A Review of LifecycleOriented Approaches and CrossSystem Integration Challenges

Abstract

The transition toward new energy vehicles introduces fundamental challenges in component reliability, as degradation mechanisms increasingly involve coupled thermal, mechanical, and electromagnetic processes that are not adequately captured by conventional single-domain approaches. This review synthesizes recent progress across three intersecting dimensions: multi-physics degradation mechanisms in battery systems, power electronics, and structurally integrated glazing; physics-based and data-driven prognostic modelling strategies for remaining useful life prediction; and the integration of reliability outputs with maintenance logistics and supply chain operations. The analysis reveals that while individual prognostic methods achieve reasonable accuracy under controlled conditions, their translation into operational decisions remains constrained by gaps between component-level predictions and system-level logistics constraints. Particular challenges arise in parameter identification for physical models, data scarcity for novel failure modes, and coordination of maintenance planning under supply chain disruptions. The review concludes that future progress depends less on algorithmic refinement and more on establishing shared validation benchmarks, uncertainty-aware decision frameworks, and cross-disciplinary collaboration between reliability engineering and operations research.