How do you know if your 3D printed part is exactly what you expected? What are the possible ways that a part could be affected by sabotage? Malicious cyber activity can occur within the 3D printing workflow to modify CAD files, G-Code or 3D printers. But what do these sabotaged parts look like and can we even detect the malicious changes before we suffer the consequences of the bad part?
Part integrity in 3D printing describes the printed part having the expected structure, features and properties with no corruption. Corruption can be malicious or unintentional due to user or equipment error. It is the ability of the part to withstand its intended loading without failing. Verification and Validation can only take us so far into detecting a loss of part integrity. A robust cybersecurity solution is needed to detect malicious activities leading to part defects.
3D printed parts can be sabotaged to cause defects in the part in different ways including:
Geometry: External geometry changes can be readily detected by quality control checks but internal geometry and wall thickness changes are not so easy to identify.
Voids: Voids are small cavities within the 3D printed part that reduce strength and are challenging to detect. A popular example of this is the drone experiment where undetectable voids were created in the propeller which caused the drone to drop from the air in mid-flight.
Infill: Changing the infill density, infill pattern or shell (outlines or outer perimeters of each printed layer) width alters the part strength and weight. Reducing part infill by as little as 10% can have a large impact on part strength.
Print orientation and Anisotropy: How a part is oriented on the print bed matters. Anisotropy refers to the property of a material or part which has different properties in different directions. FDM 3D printed parts are anisotropic due to the way they are created with layer by layer material deposition. They exhibit different properties in x, y and z axes. If a part is printed in an unintended orientation it may not exhibit the required properties and this could cause a system failure where that part is used
Material modification: As we look to the future of 3D printing, multi material printers will become more common. A part printed in an unintended material will change the part’s properties. In the near future we can expect to see more 3D printed electronics using multi material 3D printers and it is imperative that we secure these processes.
There are methods to perform quality assurance before and after the print process such as CT scans and x-rays but there is a lack of visibility from a cyber perspective. Current quality checks can not detect every type of part defect. Quality assurance checks do not give the user any information about the source of the loss of part integrity. Was it malicious activity or was it user error? The user may not know. It is imperative that 3D printing processes are secured and protected from malicious actors to prevent covert sabotage of critical parts.
What can we do from a cybersecurity perspective to validate that the manufactured object matches the initial G-Code definition? BreakPoint Labs, within it’s BISON capability, has developed a robust last-link monitoring process that passively monitors the USB-based communication link from IT infrastructure and AM devices. The custom monitoring solution reconstructs the G-Code transmitted over the wire and compares that to the original, trusted file and identifies deviations. This method of monitoring is very difficult to subvert from an attacker perspective and can provide fine-grained insight to AM operators as to why and when deviations have occurred. BPL’s BISON is an AM cybersecurity solution that can detect malicious cyber attacks, helping the 3D printer user know that their part has been compromised.