An XR-Driven Digital Twin Platform for Cybersecurity Education

Abstract

This thesis investigates the application of digital twins as an educational tool within the domain of cybersecurity, specifically targeting the infrastructure of water treatment plants. A digital twin is a precise virtual model of a physical asset, process, or system, capturing its state, behavior, and interactions in real-time. By integrating live sensor data, historical records, and predictive models, digital twins replicate their physical counterparts with high fidelity, enabling detailed simulations, monitoring, diagnostics, and analytics. This technology supports improved decision-making, predictive maintenance, and operational efficiency across industries by allowing safe testing and evaluation of modifications without altering physical assets. A case study is presented to demonstrate an immersive experiential learning platform that leverages digital twins to provide cybersecurity education. The platform aims to enhance user engagement and reinforce learning by offering hands-on experiences in a controlled virtual environment. In addition, we provide a cost-efficient hardware solution that represents the physical side of the digital twin as connecting it to the actual water treatment plant hardware is unfeasible. The study compares AI-guided learning, facilitated by a Conversational AI agent utilizing Large Language Models, against a non-AI-guided approach. This comparison evaluates the effectiveness of AI in guiding users naturally through the learning process, thereby examining the potential of digital twins to support efficient, cost-effective education across diverse sectors. The results show that presence is significantly increased with the help of an AI character while other qualities and factors remain unaffected. However, we see learning improvement overall and received positive feedback regarding the system. Users liked the digital twin concept and felt like it really helped them understand the concept thoroughly.