Michael Beverley Innovation Fellowship

Future Railway Traffic Management System

Railway industry is currently undergoing a transformative phase, with a focus on digitalisation, automation, and systematic upgrades. Modern data science is playing a pivotal role, as the entire rail sector eagerly embraces data-driven solutions to tackle complex and longstanding problems. Based on past projects established by the candidate and his group, a train rescheduling tool using a data-driven algorithm (i.e., reinforcement learning) to handle short-term requests from rail operator was developed. To further commercialise this tool, we recognise the need for a common data model (CDM) that can generalise the algorithm to a larger network and ensure its scalability. However, the railway sector faces a significant challenge in developing and deploying such a CDM due to the lack of knowledge regarding potential traffic management applications in the future.

This fellowship presents a unique opportunity for the candidate to enhance connections with industrial partners and collaborate with them to identify and develop a CDM. If successful, both the established and ongoing research could be easily validated in practrice and commercialised. Additionally, it would enhance efficient and streamlined data exchange between various railway units and pave the way for a standardised structure for future data-driven application deployments in railway traffic management system.