AI Across the Nuclear Lifecycle: Current Capabilities and the Path to Decision Support

Abstract

Artificial intelligence (AI) methods are being investigated across nuclear engineering, yet evidence from individual demonstrations does not necessarily establish readiness for operational use. This work examines nuclear datasets and computational infrastructure, surrogate and physics-informed modeling, digital twins, monitoring and prognostics, lifecycle applications, optimization and control, and trustworthy AI. Reported studies show how learning-based methods can accelerate high-fidelity calculations, extract information from multivariate measurements, and support decisions in reactor operation, maintenance, waste management, and environmental assessment. They also expose recurring limitations: scarce abnormal-condition data, differences between simulated and physical systems, uncertain generalization, and incomplete evaluation of downstream decisions. We argue that progress depends on assessing complete AI-enabled workflows rather than predictive models alone. Representative data, independent testing, calibrated uncertainty, physical and operational constraints, model traceability, and human oversight are especially important when outputs affect safety-relevant decisions. Future research should connect simulation, sensing, digital twins, and uncertainty-aware decision methods while testing their performance under realistic changes in operating conditions. These steps can help establish where AI provides dependable value across the nuclear lifecycle.