Operations Research and Artificial Intelligence for Airline Operations
PhD candidate: Antonio Montaruli
The PhD project in operations research and artificial intelligence for airline operations develops data-driven methods to support airline decision-making across strategic, tactical, and commercial planning levels. It is motivated by the increasing complexity of air transport, where volatile demand, operational constraints, congestion, disruptions, and sustainability pressures make planning more challenging. In practice, decisions such as network design, schedule development, fleet deployment, and revenue management are often addressed sequentially, which can lead to suboptimal system-wide outcomes. Conducted in collaboration with KLM Royal Dutch Airlines and the University of Twente, this research aims to develop models that better capture these interdependencies and provide more integrated decision support.
The project has two main research streams. The first focuses on network planning and schedule development, especially in hub-and-spoke and multi-hub systems. It develops optimization-based models that capture supply-demand interactions, competition, connectivity, and network spillovers, supporting decisions on routes, frequencies, fleet assignment, and schedule structure. The second stream addresses airline revenue management, exploring Deep Reinforcement Learning to learn pricing and seat-control policies in network settings and compare them with traditional benchmark approaches.
A common denominator across both research streams is the presence of network effects: decisions on one flight leg influence the attractiveness, demand allocation, and profitability of multiple connected itineraries across the system. The project, therefore, seeks to generate both methodological contributions and practical insights for airline planning.
External supervisor: Matthijs Kieskamp (KLM Royal Dutch Airlines)