Meeting Artificial Intelligence and Machine Learning for Rail Passenger Service Planning under Competition

Project overview

This subproject studies how artificial intelligence, machine learning and operations research can improve the tactical planning of passenger railway services in competitive and uncertain environments.

It forms part of the coordinated project ACoSeM@SusTran, which brings together researchers from UPC, UCLM and URJC to develop descriptive, predictive and prescriptive models for more reliable, connected and efficient passenger and freight transport.

Research focus

The project addresses the adaptation of transport planning, management and operations to deregulated markets and irregular demand. Its main research lines include:

• modelling competition between transport operators and infrastructure access policies;
• forecasting travel demand under uncertainty using machine learning and random utility models;
• integrating artificial intelligence with robust and stochastic optimisation;
• planning railway services and allocating infrastructure capacity;
• evaluating efficient and environmentally sustainable transport strategies.

The methodological approach combines Transport Engineering models, mathematical programming, stochastic and robust optimisation, machine learning, artificial intelligence and demand models supported by new data sources.

Expected contribution

The project seeks to provide decision-support methods for transport operators and infrastructure managers. These methods are designed to improve service planning, anticipate passenger behaviour and account for competition, uncertainty and rapidly changing operating conditions.