Potential applications of artificial intelligence in spent nuclear fuel reprocessing research: A review

Spent nuclear fuel reprocessing plays a pivotal role in achieving a closed nuclear fuel cycle. However, traditional research approaches face significant obstacles due to the highly radioactive environment, limited availability of experimental data, and the inherent complexity of the underlying mechanisms. Artificial intelligence (AI), with its strength in extracting intricate patterns and making predictions from sparse data, offers promising solutions to these longstanding challenges. Recent advancements have demonstrated the potential of AI in diverse domains, including image recognition, process modeling, materials discovery, process control, data analytics, and safety assessment. By addressing the limitations of conventional methods in data handling and mechanistic modeling, AI enhances system safety and operational efficiency under extreme conditions. Furthermore, it complements traditional physics-based approaches, paving the way for transformative research paradigms. As the volume of accessible data grows and algorithms continue to evolve, AI applications in spent fuel reprocessing are expected to expand—enabling fully integrated, self-optimizing control systems and proactive safety monitoring. These developments promise to significantly reshape the landscape of nuclear fuel reprocessing technologies.