Search for a command to run...
This deliverable summarizes the outcomes of Task 6.1 and provides a comprehensive review of the current landscape of Data Science, Machine Learning (ML), and Artificial Intelligence (AI) applications in the context of future metro operations. It introduces and contextualizes the foundational concepts of pillars of Work Package 6 — and explores the anticipated impact of the European Union AI Act on the development and deployment of AI-related use cases within this domain, as outlined in Section 2.Drawing on extensive desktop research, the deliverable reviews a curated selection of key industry reports and publications focused on the role of AI and data science in public transport. It identifies and analyses relevant trends, existing applications, and use cases that have emerged in the sector. Furthermore, it examines the current state-of-the-art in metro operations and AI integration from the perspective of a train manufacturer, offering insights into ongoing innovation and technological advancements. This analysis is complemented by the findings of two industrial workshops — one conducted with a train operator and the other with several European metro operators — which provided practical, real-world perspectives and industry needs related to AI implementation (as covered in Section 3). Building on this foundation, the deliverable provides a detailed examination of four representative AI use cases in metro operations: crowding prediction, demand forecasting, timetable creation support, and anomaly detection, with a specific example focused on uncleanliness detection. These use cases, grounded in existing academic and industry literature, illustrate the potential of AI technologies to enhance operational efficiency, passenger experience, and service reliability (Section 4). To move from concept to practice, these use cases are further developed through detailed implementation concepts, considering architectures, data requirements, technical challenges, and potential integration pathways (Section 5). Finally, the deliverable concludes with a synthesis of lessons learned, highlighting best practices for successful AI adoption in metro systems and identifying key technical, organizational, and regulatory challenges that must be addressed to ensure effective deployment (Section 6).