Sheffield Hallam University and Preformed Windings Limited KTP 24_25 R2

A factory's production planning system is getting a machine-learning upgrade that connects sales orders, staff schedules, and quality checks into a single, self-improving digital brain. Most manufacturers still plan production using spreadsheets, gut instinct, or rigid software that cannot adapt when a rush order arrives or a machine breaks down. This project builds a scalable system that continuously pulls data from sales, human resources, and quality control departments, then uses machine learning and artificial intelligence to forecast capacity and spot bottlenecks before they happen. The key innovation is that the system learns from ongoing operations—every new order and delay feeds back into the model, making future predictions more accurate without manual reprogramming. If successful, the system could let a factory respond to last-minute changes in minutes rather than days, reduce wasted materials, and keep delivery promises more reliably. The impact sits inside the quiet machinery of manufacturing: better coordination between departments means fewer emergency overtime shifts, less idle equipment, and a supply chain that absorbs shocks rather than amplifying them. The project is applied research with a clear commercial route—the partner company, Preformed Windings Limited, will deploy the system directly in its own production lines.