Research Seminars @ Illinois

Railway systems have long served as the backbone of regional and national transport, enabling the movement of people and goods while shaping communities and economies. Today, these systems are evolving rapidly with advances in sensing technologies, digitalization, and multi-scale computing at a remarkable speed. This seminar will introduce three cases demonstrating key AI advancements, enabled by multi-data fusion and high-fidelity machine learning frameworks grounded in real-world deployments: 1) Agent-based learning for urban rail transit; 2) Infrastructure multi-sensing for rail track; 3) Condensable failure predictions for rail wheelset. The work also proposes scalable data governance and computing pipelines designed to operate efficiently under extreme data volumes typical of modern rail systems. The seminar highlights how AI-enabled approaches can be embedded into multi-scale operations and maintenance. The seminar will also discuss data and computing challenges, and workable practice innovations tackling these challenges.