QuietPrint: Protecting 3D Printers Against Acoustic Side-Channel Attacks

The 3D printing market has experienced significant growth in recent years, with an estimated revenue of 15 billion USD for 2025. Cyber-attacks targeting the 3D printing process whether through the machine itself, the supply chain, or the fabricated components are becoming increasingly common. One major concern is intellectual property (IP) theft, where a malicious attacker gains access to the design file. One method for carrying out such theft is through side-channel attacks. In this work, we investigate the possibility of IP theft via acoustic side channels and propose a novel method to protect 3D printers against such attacks. The primary advantage of our approach is that it requires no additional hardware, such as large speakers or noise-canceling devices. Instead, it secures printed parts by minimal modifications to the G-code.

💡 Research Summary

The paper “QuietPrint: Protecting 3D Printers Against Acoustic Side‑Channel Attacks” investigates the feasibility of stealing intellectual property from additive manufacturing systems by exploiting the acoustic emissions of a 3D printer. The authors first identify three primary sources of sound: the power‑supply cooling fan (which contributes negligible information), the stepper motors that drive each axis, and the nozzle cooling fans. By recording audio with a standard laptop microphone placed near an Elegoo Neptune 3 printer, they demonstrate that the fan noise in the 6–9 kHz band correlates linearly with the nozzle’s X‑axis position, while low‑frequency spikes (0–1 kHz) correspond to sudden direction changes of the stepper motors.

A realistic threat model is defined in which an adversary controls a nearby audio‑recording device (or an insider places one) and captures the sound during printing. To enable reproducible research, the authors develop a synchronization protocol that timestamps each G‑code command using the M400 “wait for move to finish” instruction, thereby aligning recorded audio with precise nozzle coordinates.

Having established that simple acoustic features (energy of the fan signal, stepper‑motor spikes) are sufficient to reconstruct the tool path, the paper proposes a purely software‑based defense called QuietPrint. The core idea is to obfuscate the acoustic signature by making minimal, non‑intrusive modifications to the G‑code: (1) insert random micro‑movements (e.g., 0.1 mm) to break the regular stepping pattern of the motors, (2) vary feed rates slightly to randomize fan speed and thus the high‑frequency fan noise, (3) add deliberate pauses using M400 to disperse the low‑frequency spike pattern, and (4) optionally inject tiny extrusion commands to further perturb fan operation. These changes do not materially affect part geometry or mechanical properties, and the overall printing time increases by less than 5 %.

Experimental evaluation shows that, with the original G‑code, a machine‑learning model can recover the nozzle trajectory with over 90 % accuracy. After applying QuietPrint, the same model’s accuracy drops below 20 %, and reconstruction of complex internal features becomes practically impossible. The authors argue that QuietPrint offers a cost‑effective countermeasure that eliminates the need for expensive speakers or acoustic enclosures, making it suitable for both hobbyist and industrial settings.

The paper concludes by highlighting the broader implications: acoustic side‑channel attacks are a genuine risk for additive manufacturing, but they can be mitigated at the software level. Future work is suggested on extending the approach to other printer architectures (Core‑XY, delta) and on integrating defenses against multi‑modal side‑channels that combine acoustic, vibration, power, and magnetic emissions.