Enhancing System Safety and Reliability in Aviation: A Hybrid Framework Combining Prognostics, Health Management, and Probabilistic Safety Analysis

Abstract

The growing complexity of modern aircraft systems has necessitated integrating predictive and probabilistic safety approaches to ensure continuous airworthiness and operational resilience. This review critically examines the hybridisation of Prognostics and Health Management (PHM) and Probabilistic Safety Analysis (PSA) as an emerging framework for enhancing system safety, reliability, and maintenance efficiency in aviation. The study aims to evaluate the state-of-the-art methodologies, identify integration challenges, and highlight future research and policy directions for PHM–PSA synergy. Methodologically, an extensive review of recent peer-reviewed literature (2018–2026) was conducted, synthesising findings from key studies on Bayesian networks, digital twin architectures, dynamic risk modelling, and machine learning–based prognostics. Comparative analysis reveals that hybrid PHM–PSA models outperform conventional safety systems by enabling real-time risk estimation, adaptive maintenance scheduling, and lifecycle cost optimisation. Integrating probabilistic reasoning with predictive health indicators enhances early fault detection, reduces downtime by up to 40%, and supports regulatory compliance through data-driven safety assurance. However, the review identifies persistent challenges, including data heterogeneity, model interpretability, and certification barriers. It concludes that PHM–PSA integration marks a paradigm shift toward predictive, intelligent, and risk-aware aviation systems. The paper recommends establishing standardised data protocols, regulatory sandboxes, and digital twin infrastructures, alongside capacity-building initiatives to bridge the gap between innovation and certification. Collectively, these strategies will enable a globally harmonized, resilient, and proactive aviation safety ecosystem.