Security researchers identify "fingerprints" in 3D printed objects that can be used to trace their manufacturing


Originally published at:


Well, this may work providing that there is no change to the infill type, percentage, speed, filament type, filament diameter, feed multiplier, temperature, layer height, layer width, mmm… and probably a few other things I’m forgetting … orientation on the build plate, acceleration, JKN values, retraction settings. Probably more.

No doubt a fun an interesting exercise for the researchers. Not expecting a CSI: Silicon Valley any time soon. “Well detective, it looks like this was printed on a modified WanHao with an all-metal hot end, probably by a 36-yr old male with a limp.”


Oh, you know this is more a television forensics thingy than a real-life issue. The fancy huge screen that all of these shows have will show some schematic of a generic printer in wireframe, point to the nozzle, and then the smartass guy in glasses will tell the heroes that the imperfections match this printer, which is owned by Joe Blow, and they will then show his face, and some numbers, which are his credit card receipts for purchasing just the right sort of plastic which matches the gun, and so on, and so on.


They can’t even analyze rape kits with 30-year-old technology. Plus, NRA is going to defend anybody’s right to own any weapon, and use it for any reason, and accountability is for the Reds.


Reminds me of the typewriter forensics from The Lives of Others. I agree this is a bit far fetched for practical use, but it’s a good reminder that there’s a lot of scary shit in the world.


Anyone who has ever worked retail knows that the thought of the FBI instantly getting receipts for the exact thing they’re looking for is laughable.


hmmm , a reasonable quasi-random number generator for the infill g-code generation in the slicer seems possible , but silly !! of course , i am rather fond of silly !!


You mean 76.6% of the population, right?


This could be fun, though, one way or another.


Even with all the print settings changed I would think that the physical features of the printer would also be detectable in the print. The stepper motors for instance probably vibrate in a certain way. There might be an imperfection on a ball bearing. A printing axis may not be perfectly square, and so on.


Verily, you’d think that “lower-resolution, more fragile output” would present an insurmountable obstacle to consistent “slight imperfections in infill”.


That would just mean that you could loosen your steppers or rack gears and tighten then right back down and poof new fingerprint.


Just a different temperature in the room is enough to make my prints go from ok, to terrible. There’s so many things that have comparatively large effects on the print, I don’t think you could do much more than pinpoint the material used, and probably the nozzle size with any degree of accuracy.


I wonder if the smoothing technique of exposing it to acetone vapor would make things less clear. I suppose the inside infill would look the same.


filament based 3d printing is far too inconsistent for any of this to actually matter. ambient temperature and humidity create huge changes in print quality. the bed leveling is always off and always different every time you print. air bubbles in filament from manufacturing. I may use a .4mm nozzle today and tomorrow switch to .2mm. Maybe when i printed that bad thing i was using the stock hot end, but a week later upgraded to all metal.

The very notion of being able to identify who printed something by how the printer printed it, is just dumb.


Also, you’d need to track who bought what, when, where, etc. And while you might be able to do it if you had all the data, that’s still a fairly tall order.

( * inserts argument against ‘microstamping’ on firing pins *)


The very notion of being able to identify who printed something by how the printer printed it, is just dumb.

So you think the people who wrote the article are just making it up? It appears that it is indeed possible to detect the origins of a certain print just by looking at the small imperfections in it.


They wouldn’t be the first people in the ‘security’ industry flogging snake oil, mind.


This is true. I’m reminded of “lead analysis”, the debunked practice of looking at lead in a bullet and determining which batch of bullets it came from at the manufacturer.