Probabilistic System-Level Prognostics for Reliability and Risk Assessment in Aerospace Structures

System-Level Prognostics (SLP) is essential for ensuring mission success, as it focuses on predicting the System Remaining Useful Life (SRUL) rather than individual subsystem RULs. Unlike subsystem-level prognostics, SLP must account for degradation interactions and quantify inherent uncertainty. In aerospace structures, however, system-scale failure data are scarce due to cost and safety constraints, limiting the applicability of existing methods. This paper introduces the RUL Inoperabilities Model (RIM), a probabilistic framework inspired by the Inoperability Input-Output Model (IIM) that operates directly on coupon-level RUL predictions. The approach is prognostic model-agnostic, interpretable, and data-efficient, requiring only coupon-level training data and a single componentlevel degradation history to capture interdependencies. RIM propagates probabilistic coupon-level predictions to the component level, providing uncertainty-aware SRUL estimates without additional full-scale testing. The framework is validated using a component composed of three interconnected open-hole aluminum coupons, with a Hidden Semi-Markov Model (HSMM) used as the base predictor. Results show that RIM improves SRUL accuracy and uncertainty calibration compared to a naive aggregation approach. Beyond RUL-prognostics, the probabilistic SRUL outputs are used to perform conditional reliability and risk-based working subsystem analyses, enabling assessment of both remaining life and functional degradation. Overall, RIM provides a scalable and reliabilityinformed solution for system-level prognostics under limited data in the scale of interest.

Keywords: System-level prognostics, remaining useful life, uncertainty quantification, reliability, risk assessment.