Reliability assessment in complex technical systems often involves capturing the interdependency between multiple subsystems and the gradual loss of their functional effectiveness. This challenge becomes particularly critical in the context of unmanned aerial vehicles (UAVs), where sustained operational capability is essential for the safe execution of autonomous missions. This work presents a state-based methodology to estimate the future mission capability of UAVs subject to progressive component degradation. To generate a representative dataset, numerous simulated flight missions are conducted with randomly varying degradation profiles across multiple actuators. A hidden semi-Markov model (HSMM) is trained on this data to characterize the progressive reduction in system performance over time. To improve model tractability and generalization, raw flight data is first reduced to a concise set of performance-related parameters, from which critical sensor signals—such as roll, pitch, yaw, horizontal and vertical airspeeds, and current consumption—are estimated. The proposed approach enables the identification of system-wide dependencies between degradation patterns and mission-relevant behavior. By linking current operational states to the likelihood of meeting future performance requirements, it offers a quantitative basis for predictive reliability assessment. Looking ahead, this framework may support extensions toward forecasting sensor trends and in-flight anomaly detection, contributing to more anticipatory and resilient UAV operations.