Software Reduces Print Iterations and Optimizes Part Designs in Additive Manufacturing

Araqev (West Lafayette, Ind.) says its software can help users print products in only a few design iterations, reducing scrap material and machining time while improving satisfaction with the final printed products. Founded by Arman Sabbaghi, associate professor in Purdue University’s Department of Statistics in the College of Science, Araqev starts by having users upload their nominal design files and scanned point cloud data from their printed products.

The software then uses these inputs to fit machine learning models that can simulate shape deviations for future printed products and calculate modifications to the nominal designs so that when the revised designs are printed, they reportedly exhibit fewer shape deviations compared to the original.

Araqev says its algorithms also transfer knowledge over time via machine learning models across different materials, printers and shapes. The company says it verified the software’s performance in two validation experiments using a Markforged Metal X 3D printer running 17-4 PH stainless steel. Araqev reported that its algorithms reduced shape inaccuracies in those instances by 30-60%. On the company’s website, it says it has consistently achieved high quality prints in one or two design iterations over more than a dozen case studies using four different printers.

Araqev is establishing direct partnerships with 3D printing manufacturers and companies, seeking to sell licensing contracts for its program.