Enhancing Cyber-Physical Security: Integrating Virtual Fences Within Digital Twins

Virtual fencing exploits the transformation of a dynamic environment into a cyber-physical traceable system to restrict dangerous movements and create early warning zones. While most real-life applications rely on GPS systems, this article explores the use of computer vision to regulate the distribution of human crowds. The proposal includes utilizing Unmanned Aircraft Systems (UAS) to create risk-based ranked early warning zones to enhance the resilience of critical infrastructures and citizen safety/security in smart cities. The proposed system conducts surveillance of labeled areas comprising infrastructure and/or populated by citizens. In the current research, a phys-ical testbed is constructed to monitor human mobility, which simulates human behavior as a part of the cyber-physical-social dynamic environment, triggering an alert whenever a person enters a warning zone. These warning zones are categorized based on their level of risk, ranging from low to high depending on the distance from the risky area. The different warning levels can be dynamically set to leave sufficient time to react and mitigate potential hazards. The experiments to simulate the crossing of various virtual fences in the testbed demonstrate the system's high efficiency for early warning purposes. Finally, the main challenges of bringing EWZ into action are identified and discussed.