Transforming Air Traffic Control: Sequence Prediction and Modelling with Large Language Models.

Recent advances in sequence prediction with large language models have opened up new possibilities in various domains. Building upon these cutting-edge methodologies, this PhD project aims to extend their application to the realm of air traffic control (ATC).

The project seeks to develop an advanced air traffic control agent utilising a unique and rich dataset comprising over 20-years of interactions between air traffic controllers and pilots. The extraordinary ability of these models to generate new, emergent behaviour beyond their training data holds promise for developing an ATC agent that can potentially perform beyond human capabilities.

Additionally, it offers opportunities to devise novel strategies for conflict resolution and traffic flow management, leading to improved efficiency and safety in air travel.