CAST-Map: Cybersecurity Assessment Using Spatio-Temporal Mapping of Connected Electronics

Major industries such as additive manufacturing and semiconductors fabrication are rapidly moving towards Industry 4.0 setup and beyond. However, this progress is being impeded by mounting cybersecurity concerns that can lead to design/data tampering, data theft, and denial-of-service. Adding to this problem is the complexity and diversity of the digital infrastructure that is necessary to realize the Industry 4.0 dream. A cybersecurity risk analysis and mitigation of an Industry 4.0 system is not possible without first knowing the exact topology of all the electronic hardware, software, computing, sensors, and edge components that are being used in the said Industry 4.0 system. This is a particularly challenging task due to the diversity of hardware/software and the dynamic nature of these systems. In this article, we propose an automated framework (CAST-Map) that can tackle this challenge by creating a live/dynamic spatio-temporal map of all the cyber/digital components in an Industry 4.0 system. We also propose a novel unsupervised learning guided scheme for computing the system's cybersecurity risk/criticality value by leveraging these spatio-temporal maps. We implement & validate the effectiveness of this framework using a Industry 4.0 testbed & a randomized large scale system generation method.