SIGNAL+DRAHT | Ausgabe 11/2025
AI-Driven traffic management: unlocking railway capacity through deep reinforcement learning
Given the ever-growing passenger demand (PPHPD: Passenger Per Hour Per Direction), the complexity of route networks, and the tight schedules, etc., most railway systems are currently operating at full capacity, sometimes even beyond capacity. This means that even the slightest delay can have a significant impact on the network, which has no room for manoeuvre. Nowadays, operators are struggling to adapt their transport services as the ridership patterns of the past are less relevant today, thereby making it essential for them to frequently update their schedules....