The use and influence of Artificial Intelligence is growing every day and shaping industries worldwide. It appears to be able to do so much for us, but is it just the latest technological trend, or a valuable and worthwhile tool?
Our flagship AI research programme, Project BlueBird, is exploring the practical value of AI in Air Traffic Management (ATM) and its potential to support our future operations.
UK airspace is extremely complicated; among the most complex in the world, so this research focuses on how AI could help in managing uncertainties, such as adverse weather conditions, which our controllers face every day.
Project BlueBird is a collaboration between NATS, The Alan Turing Institute and the University of Exeter, which allows us to draw on and combine our collective expertise in AI, machine learning and air traffic control. The primary objective of the project is to enhance our understanding of airspace efficiency, thereby facilitating the development of new support tools, training programmes for controllers, and eventually test new airspace designs and procedures.
Over the past 2 years, our efforts have been concentrated on developing AI agents capable of simulating the control of a virtual UK airspace in a ‘digital twin’ – the first digital twin of its kind. This digital twin technology enables us to conduct tens of thousands of realistic simulations in a fraction of the time it would have taken previously.
By working closely with our air traffic controllers we have been able to apply their knowledge to the AI agents, meaning they are now capable of controlling small amounts of air traffic. This collaboration with active controllers is vital, as it allows us to explore how AI-driven tools could complement and support human controllers effectively.
Last month, we participated in AI:UK, the UK’s national showcase of data science and AI technologies and how they can be used to solve real world challenges. We showcased a gamified version of Project BlueBird’s digital twin to demonstrate how one of our AI agents can control traffic. Alongside this, attendees were invited to attempt to control the same air traffic, pitting themselves against the AI agent.
The BlueBird team demonstrating the AI agents at AI:UK
As well as attending AI:UK, the BlueBird team has also recently conducted outreach workshops with other safety critical industries. Participants from Eurocontrol, National Grid and the Transport Research Foundation engaged in extensive discussions about technology development, trustworthiness, ethics and regulations, exploring the applications and implications of AI across various sectors.
Project BlueBird demonstrates the potential use of AI in Air Traffic Management, but also underscores the ongoing learning curve. By further refining our understanding, we hope to develop innovative new support tools that can enhance safety and efficiency in airspace operations. So, trend or useful tool – the answer to that question is still evolving, but we are excited about the potential it brings.
Comments
Please respect our commenting policy and guidelines when posting on this website.