Unique Fingerprints: How Every 3D Printer Leaves Its Own Mark on Printed Parts

Scientists have found that every 3D printer leaves a unique «fingerprint» on the objects it produces, which makes it possible to accurately trace their origin. A research team from the University of Illinois, led by Professor Bill King, has developed an AI tool that can identify 3D printers based on these distinct «fingerprints» with an impressive accuracy of 98%.

According to the researchers, each 3D printer creates unique microscopic defects and characteristics on the surfaces of printed items. These distinctive features have turned out to be sufficiently unique and recognizable to pinpoint the specific printer that manufactured a given part.

Professor King’s team created an AI system trained to analyze these microscopic traits. The model was trained using 9,192 images of parts printed on 21 printers from six different brands, employing four distinct 3D printing technologies. After training, the system could match a part’s image to a specific printer with 98% accuracy.

«We are still amazed that this works. We can print the same part on two identical printers of the same model, with identical settings, using the same material—and each machine will leave a unique fingerprint that the AI model can trace,» explains Bill King.

The researchers are optimistic that their discovery will enhance the monitoring of part origin and print quality in manufacturing processes. Professor King believes that the technology can be refined to reduce the time needed for sourcing identification, significantly simplifying defect detection in large-scale production.