Years of experience are required for part optimization in sheet metal design in order to identify both errors and optimization potential in a design. "And when the expert retires, he takes all the knowledge with him," says Martina Trinczek, describing a dilemma in sheet metal production - which became the founding idea for her and Jonas Steiling. The company accelerates the design of simple sheet metal parts on its online platform with the help of machine learning and its own algorithms. The "experience" used for this are thousands of 3D CAD data, which are grouped into three categories "highly optimizable", "optimizable" and "not optimizable" and analysed using conventional design rules and internally developed feature engineering.

Users upload their CAD file, select the material from which the part is to be manufactured and activate the repair mode. They are immediately shown the optimization potential as well as errors and warnings. Warnings show visual influences on the component, such as cut-outs that are too close to the bending zone. Errors, on the other hand, alert the user to problems in the design, such as a minimum leg length that has not been reached, intersecting bends or components that cannot be unwound.

With the repair mode, Optimate already offers automated correction of the component with subsequent download of the CAD file for some features. The aim is to offer the user a fully automated solution that reduces material waste and replaces costly production steps, such as the welding process. "Our vision is part optimization that works with maximum reliability and absolute efficiency," says Jonas Steiling. "This enables our customers to design more sustainably, reduce costs and manufacture more efficiently." At Blechexpo, the founding team will also be presenting an optimization algorithm that can be used for components without rule-based design errors (warnings and errors) but with high optimization potential.