Birmingham City University and Coba Holdings Limited KTP 24_25 R3

A small manufacturer will replace human spot-checks on its production line with automated cameras that inspect every single item coming off the line. The problem is straightforward: many small and medium-sized manufacturers still rely on workers to check only a tiny fraction of what they produce. This misses defects, wastes materials, and limits how fast they can operate. The company in this project wants to switch from sampling a few items to inspecting every item, but doing that manually for large batch sizes is impractical. Computer vision and artificial intelligence can do the job instead. If this works, the manufacturer will catch faulty products immediately rather than discovering them later in the supply chain. That means less waste, fewer customer complaints, and faster production. The approach could also be adopted by other small manufacturers in sectors such as packaging, electronics assembly, or food processing—anywhere that high-volume, repetitive visual checks are still done by eye. This is an applied, commercial project. It does not aim to advance fundamental science. It takes existing computer vision and AI tools and adapts them to a specific factory-floor problem that has not yet been solved for this company.