Advancing Circular Economy Implementation for High-Speed Train Rolling Stocks by the Integration of Digital Twins and Artificial Intelligence.

This paper presents a state-of-the-art review on the integration of digital twins and artificial intelligence to advance the circular economy and the 10R principles implementation in high-speed train rolling stock. Rolling stock generates substantial waste at the end of its service life, yet the application of the circular economy and the 10R principles (Refuse, Rethink, Reduce, Reuse, Repair, Refurbish, Remanufacture, Repurpose, Recycle, and Recover) in this domain remains limited compared with infrastructure. The review analyses 47 studies retrieved from Web of Science and IEEE Xplore, focusing on digital twin applications in railway infrastructure and rolling stock, and machine learning techniques. Findings reveal that most studies concentrate on data management and efficiency improvement, while only a few explicitly address the circular economy and 10R principles. A comparative analysis of high-waste components against current machine learning applications further highlights critical gaps. To address these, an automated workflow is proposed, incorporating digital twins, artificial intelligence, and the 10R principles to support condition monitoring and sustainable resource management. The study provides insights and research directions to enhance sustainability in railway asset management.