Pith. sign in

REVIEW 3 major objections 8 minor 38 references

A hardware firewall that watches a 3D printer's motors, temperatures, fans, and endstops can catch firmware attacks that alter the print or damage the machine.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 11:21 UTC pith:BX2TJLM7

load-bearing objection Solid, non-invasive hardware prototype that actually catches the firmware attacks it claims to catch; the trusted-G-code assumption is the real limit, not a hidden flaw. the 3 major comments →

arxiv 2607.10484 v1 pith:BX2TJLM7 submitted 2026-07-11 cs.CR eess.SP

Firewall3D: A Hardware Firewall for Defending 3D Printers Against Firmware Attacks

Seyed Ali Ghazi Asgar , Narasimha Reddy This is my paper
classification cs.CR eess.SP
keywords 3D printingfirmware securityhardware firewalladditive manufacturingcybersecurityattack detectionphysical layer monitoring
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Firmware-level attacks on 3D printers can silently change motion, temperature, extrusion, or cooling so that parts come out weak, designs leak through side channels, or the printer itself is damaged. This paper claims that a dedicated intermediate board, Firewall3D, can stop those attacks by sitting between the printer motherboard and its actuators and sensors, continuously comparing physical signals against the G-code that is supposed to be running. The board measures stepper-motor currents, converts them into position and speed, and also watches hotend and bed temperatures, fan PWM, and endstop switches. When observed behavior drifts beyond small tolerances from the intended command, the system raises an alarm and can halt the print. Experiments on a Marlin-based Cartesian printer show sub-millimeter position accuracy and successful detection of length, speed, thermal, fan, and endstop-spoofing attacks without modifying the printer's own firmware or motherboard.

Core claim

Firewall3D can detect a wide range of firmware attacks that compromise print integrity, damage components, or leak intellectual property by monitoring physical-layer signals in real time and verifying that the printer's actual behavior matches the intended G-code; on mismatch it can alarm and stop the print.

What carries the argument

Firewall3D, a bump-in-the-wire hardware monitor that samples stepper-motor coil currents (decoded via atan2 into position and speed), temperature sensors, fan PWM-to-DC voltages, and endstop lines, then compares them against expected G-code values.

Load-bearing premise

The intended G-code must be known securely to Firewall3D or its host PC, and an attacker must not be able to reprogram Firewall3D itself or tamper with its communication channel.

What would settle it

Run a firmware attack that alters motion length, speed, temperature, fan duty, or endstop behavior while Firewall3D is installed; if the board fails to flag the mismatch or produces false alarms on legitimate prints across a range of feed rates and temperatures, the detection claim fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 8 minor

Summary. The paper presents Firewall3D, a custom “bump-in-the-wire” hardware monitor placed between a 3D printer’s motherboard and its actuators/sensors. Using an STM32-based PCB with Hall-effect current sensors, PWM-to-DC converters, buffered temperature/endstop paths, and an atan2-based decoder of dual-coil stepper currents (Algorithm 1, §4.9), it reconstructs nozzle motion, speed, temperatures, fan duty, and endstop events and compares them to an externally supplied intended G-code stream. Controlled experiments report ±0.1 mm position resolution (Table 1, Appendix Table 2), speed error under 1% on fixed-length moves (Fig. 10), and successful detection of hand-crafted motion-length, motion-speed, thermal-setpoint, fan-speed, and endstop-spoofing attacks (§6, Figs. 11–15). On anomaly the system can alarm and halt printing. Limitations (§7) explicitly assume a trusted G-code reference and an uncompromised Firewall3D/PC channel.

Significance. If the result holds under the stated threat model, this is a useful systems contribution: the first purpose-built, non-invasive hardware firewall for commodity FFF printers that closes a gap left by software-only and invasive (soldered) monitors. Strengths that should be credited include a fully realized PCB and circuits (§4.2–4.8), a concrete, reproducible position decoder (Algorithm 1), quantitative accuracy tables rather than qualitative claims, and coverage of three attack classes (motion, thermal/fan, endstop) drawn from prior firmware-attack literature. The work is timely given the growth of AM and documented firmware threats (Marlin, FLAW3D, bio-printer attacks). The main value is engineering demonstration and a clear physical-layer monitoring architecture, not a new cryptographic or formal security theorem.

major comments (3)
  1. §3 Threat Model and §4.6 / §7: The central detection claim (“physical behavior matches the intended G-code”) is load-bearing on a trusted, uncompromised G-code reference delivered to Firewall3D (or its host PC) while the printer may receive a maliciously altered command. All attack simulations in §4.6 and §6 are constructed exactly this way (original command to Firewall3D, modified command to the printer). The paper acknowledges this in §7 but does not design or evaluate a concrete secure G-code path (e.g., authenticated streaming, measured SD-card path through the host, or cryptographic binding). Without that, the system detects motherboard/firmware deviation only when the reference is already known to be correct; an attacker who can also influence the reference channel (or the host that supplies it) falls outside the demonstrated guarantee. This assumption should be elevated into the t
  2. §5–§6 Evaluation: Detection is shown only on short, hand-crafted G-code snippets and controlled trajectories (square paths, step temperature ramps, single-axis homing). There are no false-positive or false-negative rates on long multi-layer production prints, no characterization of threshold sensitivity under vibration, thermal drift, or filament-load variation, and no multi-printer or multi-firmware validation. Table 1 and Figs. 10–12 establish reconstruction accuracy under ideal conditions; they do not establish operational detection reliability. For a security claim that Firewall3D “can effectively detect a wide range of firmware attacks” and safely halt printing, the manuscript needs at least (i) FP statistics on several full real G-code jobs and (ii) a clear statement of detection latency and halt mechanism under continuous streaming. Without these, the “wide range / effective” clai
  3. §1 Related work and comparison to SCREAM [7]: The paper correctly notes that [7] is invasive (soldering between MCU and driver) while Firewall3D measures after the driver and is more portable. However, there is no side-by-side quantitative comparison of detection coverage, latency, position/speed error, or attack classes. Given that the authors themselves state [7] “achieves performance similar to our system,” a short comparison table (signals monitored, invasiveness, accuracy, attacks demonstrated) is needed so readers can judge the incremental contribution. As written, novelty relative to the closest prior hardware monitor remains under-specified.
minor comments (8)
  1. Abstract and §1: Market figure “exceeding $30 billion” and Apple Watch Ultra 3 example are fine for motivation but should cite the exact MarketsandMarkets / Apple sources already listed in the bibliography for reproducibility.
  2. §4.9 / Algorithm 1: The conversion “every ten π/4 increments = 0.1 mm” is printer-specific. State the steps/mm or microstepping configuration used, and note how recalibration would work on a different machine (currently only implied).
  3. §4.2 Fan circuit: PWM frequency is given as 7.8 Hz for “our 3D printer.” Confirm whether this is typical Marlin fan PWM or a board-specific setting; many boards use higher frequencies, which would require retuning the RC filter.
  4. Fig. 9 caption says temperature increments of 25°C while the body text (§5.2) says 50°C steps from 50°C to 200°C. Align caption and text.
  5. Fig. 10 caption says “motion path of length 200mm” while the surrounding text discusses 20 mm motions and Appendix Fig. 18 uses 25 mm. Clarify the length used for each speed-error plot.
  6. §6.4 thermal model: α ≈ 2.3 °C/s and β = 10% are reasonable but presented without uncertainty bounds or dependence on ambient temperature / PID gains. A short sensitivity note would strengthen the defense claim.
  7. Typos / consistency: “Firewall3d” vs “Firewall3D”; “Avgerage” in Table 1; “scenraio” in Fig. 15 caption; “commends” in Fig. 14 caption; duplicated reference [2]/[3] (same HOST 2025 bio-printer paper).
  8. §2.1: “logical high signal of 0 V” for released endstop is confusing; rephrase as active-low / open-collector behavior if that is the case.

Circularity Check

0 steps flagged

No significant circularity; physical-signal matching to known G-code is independent experimental validation, not a self-referential derivation.

full rationale

Firewall3D's core claim is an engineering result: a bump-in-the-wire monitor that samples stepper currents (via TMCS1107 + atan2 phase decoding + printer-specific scale of ten π/4 increments = 0.1 mm), PWM-to-DC fan/bed signals, NTC temperatures, and end-stop states, then flags mismatches against an externally supplied G-code reference. The conversion factor, heating-rate coefficient α ≈ 2.32 °C/s, and speed-timing windows are ordinary printer-specific calibrations obtained from controlled G-code runs (Tables 1–2, Figs. 8–10, 13); they are not fitted to attack outcomes and then re-labeled as predictions of those same outcomes. Attack scenarios (§6) are simulated by deliberately sending the original command to Firewall3D and a mutated command to the printer; detection is therefore a direct comparison of independent physical observables, not a tautology. Self-citations ([5], [6]) address prior side-channel or streaming work by the same authors and are not load-bearing for the Firewall3D detection logic or uniqueness claims. The trusted-monitor assumption (§7) is an explicit threat-model boundary, not circular reasoning. The paper is therefore self-contained against its own experimental benchmarks; score 1 only for the presence of non-load-bearing self-citations.

Axiom & Free-Parameter Ledger

3 free parameters · 4 axioms · 1 invented entities

The central detection claim rests on a small set of printer-specific calibration constants, standard physical assumptions about stepper and thermal dynamics, and an explicit attacker model that leaves the monitor itself trusted. No new physical entities are postulated; Firewall3D is an engineered artifact, not a theoretical invention.

free parameters (3)
  • position conversion factor (ten π/4 increments = 0.1 mm) = 0.1 mm per 10 increments
    Printer-specific scale used in Alg. 1 and §4.9 to turn angular steps into millimeters; must be measured for each machine.
  • heating-rate coefficient α ≈ 2.3 °C/s = ≈2.32 °C/s
    Empirically measured slope of hotend temperature rise (§6.4, Fig. 13) used to set waiting time t_wait.
  • safety factor γ and tolerance β = 10 % = β = 0.1 T_set
    Hand-chosen multipliers that define the acceptable temperature band after t_wait; alter detection sensitivity.
axioms (4)
  • domain assumption Physical signals after the motor drivers and on the sensor lines faithfully reflect the commands actually executed by the (possibly compromised) firmware.
    Core premise of the entire monitoring approach (§4.1, §3).
  • ad hoc to paper The intended G-code is known to Firewall3D / the host PC and the communication channel between them is secure.
    Stated as an assumption in §7; without it comparison is impossible.
  • ad hoc to paper Attacker cannot modify Firewall3D firmware or its wiring after installation.
    Explicit threat-model boundary in §3 and §7.
  • domain assumption Stepper motors produce approximately circular current trajectories whose phase can be recovered by atan2.
    Standard bipolar stepper physics used in §4.9 and Fig. 7.
invented entities (1)
  • Firewall3D hardware platform no independent evidence
    purpose: Real-time physical-layer monitor and anomaly detector for 3D printers.
    The engineered artifact that realizes the defense; independent evidence is the experimental measurements on the prototype, not an external physical prediction.

pith-pipeline@v1.1.0-grok45 · 21844 in / 3013 out tokens · 37937 ms · 2026-07-14T11:21:11.378049+00:00 · methodology

0 comments
read the original abstract

As the 3D printing market continues to grow rapidly, with an estimated value exceeding $30 billion, cybersecurity risks and attacks targeting additive manufacturing systems are also increasing. These attacks aim to sabotage printed components, steal intellectual property, or even physically damage the 3D printer itself. One major cybersecurity threat in this domain is firmware level attacks, which can be introduced through supply chain compromises, malicious firmware updates, or insider threats that deploy modified firmware to manipulate printer behavior. To defend against such threats, we propose a dedicated hardware based security solution,Firewall3D, that acts as a hardware firewall for 3D printers. Firewall3D continuously monitors physical layer signals, including stepper motor currents, end stop switches, nozzle and bed temperatures and cooling fans, to verify that the printer's physical behavior matches the intended G-code execution. Our experimental results demonstrate that Firewall3D can effectively detect a wide range of firmware attacks that could compromise print integrity, damage printer components, or leak intellectual property. Upon detecting abnormal behavior, the system can immediately trigger an alarm and halt the printing process, thereby preventing further damage and risks.

Figures

Figures reproduced from arXiv: 2607.10484 by Narasimha Reddy, Seyed Ali Ghazi Asgar.

Figure 1
Figure 1. Figure 1: All components are connected to a main motherboard, which controls [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Firewall3D PCB overview. printed object, leaking the design’s intellectual property through side-channel mechanisms or even damaging the 3D printer machine. To defend against these threats, we propose a dedicated hardware-based so￾lution that monitors physical layer signals including motion commands, temper￾ature sensor readings, fan speed values, and other critical signals in real time. By independently o… view at source ↗
Figure 3
Figure 3. Figure 3: Overview of the Proposed Approach to detect firmware attacks in real time. In this section, we provide an overview of the components of our hardware designed to secure the 3D printer [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Proposed circuit for converting the fan PWM signal into a DC voltage. The resulting DC signal is then read by the microcontroller’s ADC. this, we reuse the same PWM-to-DC conversion circuit shown in [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Circuit for acquiring the hotend temperature. The pull-down resistor configu￾ration is already present on the mainboard; therefore, only a buffer circuit is added to enable non-intrusive temperature measurement. 4.5 End Stop Keys Monitoring The same buffer configuration used for the temperature measurement ( [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Stepper motor current measurement design. two connectors. The first connector interfaces with the 3D printer’s motherboard, which provides the stepper motor drive signals. The current measurement sen￾sors are placed inline with these signals, and the measured currents are then routed through a second connector directly connected to the stepper motor. Since our system includes four stepper motors correspond… view at source ↗
Figure 7
Figure 7. Figure 7: Decoding stepper motor current signals to infer motion. (a) Simplified illustra￾tion of stepper motor operation showing the periodic excitation of two coils and the resulting rotor positions during a rotation cycle. (b) Representation of the measured coil currents, where the current of the first coil is plotted against the second, revealing a circular trajectory corresponding to rotational motion. (c) Angu… view at source ↗
Figure 8
Figure 8. Figure 8: The output of the duty cycle to DC voltage circuit [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Performance evaluation of the temperature measurement system. Color tran￾sitions denote the initiation of successive temperature setpoints. The temperature is increased in increments of 25 °C, with each setpoint held constant for 30 seconds. in the digital output signal. The calibration process for each axis involves two consecutive contacts with the endstop switch. Specifically, the switch transitions thr… view at source ↗
Figure 10
Figure 10. Figure 10: Speed estimation errors for a motion path of length 200 mm. The mean and maximum errors are shown for different speeds. The maximum error is less than 1%. 6 Firmware Attacks and Defense Scenarios We examine several attack scenarios established in prior research and demon￾strate the effectiveness of our defense method in mitigating these threats. Based on our observations, firmware attacks have four main t… view at source ↗
Figure 11
Figure 11. Figure 11: Comparison of normal and attacked motion execution. The left plot shows correct 30 mm movements, while the right plot highlights a compromised command with a 25 mm deviation detected by Firewall3D. Firewall3D expects a motion of 30 mm; however, only 25 mm of motion is executed [PITH_FULL_IMAGE:figures/full_fig_p016_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Comparison of 3D printer normal operation and under attack. Normal Mode maintains consistent speed ( 1500 mm/min) across all four edges, while Attack Mode shows the drop at Motion speed for the top edge drawing, with speed dropping to 1000 mm/min. Firewall3D can detect the anomaly. 6.3 Temperature Targeted Attacks Cooling fans, the hotend nozzle, and the heated bed are key components respon￾sible for temp… view at source ↗
Figure 13
Figure 13. Figure 13: Transient thermal response of the nozzle showing the rise time from 50 °C to 200 °C and from 50 °C to 150 °C [PITH_FULL_IMAGE:figures/full_fig_p018_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Temperature and cooling fan monitoring during normal operation and a firmware attack scenario. In the normal operation phase, the G-code commends as well as their effect on temperature and cooling fan is shown. During the attack phase, the temperature and fan speed goes out of the valid range. to drop to 0. Once the fan speed falls outside the expected threshold range, the system detects the anomaly and r… view at source ↗
Figure 15
Figure 15. Figure 15: The Normal Scenario (top) shows standard calibration operations where the Y-axis endstop switch is triggered twice during each calibration with the (G28 Y) command and with a On-Off-On-off pattern. This axis is calibrated twice in the Normal scenario. The Attack Scenario (bottom) demonstrates the simulated attack condition (G1 Y-85) instead of the homing procedure. In the attack scenraio, the calibration … view at source ↗
Figure 16
Figure 16. Figure 16: Simulation of the proposed circuit for converting duty-cycle values into a DC voltage for fan-speed estimation [PITH_FULL_IMAGE:figures/full_fig_p026_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Distribution of G-code motion lengths on logarithmic scale. The histogram shows 4.29 million motion commands from 80 different G-code files. Speed Sensitivity Analysis An important aspect of reporting error values and evaluating a system’s mea￾surements is to contextualize these numbers. For example, is a 3.83 mm/min speed error significant or negligible? To address this, we performed a sensitivity analys… view at source ↗
Figure 18
Figure 18. Figure 18: Speed estimation errors for a motion path of length 25 mm. The mean and maximum errors are shown for different speeds [PITH_FULL_IMAGE:figures/full_fig_p030_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Sensitivity of the measured speed to variations in path length. The plot illus￾trates how deviations in the motion length, given a 0.5 mm measurement resolution, affect the calculated speed [PITH_FULL_IMAGE:figures/full_fig_p031_19.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

38 extracted references · 3 linked inside Pith

  1. [1]

    Global 3D Printing Market Size, Share, Latest Trends & Growth Analysis, 2024-2029 — marketsandmarkets.com.https://www.marketsandmarkets.com/ Market-Reports/3d-printing-market-1276.html, [Accessed 14-07-2025]

    2024

  2. [2]

    In: 2025 IEEEInternationalSymposiumonHardwareOrientedSecurityandTrust(HOST)

    Ahsan, M., Najarro-Blancas, B., Ebode, J.T., Lewinski, N., Ahmed, I.: 3d bio- printer firmware attacks: Categorization, implementation, and impacts. In: 2025 IEEEInternationalSymposiumonHardwareOrientedSecurityandTrust(HOST). pp. 99–110. IEEE (2025)

    2025

  3. [3]

    In: 2025 IEEEInternationalSymposiumonHardwareOrientedSecurityandTrust(HOST)

    Ahsan, M., Najarro-Blancas, B., Ebode, J.T., Lewinski, N., Ahmed, I.: 3d bio- printer firmware attacks: Categorization, implementation, and impacts. In: 2025 IEEEInternationalSymposiumonHardwareOrientedSecurityandTrust(HOST). pp. 99–110 (2025).https://doi.org/10.1109/HOST64725.2025.11050047

    doi:10.1109/host64725.2025.11050047 2025

  4. [4]

    Journal of Manufacturing Processes145, 211–235 (2025)

    Ali, H., Cano, A., Ahmed, I.: Machine learning-based early detection of malicious g-code manipulations in 3d printing. Journal of Manufacturing Processes145, 211–235 (2025)

    2025

  5. [5]

    arXiv preprint arXiv:2602.02198 (2026)

    Asgar, S.A.G., Reddy, N.: Quietprint: Protecting 3d printers against acoustic side- channel attacks. arXiv preprint arXiv:2602.02198 (2026)

    Pith/arXiv arXiv 2026

  6. [6]

    arXiv preprint arXiv:2507.06421 (2025)

    Asgar, S.A.G., Reddy, N., Bukkapatnam, S.T.: Never trust the manufacturer, never trust the client: A novel method for streaming stl files for secure additive manu- facturing. arXiv preprint arXiv:2507.06421 (2025)

    Pith/arXiv arXiv 2025

  7. [7]

    In: Proceedings of the ACM Asia Conference on Computer and Communications Security

    Basu Roy, P., Blocklove, J., Bhargava, M., Pearce, H., Krishnamurthy, P., Sinanoglu, O., Gupta, N., Khorrami, F., Karri, R.: Scream: Secure channels for real-time evaluation of additive manufacturing. In: Proceedings of the ACM Asia Conference on Computer and Communications Security. pp. 234–249 (2026)

    2026

  8. [8]

    Smart and Sustainable Manufacturing Systems1(1), 142–152 (2017)

    Baumann, F.W., Ludwig, T., Darwin Abele, N., Hoffmann, S., Roller, D.: Model- Data Streaming for Additive Manufacturing Securing Intellectual Property. Smart and Sustainable Manufacturing Systems1(1), 142–152 (2017)

    2017

  9. [9]

    In: 26th USENIX Security Symposium (USENIX Security 17)

    Bayens, C., Le, T., Garcia, L., Beyah, R., Javanmard, M., Zonouz, S.: See no evil, hear no evil, feel no evil, print no evil? malicious fill patterns detection in additive manufacturing. In: 26th USENIX Security Symposium (USENIX Security 17). pp. 1181–1198 (2017)

    2017

  10. [10]

    IEEE Embedded Systems Letters14(3), 111– 114 (2021)

    Beckwith, C., Naicker, H.S., Mehta, S., Udupa, V.R., Nim, N.T., Gadre, V., Pearce, H., Mac, G., Gupta, N.: Needle in a haystack: Detecting subtle malicious edits to additive manufacturing g-code files. IEEE Embedded Systems Letters14(3), 111– 114 (2021)

    2021

  11. [11]

    arXiv preprint arXiv:2506.21897 (2025)

    Chattopadhyay, T., Ceschin, F., Garza, M.E., Zyunkin, D., Chhotaray, A., Stebner, A.P., Zonouz, S., Beyah, R.: One video to steal them all: 3d-printing ip theft through optical side-channels. arXiv preprint arXiv:2506.21897 (2025)

    Pith/arXiv arXiv 2025

  12. [12]

    Progress in materials sci- ence92, 112–224 (2018)

    DebRoy, T., Wei, H.L., Zuback, J.S., Mukherjee, T., Elmer, J.W., Milewski, J.O., Beese, A.M., de Wilson-Heid, A., De, A., Zhang, W.: Additive manufacturing of metallic components–process, structure and properties. Progress in materials sci- ence92, 112–224 (2018)

    2018

  13. [13]

    Benha Journal of Applied Sciences10(7), 9–28 (2025)

    Elgez, A.G.M., Elgedawy, N.K., Nada, O.A.E.: Intellectual property protection for 3d printing products using blockchain technology in the field of industrial design. Benha Journal of Applied Sciences10(7), 9–28 (2025)

    2025

  14. [14]

    Gao, Y., Li, B., Wang, W., Xu, W., Zhou, C., Jin, Z.: Watching and safeguarding your3dprinter:Onlineprocessmonitoringagainstcyber-physicalattacks.Proceed- ings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 2(3), 1–27 (2018) Firewall3D 23

    2018

  15. [15]

    In: Proceedings of the 24th International Symposium on Research in Attacks, Intrusions and Defenses

    Gatlin, J., Belikovetsky, S., Elovici, Y., Skjellum, A., Lubell, J., Witherell, P., Yam- polskiy, M.: Encryption is futile: Reconstructing 3d-printed models using the power side-channel. In: Proceedings of the 24th International Symposium on Research in Attacks, Intrusions and Defenses. pp. 135–147 (2021)

    2021

  16. [16]

    Gibson, I., Rosen, D.W., Stucker, B., Khorasani, M., Rosen, D., Stucker, B., Kho- rasani, M.: Additive manufacturing technologies, vol. 17. Springer (2021)

    2021

  17. [17]

    In: International Con- ference on Security and Privacy in Cyber-Physical Systems and Smart Vehicles

    Jamarani, A., Tu, Y., Hei, X.: Practitioner paper: Decoding intellectual property: Acoustic and magnetic side-channel attack on a 3d printer. In: International Con- ference on Security and Privacy in Cyber-Physical Systems and Smart Vehicles. pp. 54–74. Springer (2025)

    2025

  18. [18]

    In: Security and Privacy in Cyber-Physical Systems and Smart Vehicles: Second EAI International Conference, SmartSP 2024, New Orleans, LA, USA, November 7–8, 2024, Proceedings

    Jamaranil, A., Tu, Y., Heil, X.: Practitioner paper: Decoding intellectual prop- erty: Acoustic and magnetic. In: Security and Privacy in Cyber-Physical Systems and Smart Vehicles: Second EAI International Conference, SmartSP 2024, New Orleans, LA, USA, November 7–8, 2024, Proceedings. vol. 622, p. 54. Springer Nature (2025)

    2024

  19. [19]

    Com- puters in Industry173, 104395 (2025)

    Kumar, M., Epiphaniou, G., Maple, C.: Securing additive manufacturing with blockchain-based cryptographic anchoring and dual-lock integrity auditing. Com- puters in Industry173, 104395 (2025)

    2025

  20. [20]

    Journal of Manufacturing Processes145, 274–285 (2025)

    Mishra, A.K., Goh, S.Y., Ganapathysubramanian, B., Krishnamurthy, A.: Real time 3d reconstruction for enhanced cybersecurity of additive manufacturing pro- cesses. Journal of Manufacturing Processes145, 274–285 (2025)

    2025

  21. [21]

    In: 2016 Resilience Week (RWS)

    Moore, S., Armstrong, P., McDonald, T., Yampolskiy, M.: Vulnerability analysis of desktop 3d printer software. In: 2016 Resilience Week (RWS). pp. 46–51. IEEE (2016)

    2016

  22. [22]

    Moore, S.B., Glisson, W.B., Yampolskiy, M.: Implications of malicious 3d printer firmware (2017)

    2017

  23. [23]

    Engineering Research Express 6(2), 025404 (2024)

    Moulika, G., Palanisamy, P.: Optimizing extreme manufacturing framework: a se- cure and efficient 3d printing integration framework. Engineering Research Express 6(2), 025404 (2024)

    2024

  24. [24]

    IEEE/ASME Trans- actions on Mechatronics27(6), 5361–5370 (2022).https://doi.org/10.1109/ TMECH.2022.3179713

    Pearce, H., Yanamandra, K., Gupta, N., Karri, R.: Flaw3d: A trojan-based cyber attack on the physical outcomes of additive manufacturing. IEEE/ASME Trans- actions on Mechatronics27(6), 5361–5370 (2022).https://doi.org/10.1109/ TMECH.2022.3179713

    arXiv 2022

  25. [25]

    In: International Conference on Availability, Reliability and Security

    Puch, N., Dopfer, S., Birkel, L.: Securing the additive manufacturing process chain. In: International Conference on Availability, Reliability and Security. pp. 340–358. Springer (2025)

    2025

  26. [26]

    In: 18th USENIX WOOT Conference on Offensive Technologies (WOOT 24)

    Rais, M.H., Ahsan, M., Ahmed, I.:{SOK}: 3d printer firmware attacks on fused fil- ament fabrication. In: 18th USENIX WOOT Conference on Offensive Technologies (WOOT 24). pp. 263–282 (2024)

    2024

  27. [27]

    In: International Conference on Critical Infrastructure Protection

    Rais, M.H., Ahsan, M., Sharma, V., Barua, R., Prins, R., Ahmed, I.: Low- magnitude infill structure manipulation attacks on fused filament fabrication 3d printers. In: International Conference on Critical Infrastructure Protection. pp. 205–232. Springer (2022)

    2022

  28. [28]

    Additive Manufac- turing46, 102200 (2021)

    Rais, M.H., Li, Y., Ahmed, I.: Dynamic-thermal and localized filament-kinetic attacks on fused filament fabrication based 3d printing process. Additive Manufac- turing46, 102200 (2021)

    2021

  29. [29]

    https://www.apple.com/newsroom/2025/11/mapping-the-future- with-3d-printed-titanium-apple-watch-cases (2025), [Accessed: 14-07-2025] 24 SA

    Redding, S., Rothberg, N.: Mapping the future with 3d-printed titanium ap- ple watch cases. https://www.apple.com/newsroom/2025/11/mapping-the-future- with-3d-printed-titanium-apple-watch-cases (2025), [Accessed: 14-07-2025] 24 SA. Ghazi Asgar and N.Reddy

    2025

  30. [30]

    In: 34th USENIX Security Symposium (USENIX Security 25)

    Rossel, J., Mladenov, V., Wördenweber, N., Somorovsky, J.: Security implications of malicious{G-Codes}in 3d printing. In: 34th USENIX Security Symposium (USENIX Security 25). pp. 1867–1885 (2025)

    2025

  31. [31]

    International Journal of Computer Integrated Manufacturing pp

    Shomenov, K., Ali, M.H., Jyeniskhan, N., Al-Ashaab, A., Shehab, E.: Cost-effective sensor-based digital twin for fused deposition modeling 3d printers. International Journal of Computer Integrated Manufacturing pp. 1–20 (2025)

    2025

  32. [32]

    In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Commu- nications Security

    Song, C., Lin, F., Ba, Z., Ren, K., Zhou, C., Xu, W.: My smartphone knows what you print: Exploring smartphone-based side-channel attacks against 3d printers. In: Proceedings of the 2016 ACM SIGSAC Conference on Computer and Commu- nications Security. pp. 895–907 (2016)

    2016

  33. [33]

    Jour- nal of Materials Chemistry A13(16), 11804–11816 (2025)

    Sonigara, K.K., Vaghasiya, J.V., Mayorga-Martinez, C.C., Pumera, M.: Point-of- use upcycling of 3d printing waste for developing 3d-printed zn–i 2 batteries. Jour- nal of Materials Chemistry A13(16), 11804–11816 (2025)

    2025

  34. [34]

    Applied Sciences11(11), 5305 (2021)

    Stańczak, A., Kubiak, I., Przybysz, A., Witenberg, A.: The possibility to recreate the shapes of objects on the basis of printer vibration in the additive printing process. Applied Sciences11(11), 5305 (2021)

    2021

  35. [35]

    International Journal of Science and Research Archive14(2), 638–645 (2025)

    Thakare, S.B., Poddar, S.: Secure mechanical design and manufacturing: Pre- venting ip theft via cybersecurity. International Journal of Science and Research Archive14(2), 638–645 (2025)

    2025

  36. [36]

    Bukkapatnam, S.T.: Cybersecurity assur- ance in the emerging manufacturing-as-a-service (MaaS) paradigm: A lesson from the video streaming industry

    Tiwari, A., Narasimha Reddy, A.L., S. Bukkapatnam, S.T.: Cybersecurity assur- ance in the emerging manufacturing-as-a-service (MaaS) paradigm: A lesson from the video streaming industry. Smart and Sustainable Manufacturing Systems4(3), 324–329 (2020)

    2020

  37. [37]

    Materials today21(1), 22–37 (2018)

    Tofail, S.A.M., Koumoulos, E.P., Bandyopadhyay, A., Bose, S., O’Donoghue, L., Charitidis, C.: Additive manufacturing: scientific and technological challenges, market uptake and opportunities. Materials today21(1), 22–37 (2018)

    2018

  38. [38]

    Yuan, J., Zhou, Y., Chen, G., Xiao, K., Lu, J.: Materials vs digits: A review of embedded anti-counterfeiting fingerprints in three-dimensional printing. Materials Science and Engineering: R: Reports160, 100836 (2024) Firewall3D 25 Appendix Detecting start/stop motion Listing 1.1 demonstrates the hardware pin toggling mechanism that signals the completion...

    2024