Today 3D printers have won a great prominence, especially in the industrial production environment, although being increasingly accessible to consumers themselves.
What nobody could imagine is that sooner or later A technique capable of identifying the traces of these machines with 98 % precision.
This is so, because it turns out that each impression It has a unique fingerprint that allows it to track it to the machine that produced it.
This was discovered by researchers from the University of Illinois, who found a reliable method of associating a 3D printer with the exact machine that manufactured it.
For this they created an artificial intelligence system, capable of analyzing and identifying which machine manufactured each of the components.
To create this artificial intelligence model, The researchers provided 9192 photographs of pieces that were printed on 21 machines built jointly by six brands and with four different manufacturing processes.
Then they provided an image of 1 mm² and a printed piece, and the system managed to set the impression on the machine with a 98%accuracy.
«We are still surprised that this works: we can print the same piece design on two identical machines (same model, same process configuration, same material) and each machine leaves a unique mark that the AI model can track to the machine,» King said. «It is possible to determine exactly where and how something was manufactured. You don’t have to trust your supplier’s word in any aspect.»
This would facilitate the detection of failures in large -scale products manufacturing processes.
That is, if for example, the quality of the creation of a product is poor, those responsible could quickly track the 3D printing to the machine that manufactured it so that they could improve or discard it, finding along the way, the failures in the production chain.
Know How we work in NoticiasVE.
Tags: Artificial intelligence