AI for Net Zero is integrating artificial intelligence (AI) into three projects focused on wind, road transport and aviation to accelerate decarbonization by increasing their efficiency through optimisation and employing digital twin modelling. 

Intelligent road vehicle aerodynamics


Recent statistics indicates that the transportation sector is responsible for 24% of greenhouse gas emissions in the UK, with over 80% coming from road vehicles. When vehicles move, up to 65% of the energy they use is to overcome aerodynamic drag.

The overarching goal of this work package is to employ digital twins and sparse measurements to accelerate and enable real-time active aerodynamic optimisation. This will help reduce emissions and increase the range of electric cars.  

Database from wind-tunnel experiments: A series of experiments at different flow regimes for single/multiple vehicle formations are carried out in the 10x5 wind tunnel at Imperial College London.

The wind tunnel is equipped with a rolling road to simulate the ground effect, enabling the study of the aerodynamics of the vehicle under realistic conditions. We will use Particle Image Velocimetry and pressure measurements to track and define the vehicle’s wake. This data will help us create a novel database to explore real-world models that surpass the current capabilities of Computational Fluid Dynamics (CFD) tools.  

Active intelligent aerodynamics: When attached to vehicles, aerodynamic devices can reduce aerodynamic resistance up to 9%. However, current devices aren’t efficient because they are static and configured for limited operating conditions.  

We will use state of the art reinforcement learning algorithms to discover active control strategies, helping road vehicles improve their aerodynamics in different operating conditions.  

 

George Rigas 

Principle Investigator 

George Rigas is a Senior Lecturer in Aerodynamics in the Department of Aeronautics at Imperial College London. He models and controls transitional and turbulent flows for sustainable transport with experiments, reinforcement learning, and theory. 

Isabella Fumarola 

Research Co-Investigator 

Isabella Fumarola is a Research Fellow in Flow Control at the Department of Aeronautics at Imperial College London. She works on experimental aerodynamics focusing on drag reduction and flow control. 

 

Max Weissenbacher 

Research Associate  

Max Weissenbacher is a Research Associate at the Department of Mathematics at Imperial College London. His research lies at the interface between mathematical physics and deep learning. 

 

Adaptive optimisation of wind farms

Modern large–scale wind farms consist of multiple turbines clustered together. It is difficult to understand the optimal design of a wind farm as you must consider maximising the power output whilst minimising maintenance costs. 

This work package will investigate:  

- Wake steering: we will see if the power output increases if the upstream wind turbines are misaligned. This will redirect their wakes away from downstream turbines. 

- research layout optimisation of wind turbine formations.  

 

 

Professor Sylvain Laizet

Principle Investigator

Sylvain Laizet is the Professor in Computational Fluid Mechanics in the 
Department of Aeronautics at Imperial College London.

George Rigas 

Co-lead researcher

George Rigas is a Senior Lecturer in Aerodynamics in the Department of Aeronautics at Imperial College London. He models and controls transitional and turbulent flows for sustainable transport with experiments, reinforcement learning, and theory. 

Dr Andrew Mole

Research Associate

Andrew Mole is a Research Associate at the Department of Aeronautics at Imperial College London.

 

Safe Hydrogen Operation for Sustainable Transportation

Hydrogen (H2) and its carriers - such as ammonia - have been identified as the key fuels to achieve decarbonisation. However, H2 experiments are hard to perform because H2 is stored in cryogenic conditions and is extremely flammable. This means that there are limited data sets about H2 which engineers can use to improve future systems. 

This work packages main goal is to design digital twins with surrogate modelling that extrapolate well outside current datasets. This will maximise the reliability, safety and operational efficiency of hydrogen based systems. 

 

 

Professor Konstantina Vogiatzaki

Principle Investigator

Konstantina Vogiatzaki is an Associate Professor at The University of Oxford. 

Dr Giovanni Tretola

PG Research Fellow

Giovanni Tretola is a postdoctoral research fellow at The University of Oxford.