UK’s AI ‘twin’ will use big data to show the way to net zero

The University of Leicester is leading a new research project to develop a digital ‘twin’ of the UK that harnesses artificial intelligence and big data to meet its net zero target.

Funded by a £2.5 million research grant from UK Research and Innovation, the project will be putting artificial intelligence to use to reduce emissions from land use. 

The world must urgently reduce its greenhouse gas emissions to prevent dangerous climate change. The UK aims to achieve net zero greenhouse gas emissions by 2050. 

Emissions from cattle and sheep farming and degraded peatlands together contribute about 10% of the UK’s emissions. The researchers will work together with industry leaders and the farming community to deliver intelligent computer code that can produce advice how to reduce these emissions.

The digital twin is an up-to-date representation of the real world. It is updated based on satellite information, field instrumentation and other data on an ongoing basis. A self-learning digital twin is clever enough to improve its knowledge of how the environment works based on these data. This way, the technology will be able to inform decision making for more environmentally friendly land use.

The Self-Learning Digital Twins for Sustainable Land Management project, announced today by UK Research and Innovation, started on 1 May 2023 and runs for 23 months. The team brings together leading researchers from three research-intensive universities and the National Centre for Earth Observation (NCEO) with complementary expertise in AI, environmental research, greenhouse gas measurement and modelling, behavioural and social science. 

Lead investigator Professor Heiko Balzter from the University of Leicester and NCEO said: “The UK is not yet on track towards achieving net zero emissions by 2050. Our research in this project will make a vital contribution to new technologies that are needed to get us on a path to net zero. 

“Our team comprises not only environmental and computer scientists but also social scientists, behavioural scientists and media and communications experts. This way, we can understand what the barriers to using artificial intelligence are and how we can work with potential users to overcome them. Our industry partners will make important contributions to the project. Our team will consider how artificial intelligence algorithms can be used in an ethical and appropriate way. We aim to make our new technologies useful for companies working in the net zero realm.” 

Professor Ashiq Anjum from the University of Leicester is leading the artificial intelligence research at Space Park Leicester and will lead the digital twin investigations in the new project. 

He said: “We are pleased to see our excellent research work in digital twins is getting recognized in competitive research projects. We have been actively investigating self-learning digital twins at Leicester and this project is a great example of how we can enable computational models to co-exist with distributed machine learning approaches for solving ambitious research problems such as accurate modelling of net zero scenarios. 

“We aim to use AI to calibrate computational models as well as enable computational models to control the AI functionality. This will lead to accurate, trustworthy and agile AI that has mathematical foundations and is capable of solving most pressing challenges to achieve the net zero targets.”

Professor Matthew Rigby from the University of Bristol said: “Machine learning is providing exciting new ways to use and analyse environmental data. The new advances in this project will allow us to bring together multiple data streams in ways that have not been possible before, to learn about the UK’s greenhouse gas emissions.”

Professor Baihua Li from Computer Science at Loughborough University said: “Smart AgriTech for efficient and sustainable farming provides a challenging research agenda. We are pleased our research in machine learning and digital twin technology will help to facilitate the digital transformation of precision agriculture and address numerous challenges for wider net zero carbon ambitions. 

“Livestock is a major area in agriculture farming. Intelligent technology and innovation integrated with a digital twin will help us to gain insight of the complex agroecosystems, and improve livestock welfare and productivity associated together with soil and pasture management to reduce greenhouse gas emissions.”   

Professor John Remedios, Director of the National Centre for Earth Observation, said: “In our work on net-zero verification for the UK we are recognising that both observations and models are key to characterisation of emissions and to predictability. A digital twin in this area will capture a key interaction between food and carbon, allowing us to test not only current outcomes but future scenarios. 

“It is very likely that the drive to reach net-zero will require a number of measures and this project will provide not only a major step forward but also an exemplar for investigation of other net-zero measures. The UK can be proud of the innovative studies, such as this project, that are a sign of the scientific community’s commitment to support healthy and productive environments.” 

Project partners include software giant CGI, who are working with clients in over 400 locations worldwide to support net zero, software developer FNK Designs, information provider Geospatial Insight, the leading global satellite companies MAXAR and Planet Labs. The team will work with the Cattle Information Service on dairy cattle, and Silvasheep will provide their enterprise on regenerative agroforestry farming.