Brunel University London and Hadley Group Holdings Limited KTP 24_25 R2

A British manufacturer is teaching artificial intelligence to design the complex steel rollers that bend flat sheet metal into precise, three-dimensional shapes. Roll forming is a quiet workhorse of modern manufacturing. It produces everything from car body panels to structural beams, window frames, and shelving. But designing the tooling—the sequence of rollers that gradually bends a flat strip into a finished profile—remains a slow, expert-driven craft. Engineers rely on trial and error, accumulated experience, and physical testing. Mistakes are expensive: a poorly designed roller can scrap thousands of pounds of steel before the problem is caught. This project aims to replace that guesswork with AI. By training machine learning models on successful and failed tooling designs, the system will learn to predict the optimal roller geometry for a given profile. The goal is to cut design time from weeks to hours, reduce material waste, and allow the company to take on more complex, custom orders that competitors cannot handle efficiently. If it works, the impact is straightforward: a British engineering firm stays competitive in a global market where speed and precision decide who wins contracts. The public will not see the AI, but they will see the products it enables—lighter cars, stronger buildings, and cheaper shelving—made faster and with less waste.