century Limited KTP 24_25 R2

A mobile phone video of a Sunday league match could soon be enough to flag a future professional footballer. This project builds an AI system that watches amateur football footage and recognises players’ movements, skills, and potential — the kind of assessment currently reserved for scouts at elite academies. Talent identification today is patchy and biased: clubs focus on a tiny pool of players who can afford coaching and travel, while thousands of promising kids in under-resourced areas are never seen. The AI aims to democratise that process by analysing any game filmed on a phone, spotting technical and tactical ability without requiring expensive equipment or expert observers. If it works, the system could widen the talent pipeline for professional clubs, but its real impact may be on the grassroots infrastructure that quietly feeds the sport. Local leagues, school teams, and youth clubs could upload footage and receive objective assessments, giving coaches in low-income regions a tool they currently lack. The project is applied — it builds a working product — but its success depends on solving fundamental computer vision problems: recognising individual players in chaotic, low-quality footage and distinguishing meaningful skill from noise.