Some of the biggest concerns surrounding the proliferation of 3D printing technologies have to do with the possibility of 3D printed weapons and of counterfeit 3D printed goods. Though many solutions to the counterfeit risk involve hidden identification codes within 3D models, a team of researchers has developed a slightly different approach.
The approach, developed by a team from the University at Buffalo, comes in the form of a project named PrinTracker, which is capable of identifying what 3D printer a given part was created on based on inherent imperfections in infill patterns. The novel approach is believed to be the first technique for accurately tracing an object to the 3D printer it was made on.
Though still in its development phase, PrinTracker could eventually be employed by law enforcement and intelligence agencies to track where 3D printed firearms and counterfeit goods originate from.
“3D printing has many wonderful uses, but it’s also a counterfeiter’s dream,” commented Wenyao Xu, PhD, an associate professor of computer science and engineering at UB’s School of Engineering and Applied Sciences and the study’s lead author. “Even more concerning, it has the potential to make firearms more readily available to people who are not allowed to possess them.”
With this in mind, Xu and his team devised a method for identifying the subtle differences between objects printed on different 3D printers. These differences, the researchers explain, are found on a submillimeter scale in infill patterns. Though the patterns are supposed to be uniform and dictated by the digital 3D model of the printed object, a number of elements (such as printer model, filament type, nozzle size etc.) can cause tiny imperfections in the print.
Interestingly, it turns out that a given 3D printer model will produce parts with unique and repeatable infill imperfections, creating something of a fingerprint within 3D printed parts. Xu elaborates: “3D printers are built to be the same. But there are slight variations in their hardware created during the manufacturing process that lead to unique, inevitable and unchangeable patterns in every object they print.”
In a recent study, the team tested PrinTracker by using 14 3D printer models (10 FDM and four SLA) to 3D print a series of five keys each. Once the prints were complete, an inkjet scanner was used to digitally capture each of the printed keys. These scans were then enhanced and filtered so that the infill patterns were distinguishable.
From there, the team used an algorithm to align and calculate the variations of each key’s print pattern to check the authenticity of the “fingerprint,” which enabled them to establish a database of fingerprints for each of the 14 3D printers in the test.
In running the printed keys through the algorithm, PrinTracker able to match the key to the right 3D printer 99.8% of the time. A follow-up test 10 months later further demonstrated that even with more use of the 3D printers, the fingerprints remained the same and were identifiable.
Going a step further, the team even tested keys that had been damaged in some way, and PrinTracker was still able to identify their originating 3D printer with 92% accuracy. “We’ve demonstrated that PrinTracker is an effective, robust and reliable way that law enforcement agencies, as well as businesses concerned about intellectual property, can trace the origin of 3D-printed goods,” Xu concluded.
Xu and his team will be presenting the innovative PrinTracker project at the Association for Computing Machinery’s Conference on Computer and Communications Security this week in Toronto. The team also includes researchers from Rutgers University and Northeastern University.