1.4m grant for AI to help achieve net zero in energy and transport

Researchers have been awarded £1.4 million to lead a project using AI to help the UK-s energy and transport sectors achieve carbon net zero.

The funding from the Engineering and Physical Sciences Research Council (EPSRC) was announced by Secretary of State for Science, Innovation and Technology, Chloe Smith, as part of a £54 million government investment to develop trustworthy artificial intelligence (AI) research.

Collaborative venture

The project brings together academics from the universities of Cambridge, Oxford and Edinburgh under the overall leadership of Dr Georgios Rigas from Imperial College London’s Department of Aeronautics. It will develop exact virtual replicas of aeroplanes, wind farms and lorries and test adaptations to these in a range of aerodynamic conditions.

Dr Rigas said: "Our project will harness machine learning to solve currently intractable problems in wind energy, hydrogen combustion, and road transportation, helping the UK towards its objective of becoming a net zero economy. This is a truly collaborative venture between physicists, aeronautical engineers, mathematicians, politicians and industry leaders."

Developing ’digital twins’

The most significant bottleneck of most scientific machine learning (a form of AI) is that systems need time to be re-trained offline when new data becomes available. In this project, the interdisciplinary team will work together to develop real-time, sustainable and robust ’digital twins’ - exact virtual replicas - that learn from both emerging data and prior knowledge and use these to make refinements and adaptations. Using AI algorithms, researchers will be able to test the digital twins in a range of aerodynamic conditions and predict in real-time how they would perform in real life.

Data will be provided by high-fidelity simulations and experiments, from state-of-the-art UK facilities and software - including the National Wind Tunnel facility at Imperial. Researchers will work to maximise the efficiency of machine learning training so that the algorithms require minimal energy and therefore produce minimal emissions.

Influencing policy

The researchers will share open-source software with policy-makers and industry to enable fast decision-making and adaptations to wind farm layouts, road vehicle aerodynamics and optimal use of hydrogen in aviation.

An equally significant contribution to the community will be the ideas developed in this work (real time optimisation via digital twins) which could potentially be applied to other fields of research.

Sir John Aston, Harding Professor of Statistics in Public Life at Cambridge University, who is policy lead for the project explained: "A project such as this has the potential to be of considerable relevance well beyond academia. As part of the project. we’re carrying out research to examine the integration of climate science communication, policymaking and the AI research we will deliver, to ensure we develop best practice in these areas, allowing our work to make a difference."