Answers are grounded in this paper and cite their sources.
Maintenance decision-making in complex infrastructure systems is inherently conducted under uncertainty due to stochastic degradation processes and imperfect condition observations. Although Condition-Based Maintenance (CBM) has been widely adopted to improve maintenance efficiency, many implementations still rely on fixed decision thresholds that do not explicitly reflect organizational risk preferences. Consequently, maintenance policies often remain weakly connected to strategic maintenance governance. This study proposes a risk-sensitive maintenance framework that integrates stochastic degradation dynamics, failure penalties, and organizational risk sensitivity within a unified decision structure. A Monte Carlo simulation environment is developed to analyze system behavior under uncertain degradation and to evaluate maintenance policies using a Risk-Weighted Availability (RWA) metric. The results demonstrate that the optimal CBM threshold varies systematically with the level of risk sensitivity. As the penalty associated with failure increases, maintenance interventions occur earlier, leading to more conservative maintenance strategies. The analysis further reveals a structured policy landscape in which the relative performances of CBM and Time-Based Maintenance (TBM) change across different risk regimes. More importantly, the results reveal that the optimal intervention threshold is an endogenous function of organizational risk sensitivity, θ*(λ). As risk sensitivity increases, the optimal threshold decreases systematically, triggering earlier preventive maintenance actions and progressively shifting maintenance governance toward more conservative policies. These findings demonstrate that maintenance thresholds should be interpreted as strategic governance variables rather than fixed engineering parameters, thereby providing a theoretical bridge between stochastic condition monitoring and risk-sensitive maintenance governance in complex railway systems.
No takes yet. Share an insight, caveat, or question.
Kyoung-Soo Seo (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: