Dynamic Mission Risk Management via Adaptive Information Acquisition and Online Uncertainty Calibration

Mission reliability and system survivability are jointly shaped by how risk management policies respond to evolving degradation and uncertainty in health information. Existing studies, however, predominantly rely on fixed information-acquisition schedules and pre-specified degradation trajectories, which limits the ability to adapt risk decisions to operating environments and heterogeneous degradation. This paper develops a dynamic mission risk management framework that integrates adaptive information acquisition with online uncertainty calibration. Specifically, the framework dynamically determines when additional degradation information should be acquired according to the current degradation state and mission progress, while sequentially updating uncertain degradation parameters through parameter learning. Under the specified random-effects Wiener model with exact observations, we establish that the latest degradation observation and its acquisition time are sufficient for Bayesian posterior updating. This property supports adaptive information acquisition, while intermediate observations remain valuable for timely risk assessment and intervention. A key insight is that the resulting optimal mission risk management policy admits a tractable threshold structure characterized by the system degradation level and mission progress. Exploiting these structural properties, an efficient solution algorithm is developed to streamline policy computation. Comparative experiments based on shipborne radar-driver degradation data demonstrate that the proposed framework consistently reduces risk-related losses relative to benchmark policies, highlighting the value of jointly adapting information acquisition and uncertainty calibration for dynamic mission risk management.