Nearly every machine has a distinct sound signature and that includes 3D printers with their multiple stepper motors. Thieves who are able to get physically close to a fused deposition modeling (FDM) 3D printer as it prints would be able to record the sounds as the model prints and then, by using sophisticated algorithms, reverse engineer that model and print it.
Dr. Ismail Fidan, Professor of the Department of Manufacturing and Engineering Technology and College of Engineering-Faculty Fellow in Innovation and Techno-Entrepreneurship at Tennessee Technological University and his team conducted a cybersecurity research project that was able to record 3D printer sounds as an object was printed and later replicate that model on another printer by performing a cyber acoustic analysis. A classification and regression model was completed on the audio recorded during the 3D printing process and G-Code data was created to attempt another 3D print.
To date, according to the research paper and results, “Current security measures have focused on securing machines against cyber based attacks with cloud based resources and software programs. Nevertheless, attacks in the physical domain have also occurred… The physical component of AM machines opens up the system to vulnerabilities due to side-channels. Side-channels are indirect pathways that lead to the access of desired data such as obtaining G-Code from vibrational, acoustic, magnetic, or power emissions. Previous analysis of side-channels has been used to infer information about cyber domain data. Therefore, it is important to analyze these side-channels to better secure the system and prevent leakage of IP.”
As other researchers have pointed out, these intellectual property risks are not far-fetched and present a real problem for additive manufacturing. Dr. Fidan and team are proving that it is possible to record and reproduce a 3D print. Built into that research, IP protection methods are sure to evolve. Keep track of Dr. Fidan’s work at Tennessee Technological University and its Additive Manufacturing (aka 3D Printing) efforts.
The University of California, Irvine reported on a similar research project in 2016: Bad vibrations: UCI researchers find security breach in 3-D printing process.
“The team, led by Mohammad Al Faruque, director of UCI’s Advanced Integrated Cyber-Physical Systems Lab, showed that a device as ordinary and ubiquitous as a smartphone can be placed next to a machine and capture acoustic signals that carry information about the precise movements of the printer’s nozzle. The recording can then be used to reverse engineer the object being printed and re-create it elsewhere. Detailed processes may be deciphered through this new kind of cyberattack, presenting significant security risks.”
“Unlike most security hacks, the researchers did not simulate a cyberattack. Many 3-D printers have features, such as encryption and watermarks, designed to foil such incursions. Instead, the researchers programmed a common smartphone’s built-in sensors to measure electromagnetic energy and acoustic waves that emanate from 3-D printers. These sensors can infer the location of the print nozzle as it moves to create the three-dimensional object being printed.”