TL;DR
This paper provides a comprehensive security comparison between additive manufacturing (AM) and traditional subtractive manufacturing (SM). It introduces a framework for analyzing cyber-physical attacks on these systems, highlighting that while many attack vectors overlap, AM's unique characteristics like its dynamic supply chain and open access to blueprints create distinct vulnerabilities, particularly in intellectual property theft and sabotage.
What Problem Does This Paper Address?
The rapid growth of additive manufacturing (AM), also known as 3D printing, has raised significant security concerns. As AM becomes increasingly prevalent in critical infrastructure sectors like aerospace and automotive, understanding its security landscape is paramount. However, the security implications of AM have not been thoroughly compared to traditional subtractive manufacturing (SM), which has been the industry standard for decades. This paper aims to bridge this gap by analyzing the workflows, actors, and potential attack vectors of both AM and SM to identify their similarities and differences from a security perspective, with a particular focus on the emerging threats posed by AM.
Methods
The authors employ a comparative analysis of additive and subtractive manufacturing workflows. They propose an "attack analysis framework" to systematically evaluate potential threats. This framework breaks down attacks into key components: attack vectors, compromised elements, manipulations, and effects/attack targets. The paper then applies this framework to both AM and SM, identifying commonalities and unique vulnerabilities. They categorize threats into three main areas: theft of technical data, sabotage, and illegal object manufacturing.
Datasets
Not discussed in the paper.
Key Insights and Ideas
- AM's unique attack surface: While both AM and SM are cyber-physical systems and share many attack vectors (e.g., hacking, social engineering), AM's decentralized nature, dynamic supply chains, and readily available blueprints (especially for 3D printing) create new avenues for adversaries.
- Intellectual Property (IP) is defined differently: In SM, IP is largely tied to the design files. In AM, IP extends beyond geometry to include material properties and manufacturing process specifications, making IP theft more complex and encompassing.
- Sabotage is more nuanced in AM: AM allows for more subtle sabotage, such as altering internal geometries, introducing voids, or manipulating material properties during the printing process itself, which can be harder to detect than in SM.
- Illegal manufacturing is democratized: The increasing accessibility of high-quality AM equipment and readily available blueprints online significantly lowers the barrier to entry for manufacturing illegal items, a threat that is less pronounced in SM due to the specialized and proprietary nature of its equipment and designs.
- Immaturity of AM systems: The relative novelty of AM means its software and firmware are often less mature and more vulnerable to cyber attacks compared to established SM systems.
Key Results
The paper identifies that while many attack vectors are common to both AM and SM, the impact and probability of certain attacks differ significantly.
- Theft of Technical Data: While both can suffer from IP theft, AM's broader definition of IP (including material properties and process specs) makes it a richer target.
- Sabotage: AM is more susceptible to subtle sabotage that degrades part quality or even damages equipment, due to factors like material recycling and the ability to manipulate parameters during the build process. Specific examples include introducing voids or altering material composition.
- Illegal Object Manufacturing: This threat is amplified in AM due to the increased accessibility of equipment and blueprints, making it easier to produce items like firearms.
Limitations and Future Work
The authors acknowledge that the research literature on attacks specifically targeting subtractive manufacturing systems is limited, making direct comparisons sometimes challenging. They also note that while the framework is presented, a full empirical evaluation of all identified threats is beyond the scope of this chapter.
Future work could involve:
- Developing more detailed attack scenarios and mitigation strategies for AM.
- Conducting empirical studies to validate the identified vulnerabilities and the effectiveness of proposed defenses.
- Further exploring the legal and ethical implications of AM security, particularly concerning intellectual property and the manufacturing of illegal items.
- Investigating the security of emerging AM materials and processes.
Summary updated