Digital Twins for Automotive Cybersecurity Engineering and Validation, Abstract
One of the global trends in digitising industry is to connect the physical assets of the shop floor with the existing enterprise information systems. This trend increasingly includes Cyber Physical Systems (CPS) and Digital Twins, thus expanding the digital representation of assets into executable models of manufacturing processes and workflows. Concurrent with the above trend, network-centric computing has evolved
into cloud computing enabling large-scale data analytics and Machine Learning (ML), as prerequisite for desirable new solutions for autonomous decision making, adaptation and self-management of industry 4.0 systems, predictive and intelligent forecasting capabilities. With the huge increase in connected devices and their business processes and workflows, comes a commensurate rise of cybersecurity risks and challenges. Traditional security solutions have proven inadequate when used in cloud environments as has been seen with a number of SCADA based industry systems. Therefore, strengthened cybersecurity for cloud services has become a necessity, which requires novel design solutions with respect to the modelling of cybersecurity threats and countermeasures. In our research carried out in the IoT4CPS project (Trustworthy IoT for Cyber-Physical Systems) we investigate the concept of a Digital Twin as a means of validating security and functional safety measures, and their interplay with physical assets in Industry 4.0 settings. Our motivation is twofold: on the one hand, it is about designing and implementing a demonstrator for the validation of our methods, and on the other hand, we aim to augment current security- and safety-critical industrial systems by taking a data-centric approach to asset identification and monitoring, decision-making and virtualization.