Modeling and Safety Analysis of Automated Vehicle Teleoperation with Dynamic Fault Trees

Teleoperation is emerging as a key enabler for deploying automated vehicles in safety-critical environments where full autonomy is not yet feasible. Ensuring safety in such systems is challenging due to their reliance on, for example, human operators, wireless communication networks, mixed-traffic interactions, and onboard automated systems, as well as the interdependencies among these components that can trigger cascading failures. To address this, we present a case study on the safety analysis of teleoperated driving in underground mines. We model the teleoperation architecture using a Dynamic Fault Tree (DFT), an extension of traditional static fault trees with explicit modeling support for functional dependencies, order-dependent failures, and redundancy mechanisms. The DFT model is rigorously evaluated through probabilistic model checking across a broad range of system-wide metrics, including degradation behavior and criticality of components. The results provide insights into system reliability, the sensitivity of failure probabilities with respect to different component failure rates, and the behavior of the system in fully functional, fail-operational, and degraded modes. Overall, the findings highlight the critical components that dictate system dependability and establish a structured basis for enhancing the safety of teleoperated vehicle systems to comply with industrial safety standards.