Researchers at the Grainger College of Engineering at the University of Illinois Urbana-Champaign, US, have developed a new artificial intelligence system, capable of precisely identifying the 3D printer that manufactured an individual part, thanks to "fingerprints” invisible to the naked eye but nevertheless left on each printed item.

It may seem surprising, but these researchers have discovered that every 3D printer (even among identical models) leaves unique micro-variations on the surfaces of the parts they produce. These microscopic differences can now be detected by an AI system trained to recognise these specific signatures.

To test their tool, the researchers analysed some 9,192 photos of parts printed by 21 different machines, covering a total of four additive manufacturing processes and six different types of material. In this context, the AI model is currently able to achieve 98% accuracy in identifying the origin of a part from a simple photograph of one square millimeter of its surface. The idea now is to extend the study to a wider range of printing machines and manufacturing processes.

This new technology offers many promising prospects. For example, it could help to improve the traceability of objects, by ensuring the authenticity of parts and detecting counterfeits. It could also help quality control by easily identifying the machines responsible for any manufacturing defects. It could even be used in criminal investigations to trace the origin of 3D-printed weapons.

For manufacturers, it could optimise the supply chain, with enhanced monitoring of suppliers and compliance with agreed manufacturing processes.  – AFP Relaxnews