#16 AI: The Next Big Use Case for Rail

About Digitale Schiene Deutschland Digitale Schiene Deutschland (DSD) is a sector initiative aimed at transforming the German rail network through digitalization. DSD’s primary goals are to increase rail capacity, enhance service quality, and improve efficiency across the network without extensive physical infrastructure expansion. Key components include the rollout of the European Train Control System (ETCS) for digital signalling, automated train operations, a vehicle-centric safety logic for driving at optimal headways, and highly automated, AI-driven planning and real-time traffic management. By developing and integrating advanced technologies, Digitale Schiene Deutschland aims to create a more resilient, efficient, and sustainable rail system that supports Germany's environmental and mobility goals. The deployment of these systems is expected over the next decade. https://digitale-schiene-deutschland.de/de About Michael Küpper Michael joined Digitale Schiene Deutschland (DSD) at Deutsche Bahn in 2017. As Product Manager, he has built and led the scaled-agile team-of-teams that implements DSD’s Capacity & Traffic Management System (CTMS). In parallel, he has led the implementation of agile and self-organized principles across several innovation departments at DSD. Since 2023, Michael has served as Stakeholder Manager to drive the strategic vision of CTMS and its enabling technological foundations within the railway sector at large. Michael holds a PhD in physics from The Weizmann Institute of Science in Israel and has over 10 years of experience as strategy and management consultant. Throughout his career, he has introduced Artificial Intelligence (AI) in environments, where AI had not been previously applied, from particle physics analysis to housing price prediction to rail traffic management. Key Takeaways Focus on “Weak AI”: Current applications rely on weak AI to address specific problems, while strong or “General AI” remains a distant possibility. Two Key AI Use Cases: In addition to predictive maintenance (e.g. for rolling stock) and similar applications of pattern recognition, active traffic management is emerging as a significant new application of AI in the rail industry. Active Traffic Management: This approach allows for quick decision-making in complex environments, enabling disruptions to be managed efficiently—new schedules can be generated and implemented in minutes or less. Holistic Approach: The CTMS will take a comprehensive view, considering multiple factors like speed, energy consumption, and connections. Long-Term Implementation: While AI can already support planning, a fully implemented CTMS is expected to take another 5-10 years. High Failure Rate of AI Projects: Up to 85% of AI projects fail. Costly AI Projects: AI initiatives are often expensive to execute. Focus on Automation: The goal should not be AI itself, but rather automation and new processes. Data Challenges: A key hurdle for AI projects is cleaning and consolidating data from outdated systems. It often takes a longer breath than many management teams or investors are ready to accept. The Answer of AI Is AI: The phrase suggests that AI is often seen as the go-to solution for various challenges, including the ones created by AI itself Sebastian Sperker www.railup.club [email protected]