Improving safety management in railway stations through a simulation-based digital twin approach

Abstract

In the dynamic milieu of modern, bustling railway stations, the emphasis on safety and efficient pedestrian traffic management has never been more crucial, elevating the significance of Digital Twins (DTs). DTs bring transformative changes in the way we approach the design, operation, control, and maintenance of complex systems, such as transportation hubs. However, documented real-world applications are lacking, leaving scientists and practitioners without valuable insights on DT implementation. Drawing from a case study centered around a prominent Italian railway station, this paper showcases the prototype of a simulation-based DT that empowers station managers with capabilities such as pedestrian flow prediction, early congestion warnings, evacuation response planning, layout optimization, and intelligent gate management. The simulation-based DT acts as a sophisticated mirror reflecting the dynamics of a station's crowd, drawing data from its physical counterpart with a certain frequency, simulating the evolution of the system over a given time horizon, returning warning messages to the decision maker if issues are identified, and evaluating ex-ante different corrective solutions. The latter becomes crucial for controlling the physical world, providing a closed-loop system. Our case-based framework validates the effectiveness of methods and technical solutions engineered ad-hoc to achieve an acceptable tradeoff among accuracy, performance, and scalability. This includes addressing challenges related to the synchronization of the virtual and physical world, data integration and interoperability with existing systems, and human-system interaction. Ultimately, this contribution serves as a valuable resource, bridging the gap between theoretical concepts and tangible applications in the realm of transportation hub management.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 64 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.