The prospect of 3D-printed guns terrifies law enforcement. Along with being DIY for anyone with a 3D printer, the weapons are, in theory, untraceable because they don’t contain the identifying serial number that regular firearms have.
Researchers at University at Buffalo, however, report they’ve discovered the first way to track 3D-printed objects using the unique “fingerprints” that 3D printers leave on whatever objects they produce.
Lead researcher Wenyao Xu told VICE News that he started looking into ways to track 3D-printed objects in 2015 after learning of self-described crypto-anarchist Cody Wilson’s legal battle against the U.S. State Department. Wilson sued the State Department on free speech grounds after they forced him to take his 3D-printed gun blueprints offline.
During their research, Xu’s team noticed that 3D-printed guns have slight imperfections, much like the variations on fingers. They’re unique to each printer but uniform across all the objects created by it.
“3D printers are built to be the same,” Xu wrote in a press release. “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.”
Here’s how the researchers’ tracking system, dubbed the “PrinTracker,” works: 3D printers discharge material, usually plastic, in layers until they form an object. According to researchers, each layer has tiny wrinkles, which are supposed to be uniform. But each printer, together with its nozzle size and the type of plastic used, causes miniscule imperfections. Researchers refer to that as its “fingerprint,” which can be as tiny as a half millimeter.
The research team tested their system by printing five door keys each from 14 different commercially available 3D printers. After creating digital images of each key, they developed an algorithm to calculate variations of the imprints on the key down to the millimeter. They then added that information to a database.
According to the study, the researchers were able to match the key to its printer 99.8 percent of the time. They ran another round of tests 10 months later “to determine if additional use of the printers would affect PrinTracker’s ability to match objects to their machine of origin,” according to the study’s release. The results were exactly the same.
“Like human beings, they [3D guns] have their own writing signature,” Xu said. “Finding these signatures could help forensic experts and police.”
But the team’s method has its limitations. “Fingerprinting” 3D guns requires the same tool used for fingerprinting people: a database.
“For this to be useful to law enforcement who would be tracing guns, you’d have to have a record of the unique signature of every 3D printer being sold,” said David Chipman, senior policy adviser at Giffords Law Center to Prevent Gun Violence, who spent 25 years at the Bureau of Alcohol, Tobacco, Firearms and Explosives. “We can only match it if we already have a record of it,” Chipman said.
When tracing regular firearms, law enforcement can also rely on the serial number that's already, in theory, linked to its owner. On the other hand, “fingerprint” recovery is a forensic process that’s usually done after a crime has been committed. The only way “fingerprinting” 3D guns could compare to the traditional method of tracing firearms is if 3D printers were regulated to the extent that their “fingerprints” were logged in a database alongside identifying information of whoever buys it.
Xu’s “PrinTracker” study, which was co-authored by researchers at Rutgers University and Northeastern University, will be presented later this week in Toronto at the Association for Computing Machinery’s Conference on Computer and Communications Security.
Cover image: Cody Wilson, with Defense Distributed, holds a 3D-printed gun called the Liberator at his shop, Wednesday, Aug. 1, 2018, in Austin, Texas. (AP Photo/Eric Gay)