Machine learning algorithms improve product accuracy in additive manufacturing

According to foreign media reports, Purdue University and the University of Southern California have created a machine learning algorithm designed to improve the quality control of 3D printing. The idea is to assemble the various components of the vehicle more precisely and reduce test time.

The new software is designed to solve the shortcomings of additive manufacturing, which often faces a big problem: the parts produced need to achieve "high" accuracy and repeatability. The software program utilizes a "model-building algorithm" to analyze product data and correct computer-aided design modeling to improve geometric accuracy. Therefore, the printed parts produced by it can exhibit high precision.

Arman Sabbaghi, Assistant Professor of Data Analysis at Purdue University, USA: “The software has been used in many industries, such as aerospace, where equipment in the field needs to be extremely precise in geometry, designed to ensure reliability and safety. ""

He added that this innovative technology is important to “make everyone a manufacturer”.

The researchers worked with the Purdue Office of Technology Commercialization to apply for a patent for the technology, and the parties are still looking for partners to continue the technology development.

The project received financial support from the National Science Foundation.