Schneider, Stefan (DB InfraGO AG), Ramesh, Anirudha (InstaDeep Ltd), Roets, Anne (DB Systel GmbH), Stirbu, Ciprian (InstaDeep Ltd), Safaei, Farhad (DB InfraGO AG), Ghriss, Faten (InstaDeep Ltd), Wülfing, Jan (DB InfraGO AG), Güral, Mehmet (DB InfraGO AG), Siboni, Nima (InstaDeep Ltd), Gentry, Rick (InstaDeep Ltd), Liessner, Roman (DB InfraGO AG), Hustache, Thomas (InstaDeep Ltd), Thomas, InstaDeep Ltd (69385), Umashankar, DB Systel GmbH (69386), Valerii, InstaDeep Ltd (69387), Victor, InstaDeep Ltd (69388), Wissam, InstaDeep Ltd (62515), Michael, Deutsche Bahn (62423), Irene, DB Netz ()

Intelligent Railway Capacity and Traffic Management Using Multi-Agent Deep Reinforcement Learning

Scheduled for presentation during the Regular Session "Rail Traffic Management I" (ThAT7), Thursday, September 26, 2024, 10:30−10:50, Salon 15

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 7, 2024

Abstract

A more attractive future railway system needs to offer more capacity in the railway network and improve quality and punctuality. A fundamental centerpiece of future digitized railway network operations is automated and optimized planning and dispatching. The sector initiative “Digitale Schiene Deutschland” (DSD) develops a holistic and intelligent Capacity & Traffic Management System (CTMS) that can automatically plan and continuously optimize railway traffic at scale. Both, planning and dispatching tasks, are highly complex and, today, require human expertise and oversight.

Our main contribution is a multi-agent deep reinforcement learning approach at the core of the envisioned CTMS, which learns from interaction with a realistic, microscopic railway simulation. Our results demonstrate that the proposed approach flexibly solves planning and re-scheduling tasks in the realistic setting of a medium-sized part of the German railway network. It exhibits response times and scaling properties that make it a promising candidate for future applications in railway operations at scale.